Note: Before we start, I want to note that as I've worked through this period of development, I've slightly changed how I'm defining this section. I initially said that Period 4a consisted of every Pokemon created before May 1999; however, that turned out to be a difficult marker to always quantify, and it also led to a very large section that didn't properly distinguish between Pokemon created at the beginning of the 1999 reboot and those created later. As a result, I've moved to a slightly different definition: Period 4a consists of every Pokemon which predates our first snapshot of rebooted development, on April 1st, 1999. This ends of leading to slight reshuffling of the Pokemon I've already covered, and of the list that I provided at the end of Era III. Most notably, Sentret and Furret should be the first two Pokemon in Period 4b, as we know that they were first given stats twelve days after April 1st, on April 13th. Likewise, I'm going to begin this section by discussing Sunkern, but looking through the data, it becomes obvious that some version of Sunkern existed very early in the reboot, potentially even predating 1999. As a result, it should probably be earlier in Period 4a, around Heracross. I'm sure that I'm the only one who cares all that much about the timeline I'm trying to construct of the creation of these reboot-mon, but I want to do my best to be as transparent as possible when I change my mind about things. When I eventually create the chart of Generation II development that includes these guys, these changes will be included. ID X16: SunkernSunkern, the seed Pokemon we all know and love, turns out to have had a surprisingly bumpy development. It appears that the team created and mostly finished Sunkern, then decided they didn’t like the idea, completely wiping Sunkern from the data. However, late the day Sunkern reappeared, ready to take its rightful place in the Pokedex. They buried Sunkern, but they didn’t know it was a seed. Obviously Sunkern’s origins are straightforward. It looks like a seed; specifically, the stripes on Sunkern make it look like a Sunflower seed. Which, of course, makes perfect sense, as it was created to be the pre-evolution of Sunflora, the Sunflower Pokemon. It’s Japanese name, “Himanuts” is just a portmanteau of Himawari (sunflower) and nut; It’s English name is probably a portmanteau that means “kernel of a sunflower.” So there’s not much that’s tricky going on here. The Pokemon we'll look at next, Shuckle, has a concept that seems muddled from start to finish, Sunkern’s exactly what it says on the tin. There’s a lot of evidence that Sunkern was one of the very first Pokemon designed at the beginning of the 1999 reboot, or even possibly before that in late 1998. For instance, on April 1st, 1999 (once again, our first snapshot of rebooted development) Sunkern’s already gotten unique stats and a Grass-typing, while Sunflora is still using the flat 50s in every stat that it had in Spaceworld ’97. It also evolved into Sunflora at level 20, and it had a unique name: Bayleef, which was eventually repurposed for, well, Bayleef. In retrospect, I should have put Sunkern at the very beginning of Period 4a, right next to Heracross. The reason I didn’t initially due this was because I was thrown off by Sunkern's weird development history. Despite the gameplay (ie, the stats, evolution, and typing) of Sunkern being completely done by April 1999, we have no evidence of what its sprite looked like, or if it even had one. And then, by June, Sunkern no longer had a name (only called Mitei 01, the placeholder the team used for blank slots) and we know that it had only placeholder stats by the end of July. So Sunkern got deleted! And then it's back, in the middle of August, this time with a goofy joke sprite that looks like a little dude with a sprout coming out of his mouth. Here's the sprite, along with RacieB's wonderful art of it: What an absolute goofball! This art for Sunkern was clearly made in a rush, and probably only meant as a placeholder while the team decided what Sunkern would really look like. There's a bunch of goofy sketch sprites like this in Spaceworld '99, where this Sunkern comes from, all of which have no shading and look like just a basic idea of what the Pokemon might look like. In particular, I think there's enough similarities between the sketchy early sprite of Shuckle and this sprite of Sunkern to speculate that there might be a connection. We know Shigeki Morimoto made Shuckle; maybe he also designed Sunkern? Obviously the early Shuckle sprite has a bit more work put into it, but the eyes and mouth are done almost identically, and both don't bother with a black outline around parts of the sketch (the sprout's stem, Shuckle's white spots). The question that arises, though, is why the team needed to come up with a new sprite at all. Sunkern doesn't have a sprite in the June 1999 spritebank we have, so it's very possible that it never had one, and only got this sketchy one after being reimplemented into the game. But as I wrote above, Sunkern was basically finished in April 1999, which implies it had to have been using some sprite or another. In addition, in the June spritebank, Sunkern isn't using an old sprite from Spaceworld '97 as a placeholder, like almost every other Pokemon without a sprite is using (see Shuckle's backsprite, Girafarig-3, Yanma-2, Pupitar and Larvitar, and others). Instead, it has a shadow outline of Pikachu, with a green palette. This implies there was a sprite there--probably one using the green palette the Pikachu outline is using--and it had been actively deleted. So what was there? I’ve said before that the strange eye-flower that was ID #404 in the Korean Index may have been an early Sunkern. That sprite didn’t appear in Spaceworld ’97 at all and neither did any form of Sunkern. Sunflora was a single stage Pokemon with the Grass/Psychic typing, if you'll remember. But it is possible that Sunkern, with #404's sprite, could have been in an earlier build, removed before Spaceworld '97, and readded later, probably during 1998, a hazy period of development that we don't have much data on. While this seems complicated, it also wouldn’t surprise me at all that the development team went back and forth on Sunkern’s inclusion in SW97, given how they almost deleted it again in 1999. Saying that, I’m skeptical that ID 404 was ever Sunkern, mostly because of it’s placement in the Korean Index. ID 404 was found in Period 2d, which was otherwise filled with strange, unrefined designs probably done by minor developer whose designs were mostly discarded. Given how none of Period 2d appeared in SW97, I don’t think it’s likely that ID 404 just happened to be Sunflora’s first form while the rest of that section were forgotten. Another crazy idea I have is that, possibly, Hanamogura (the strange second stage of Chikorita) had been repurposed by this point into a pre-evolution for Sunflora. Hanamogura is generally just a mystery regardless, but if you go back to that entry, you can read my explanation of what I think was going on there. In short, it’s possible that Hanamogura influenced Meganium’s design, but then was seen as an outlier and dropped from the evolutionary line later, because Meganium was such a good design that the first stage, Chikorita, took its cues from Meganium's design and made no attempt to build into Hanamogura. It's very likely the team already knew they’d have to rework Hanamogura into a better second stage even during the Spaceworld ’97 timeframe, and by the time of the 1999 rework, Hanamogura had already been dropped: it’s name was Mitei 01 (Pending 01) by April 1999, it's name data was completely deleted until mid-June 1999, and then finally it was renamed Bayleef. The assumption most fans have made was that the starter Bayleef got its name from Sunkern’s early name, after Sunkern was temporarily removed from the Pokedex. But what if it was the other way around? What if Hanamogura had been renamed “Bayleef” sometime in 1998, and early in the reboot it had been moved into a slot before Sunflora, eventually becoming the early Sunkern? Thus, when the developers actually came up with a new second form for Chikorita, they simply brought back the name they had already been using for the second stage and taking it from the slot Hanamogura/Sunkern was using. Admittedly, Hanamogura doesn’t look great as a Sunflora pre-evolution, but in this scenario the team probably already knew that Hanamogura would have had to be redrawn to better fit with Sunflora. When Sunkern was deleted, and then re-added to the Pokedex, Game Freak thus needed a new sprite, because Hanamogura's was already slated for rework before Sunkern had been temporarily removed. Note, also, that the Pikachu outline sprite used for Sunkern uses a dark green palette (on the question mark, most obviously). That palette would have been a perfect fit for Hanamogura if it had been updated for the Gameboy Color. The third option—and I have to stress, the most likely option—is that despite how finished Sunkern’s stats were in April 1999, the team hadn’t yet developed a sprite for it, and that it was using those placeholder Pikachu sprites since its creation. This isn’t the flashiest or most exciting answer, but it does explain why Sunkern uses a sketchy unshaded sprite when it got readded to the game in August 1999, rather than the old sprite it had previously been using. Most likely, the team had to use a sketch because they had never had a Sunkern sprite to use in the first place. Anyway, let's talk in a bit more detail about the team's ambivalence over Sunkern. Like I said above, Sunkern’s development indicates that it was on the chopping block through June and July 1999, only brought back into the games in the last couple months. Though it was almost complete by April 1999 but by July 30th, all of its stats had been deleted and replaced with placeholder 10s across the board. Then, fifteen days later, by August Sunkern had lost it’s typing and had just become Normal-type. At the same time, it stopped being able to evolve in Sunflora. By mid-August, Sunkern was completely deleted from the game. However, just seven days later, the team changed their minds. On August 22nd, Sunkern regained its typing, got new stats, and became one of the only Pokemon to use the Sun Stone to evolve. The team may have decided to bring it back a few days before this: Sunkern's sketchy prototype sprite first appears to us on August 17th, though given the gaps in our knowledge, that sprite could have been made back as early as late June, right before the decision was made to axe it. Or it could have been made quickly after August 14th, once the team decided to bring Sunkern back; it's very difficult to tell. When it did make its glorious return, the team replaced Sunkern's old stats (55 HP, 45 Attack and Defense, 65 Attack and Defense, 30 Speed) with 30s across the board, making it one of the absolute weakest Pokemon in the entire series, even to this day. Those old stats seem to have been derived from subtracting twenty points from all of Sunflora's stats at the time except Special Attack (Sunkern's was forty lower than Sunflora's). These stats were nothing special, but the new stats given new Sunkern are absurdly low. My only explanation is that when the team revived Sunkern they probably had reimagined it as a pseudo-baby Pokemon and given it the stats to match. Given that Sunkern has no arms or legs, and given that the only job of a seed it to flower, there is some sort of logic to making sure Sunkern is more or less useless before it becomes a Sunflora. So why did the team bring Sunkern back, after getting so close to deleting the seed entirely? I think there are two likely reasons. First, I think that that the team just put off replacing Sunkern with anything for so long that by August 14th, when they turned their eyes back to it, they just didn't have the time to develop a good replacement. By that point, the developers were hard at work replacing other slots that had needed fixing: Hironobu Yoshida was replacing the slot for a Shuckle evolution with Dunsparce, he was reworking Twinz into Wobbuffet, and was working on making Celebi interesting. The team was also developing Lanturn, Chinchou, Slugma, Foretress, and trying to figure out what to do with the slot previously used by a Yanma evolution. They hadn't paid much attention to the Sunkern slot for months while developing other aspects of the game, and I imagine that when they turned back to it, the team just decided that the concept of a 1st evolution for Sunflora was "good enough." The idea worked, it wouldn't take much development work to fix, and they already had a concept nearly done. They just needed to draw a sprite. Why make more work for themselves as they got closer to the release date? If it were me I would have used this slot to save Tsubomitto or Plux instead, but who knows what their thought process was there. Tsubomitto and Plux probably felt too similar to Bellossom and Scizor, hence their deletion; Sunkern, on the other hand, added some depth to Sunflora, and for that reason might have been deemed more important to the game. The second reason was suggested by a commenter on an earlier post, though I can't find the comment at the moment. This commenter speculated that Sunkern was added back into the game because the team realized that Gold and Silver only had one Pokemon that could evolve with the Sun Stone: Gloom, into Bellossom. The second Sun Stone user, Tsubomitto, had been slated for removal at almost exactly the same time that Sunkern was brought back from the dead, so Tsubomitto’s deletion could have triggered the team to bring Sunkern back to make the Sun Stone have some use. If that was the case, that explains why Sunkern’s goofy joke-sprite is so sketchy: the decision to bring back Sunkern might have been made very quickly after Tsubomitto’s deletion, and the sprite was created on the fly so they wouldn’t have to use placeholder sprites. All of this seems like a very complicated origin story for such a mediocre Pokemon. But maybe that’s exactly the point. The designers didn’t face this sort of indecision about Donphan, or Slowking, or Espeon, because those were good designs that evoke a lot of passion in players. Sunkern is a bottom barrel design, just barely good enough to justify its existence. Of course the designers would be on the fence about including it into the game. In the final game, Sunkern barely leaves an impression on players, and while Sunflora has had a lot of reappearances (even appearing in that minigame in Pokemon Scarlet), Sunkern is, to this day, pretty forgettable. Sunkern, as it turns out, didn’t have a complicated design story in spite of it’s boring design; in fact, Sunkern had a complicated origin precisely because it was so middle of the road that it barely made it to the finish line. (Fan Art by Javiera) ID X17: ShuckleWhile Shuckle existed prior to April 1999, and thus fits into Period 4a, Shuckle's journey through development is, like Sunkern's, a strange one. First appearing as more or less a generic-defensively specced Geodude lookalike, nearly everything about Shuckle was changed by the final game. So what's going on with Shuckle, the turtle/snake/bug/wine/rock Pokemon? Even in the final games, Shuckle’s a really strange Pokemon. Growing up, I was sure it was a cool turtle Pokemon, but there's nothing about Shuckle except its sprite that actually has to do with turtles, and the evidence is that the developers didn't even see the resemblance to a turtle. In fact, Shuckle is a Bug! But it's a bug-type Pokemon that...learns no Bug-type moves. It's also a Rock-type, even though it learns no Rock-type moves through level up, and only one--Rollout--through TM. Interestingly, Shuckle's Bug/Rock typing was also completely unique as of Generation II, and even in the present day, the only other Bug/Rock Pokemon are Dwebble, Crustle, and Kleavor. In the final games, Shuckle is most notorious because it is a gimmicky defensively specced Pokemon. Shuckle, of all Pokemon, has the highest Defense and Special Defense in the series, nearly maxxed out at 230. It trades those stats for some of the worst HP, Attack, Special Attack, and Speed in the entire series, making it more or less unusable except in the silliest gimmick runs. Even its moveset is odd: all Shuckle gets are Wrap and Constrict, two moves with the lowest base power in the series, and a bunch of random defensive buffs. What I'm trying to say is this: Shuckle is certainly odd. Part of the strangeness (some might call it uniqueness) of Shuckle can be traced to it’s development. Shuckle seems to have been created to fit a role almost entirely different than the one it fills in the final games. In our first view of Shuckle in April 1999, Shuckle was a Rock/Ground type. the typical typing that Rock Pokemon got in Generation I (likely because of how late the Rock type was created). Shuckle also didn’t have its strange stat spread, so it wasn’t the defensive wall it later became: instead, it had low, generic stats all around, though it’s Defense and Special defense were already it’s highest stats. Originally, Shuckle also evolved into a stronger Shuckle! We’ll talk about that more in the next entry. (Table thanks to TCRF) Quite frankly, these stats are awful. But they do seem to fit the general role that Rock/Ground Pokemon (most notably, Geodude) were used for in Generation I. It's stats probably needed a buff, or an evolution (see the next entry), but I could see this version of Shuckle as Generation II's early game defensive Rock-type; its Geodude equivalent. I’m a bit curious about what Shuckle’s moveset looked like at this early stage as well. It’s final moveset is absolutely goofy, and bad; Though, kudos to the developers, it fits Shuckle’s flavor incredibly well by stressing just how much this thing is a wall. Given that the earlier version of Shuckle was not nearly as focused, I wonder if it had a more generic Geodude-like moveset? Or, since its scrapped evolved form only ever got a placeholder Tackle as its only move, maybe that’s all that Shuckle had at that point. At the time Shuckle was Rock/Ground, its sprite was also significantly different. Compare: At first glance, this still looks like Shuckle, but there are a number of important differences. First, Shuckle is more obviously some sort of snake-like creature: you can see it slithering out of the pot it lives in. Second, of course, the rock, or whatever the final version of Shuckle lives in (the Pokedex entry calls it a shell, but I can't tell if it's part of Shuckle or something it crawls into) is a pot in these early sprites! And while the final Shuckle seems to have four legs coming out of it’s shell, in the early sprite they’re much more clearly tentacles of some sort, making Shuckle a lot stranger. I don’t think we pay enough attention to how absolutely rough this early Shuckle sprite is. If you look at it close, you’ll realize that this is a sketch, barely more polished than the infamous early Wooper sprite or the goofy Sunkern sprite, as I compared it to above. There’s almost no shading on this prototype sprite, and the perspective is basically flat. I think that if early Shuckle looked noticeably alien compared to the final design, it’d be easier to note just how much of a placeholder sketch this sprite is; since it looks close to the final design here, that makes it easier to overlook some of the sketchy details. Shigeki Morimoto created Shuckle, and I’ve always been curious about exactly what was going through Morimoto’s mind as he shaped the design. I think that part of the reason the final Shuckle's design is all over the place is because Morimoto didn't have a strong concept for a Rock/Ground Pokemon, leading the team to change its typing and eventually its stats and take it in a different direction. But I have a few possibilities for where Shuckle's origins lie. One obvious thought is that this early sprite, of a snake coming out of a pottery jar, looks a lot like the snake that a snake charmer would use: kind of like how Nameeru appears to be a snake charmer itself. Bulbapedia also suggests that Shuckle might derive from "Scale insects," which are apparently (I'd never heard of them before) insects that live under a small shell and produce honeydew to attract other insects for food. That description certainly seems to fit Shuckle. But I personally like another explanation. Though OrangeFrench strenuously disagrees with this interpretation, a lot of people think that Shuckle was originally developed from the idea of making a Pokemon out of Snake Wine. Snake Wine, a pretty exotic drink developed in Vietnam and China, is a type of alcohol in which a poisonous snake’s body is left in the container while the alcohol ferments. The theory is that the venom from the snake dissolves into the alcoholic drink, giving it a unique taste and effect on the imbiber. It’s a pretty weird inspiration for a Pokemon, but given that the early sprite of Shuckle looks like exactly the sort of pot you’d ferment wine in, the early development information we have seems to lend strength to this theory. Of course, there’s one giant flaw in the middle of this snake wine theory: why is Shuckle a Bug? If it was actually a snake, I could imagine the team making Shuckle into a Poison/Rock Pokemon, or maybe Poison/Ground, to reflect the idea that the snake in the jar is dissolving venom into the drink. Even Shuckle’s original Rock/Ground typing makes some sort of sense, if you just imagine that they’re playing up the aspects of Shuckle that have to do with earthenware clay. But Bug/Rock is a thoroughly bizarre choice for this origin story. OrangeFrench, on the other hand, has a competing theory for Shuckle’s origins that explains this Bug type. In the wild, there are a bunch of animals—notably, ants—which collect berries and other fruit and then guard them while they ferment, consuming them only after they have a level of alcohol in them. If ants are well known as an animal that does this, then Shuckle might be related to real world bugs that ferment fruits. Thus, though Shuckle may not look like it, Shuckle’s just a weird bug waiting for its berries to ferment. In early drafts, Shuckle was also known as the “pottery” Pokemon, but in the final, it became the “Fermentation” Pokemon, a title that could be evidence for either theory. There’s another oddity about Shuckle which could potentially lend credence to either. In the final game, Shuckle has a special ability no other Pokemon has. If a Shuckle is holding a berry, it has a 1/16th chance after every battle to transform that berry into berry juice. It’s a weird thing to program, and feels much more suited to Generation III, when Game Freak introduced passive abilities that each Pokemon had; in some ways, Shuckle’s ability to create berry juice feels like a prototype of that idea to give every Pokemon a unique passive trait. But this act of fermentation could prove that Shuckle is exactly the type of bug that OrangeFrench suggests, hoarding berries until they turn into "juice." It could also be the case that this juice is just a euphemism for wine, and that this proves that Shuckle is creating snake wine in its jar/shell. Shuckle's early Pokedex entry referenced the pottery it lived in, explaining that “It sneaks into jars unnoticed. Nobody knows what its true form looks like.” It’s kind of a strange entry: unlike other unknowable Pokemon like Forrestress, or Tangela, or Mimikyu, it's not like Shuckle's face or body is obscured. You can absolutely tell that Shuckle is just a weird snakey guy. Given that he isn’t really hidden, nor does he live in a jar in the final game anymore, the team changed Shuckle’s Pokedex information to something a bit more relevant. In the final, its entry focuses much more on Shuckle's relationship to berries: It stores berries in its shell. It hides motionlessly under rocks to avoid being attacked. It also mentions the most common way to find a Shuckle in the final games: using Rock Smash, a player might find this cowardly bug hiding under a boulder! Finally, there's one last important difference between early Shuckle and what it became. When first created, Shuckle was supposed to have an evolution! So let’s move on, and consider what a super-Shuckle may have been like. ID X18: Shuckle-2Yes, there was originally a giant, super-powered version of…whatever Shuckle’s supposed to be. Let’s go through what we know, which unfortunately isn’t a lot. Unfortunately, we don't have a sprite for Shuckle-2. As a result, in my head, I'm just imagining a giant kaiju Shuckle. I invite you to do the same. (Pictured above: Regular Shuckle next to its monstrous brethren) Anyway, the earliest data we have for Shuckle-2 are from April 1999, our first glimpse into the reboot period of development. In this snapshot, Shuckle-2’s stats are available: it has exactly double Shuckle’s stats in each category. This means that Shuckle-2 had pretty good defense: Shuckle had 60 in both defensive stats, which means that Shuckle-2 had 120 Defense and Special Defense. However, because of the rest of Shuckle’s early stats were terrible, Shuckle-2 was nothing to get excited about. It had 70 HP, 50 Attack, and 30 Speed, none of which was particularly good, especially for a fully evolved Pokemon. Shuckle-2 was also a Rock/Ground type, just like Shuckle-Classic, and Shuckle evolved into it at level 22. Unfortunately, Shuckle-2 had nothing but Tackle as its moveset, though I’m not sure if, at this point, Shuckle had anything more than that placeholder Tackle as well. In the case of Piloswine-3, the Scratchpads data we have left us some evidence that at least preliminary designs for a third Piloswine evolution were being worked on, but it’s very unlikely the same can be said for Shuckle-2. By June 1999, when we have our first snapshot of sprites, Shuckle-2 was using the old sprites of Berunrun, which it replaced in the Pokedex. Given that it was still using these sprites, they were probably being used as placeholders until a sprite was drawn for Shuckle-2. Since no sprite ever was drawn, those old Berurun sprites were still used until Shuckle-2 was rejected, and then they were replaced with early Dunsparce sprites. Shuckle-2 was in this slot until at least July 18th, which was three months after it first appeared, so there’s a small chance that it briefly got a sprite in July that was deleted by August 1999, when we have our next snapshot of the sprites. However, it isn't very likely that a sprite was ever made. Remember that Shuckle’s sprite as of June 1999 was incredibly sketchy, and it kept that sprite all the way until just a month or so before the final build. If the team hadn’t bothered updating Shuckle’s sprite, they probably hadn’t gotten around to making even a sketch for Shuckle-2. It's hard to stress just how sketchy Shuckle-2 was: it clearly wasn’t a Pokemon, but more or less an idea for an evolution that just never happened. A bunch of times, I’ve mentioned that Masuda’s method for designing the roster was to reserve Pokedex slots for a particular idea, and then have the designers fill in the gaps by creating a Pokemon to fill that role. Here, it seems pretty clear that Shigeki Morimoto was asked to design a two-stage Rock-type (or Rock/Ground type) Pokemon, and that the first idea he came up with was Shuckle. Given that it’s defense was its only notably good stat as it was first constructed, I’m sure part of the idea was to design an early game defensive Rock-type. However, it seems that Morimoto didn’t quite know what to do Shuckle beyond its first form. If Shuckle was designed around the goofy idea of a living snake-wine snake, then what would an evolved form of that even be like? I could imagine a world in which Shuckle’s evolution was a giant serpent, with teeth and fangs, overflowing out of a pottery jar, but Morimoto clearly seemed to think of Shuckle as something whimsical and funny, and this sort of concept just wouldn’t have worked for Shuckle. So I suspect that Morimoto left the second evolution empty until he could come up with a good idea to fit the spot. Saying that, what actually killed the idea for Shuckle-2 was the decision to make Shuckle a gimmick Pokemon based upon maxed out defense stats. It’s no coincidence that the last update to Shuckle-2 was made on July 18th, and it was transformed into Dunsparce by August 14th: after all Shuckle’s stats were changed to give it 180 Defense and Special Defense at the exact same time Shuckle-2 ceased to exist. My guess is that while Morimoto liked his little goofy pottery-snake, the team were scratching their heads about what role Shuckle could actually fill in the completed games. Once they finally figured out its role—to be a super-defensive wall—it made the idea of giving Shuckle an evolution obsolete. Sure, they could have kept the Defense and Special Defense of Shuckle at 180 and upgraded them to 230 when Shuckle evolved, but that's pretty boring: not only is the evolution not a very exciting upgrade (I guess it's more impossible to kill now?) but it also makes Shuckle less interesting, since it doesn't even have the notoriety of being the ultra defensive Pokemon anymore. Once the team had figured out Shuckle’s deal, that left the next Pokedex slot completely open. At this point, it looks like the empty slot was handed over to Hironobu Yoshida, who designed Dunsparce as a one-stage Pokemon that could fill the empty slot. I discussed how this happened in a ton of detail back when we were discussing Proto-Dunsparce, but the short version is that Yoshida—who will go on to be one of the most important Pokemon developers in later generations—seems to have been brought on to the Pokemon design team pretty late in the day and was then handed the Pokedex slots which the team needed a quick replacement Pokemon to fill. Yoshida’s three known contributions are Dunsparce—which replaced this unused Shuckle-2—Wobbuffet—which replaced Twinz after the idea for Girafarig to have a first stage fell through late in development—and Celebi—the team probably knew they wanted a mythical Mew-like Pokemon at the end of the Pokedex but no one had yet come up with a design. It seems that Masuda handed particular assignments to the other developers, such as asking Morimoto to come up with a Rock-type defensive Pokemon similar to Geodude, asking Sugimori to develop an analogue to Dragonite (Tyranitar), or asking some developer to develop an early game Normal Pokemon that could show off the idea of berries (Sentret). But in the case of Yoshida, who joined the team as most of these slots were already assigned or filled, he seems to have been given the problem assignments and just told to create whatever he wanted to fill those gaps. Like I’ve previously mentioned, it seems that Yoshida, looking for inspiration, would glance back at the discarded designs that were still in the Korean Index when deciding how to fill these slots. His early Celebi strongly resembles the discarded Kokopelli design from the Korean Index, and of course Dunsparce is strikingly close to ID #412. In this case, since Shuckle-2’s stats were replaced with Dunsparce’s final stats the second it was replaced, it seems like Yoshida got his inspiration from the Tsukinoko sprites in the Korean Index, drew up a stat spread he thought would reflect that mythical creature, and then got to work on the sprite immediately. Unlike the Piloswine evolutions, Shuckle-2 was barely a sparkle in the team’s eye before it got deleted. It had no movepool, no sprite, and barely anything distinctive about its stats. It also barely looks like the final Shuckle, having been created at a time when the team still didn’t really know how Shuckle would fit in the final games. So it’s hard for me to get too wound up about this loss. There was barely anything here, and then it was gone. (Mega Shuckle Concept, by Rollingstone51 on Reddit. Thanks for this awesome design!) ID X19: Girafarig-3It’s possible that Pokemon with the most complicated design history in Generation II was Girafarig. Despite being created at the very beginning of development, all the way back in Period 1d, Girafarig went through an incredible amount of changes. It was a Dark/Normal type, a Ghost/Normal type, and eventually a Normal/Psychic type. It's concept changed from an Giraffe with an evil twin into a Giraffe with a funny tail, and its pre-evolution went from a duo of ghosts, to a swirling Haunter-like Ghost, to a baby Giraffe, to nothing at all. Girafarig went from being a two-stage Pokemon family, to a three-stage family, back to a two-stage and eventually just a single-stage Pokemon. The design team really wanted Girafarig to have evolutionary relatives, but when those didn’t work out, they eventually went with the most boring version of the “evil ghost Giraffe” concept. Finally, 23 years later, Girafarig finally got a evolved form, in Farigiraf. Farigiraf's a recent idea, but there was a push all the way back in 1999 to give Girafarig an evolved form. We don’t know much about this Pokemon as the only data we have are some stats. Girafarig-3 got about as far as Shuckle-2 did, meaning that it was more or less a placeholder slot for a third Girafarig form more than it was an actual deleted Pokemon. Still, there are at least a few things we can say about Girafarig-3. First of all, let’s look at the stat comparisons. In April 1999 until Girafarig-3 was deleted, Twinz, Girafarig, and Girafarig-3 formed a three-stage line that followed more or less a typical spread of stat upgrades as they evolved. Girafarig added twenty points to all of Twinz’s stats, except speed (which only gained 10); Girafarig-3, in turn, added five to all of Girafarig’s stats except Special Attack and Special Defense, which both got a 15 point upgrade. Twinz was a pure Ghost-type, but Girafarig and Girafarig-3 were Ghost/Normal, on account of them both (presumably) having a regular Giraffe half and an evil Ghost half. Twinz evolved into Girafarig at level 17, and Girafarig evolved seventeen levels later, at level 34. All three of them only learned a placeholder Tackle as their only move, which probably reflected how difficult it was for the team to come up with a moveset that suited both the Ghostly, ethereal Twinz and the four-legged Giraffe-shaped Girafarig. Like I mentioned back in the Twinz entry, Twinz and Girafarig were the only Pokemon in Spaceworld ’97 not to have a full moveset, which was very odd and probably down to this difficulty. Evidently, even by the beginning of the 1999, the team hadn’t solved this problem: all they’d done was change the family to the Ghost type and added on one more member. Notice, as well, that the final Girafarig has higher stats than Girafarig-3 did across the board, and that Farigiraf's stats blow it completely out of the water. Clearly, Girafarig-3 would have gotten a boost before the end of development had it not been culled.
The Girafarig family stayed more or less exactly like this until late July; unfortunately, there’s no indication whether Girafarig-3 ever got even a basic sketch of a sprite. In the June 1999 sprite bank, Koonya’s—the discarded baby Meowth from Spaceworld ’97—sprites are being used in Girafarig-3’s slot. Koonya’s sprite has been slightly updated to fit the new Gameboy Color style shading of the reboot, suggesting that it was worked on at least a bit after Masuda rebooted development before being replaced with Girafarig-3. This, interestingly, suggests that there was some work done in early 1999 or late 1998 that we're not privy to; I would love to see a build dated from January 1999, for instance. I bet we'd see some very interesting variations on old Spaceworld '97 'mons, along with a few new concepts gone by April 1999. Koonya’s still using a Meowth-like palette by June 1999, which suggests that the team hadn’t begun to get to work on a Girafarig-3 sprite. While it’s certainly possible that in the next month the team created something, I find it pretty unlikely. Most plausibly, Girafarig-3 was never worked on to the point that the team had a reason to create a sprite, and it was deleted before anyone started doing sketches. Pineco, the eventual replacement of Girafarig-3, has the old sprites of Rinrin (the black cat) on its scratchpads, despite Rinrin probably never having inhabited this Pokedex slot. That could indicate that in July 1999, Girafarig-3 had used Rinrin’s sprites instead of Koonya’s before being deleted. This is speculative, but there’s a possibility that Girafarig-3 didn’t get a sprite because the team wasn’t absolutely sure whether Girafarig would be the final stage evolution in its three-stage line, or whether it would stay in the second spot in its evolutionary tree. As we saw in the case of Piloswine, sometimes during the reboot, multiple slots in the Pokedex were designated as part of one evolutionary line, but the designers weren’t sure exactly which order the evolutions would be in. In Piloswine’s case, Morimoto just made a sprite for all three to test out how it looked as a small, medium, or large Pokemon. In the case of Girafarig, the team could have been considering a transition Pokemon between Twinz and Girafarig, to make the evolution from one to the other feel more natural. Those two were so different and incongruous as is, so the thought may have been to make a Pokemon with traits of both. Thus, the missing Pokemon here may not have actually been Girafarig-3, which would have just been Girafarig. Maybe the one we're actually missing is the transitory stage two between a double headed ghost and an evil Giraffe. Saying that, we do know that Twinz’s slot was eventually repurposed for a baby Girafarig (which we’ll talk about in Period 2b); it’s not crazy to think that the baby Girafarig was supposed to be the missing link between the two rather than overwriting Twinz itself (though of course, in the data we have, the baby Girafarig does absolutely overwrite Twinz, but that could be a reflection of the team changing their mind). Thus, the family could have been Twinz -> Baby Giraffe -> Girafarig. Girafarig-3 only lasted as a concept until late July or early August, when it was replaced by Pineco, a completely different Pokemon that we’ll discuss in Period 4b. This is, not coincidentally, almost the exact same time that Twinz was replaced with Girafarig-1, which was also the same time that both it and Girafarig became Normal/Psychic types and dropped the Ghost-typing. There are two possibilities about what happened here. The first is just that when Twinz was updated into Girafarig-1, the team cleaned up everything surrounding Girafarig, including its unused third Pokedex slot. Under this theory, the idea of a third Girafarig evolution had been defunct since at least early July, but the team had been distracted with only priorities: when they added Girafarig-1 into the Pokedex, they simply deleted Girafarig-3’s slot as well, updating the Pokedex to their current plans. Pineco got its name by August 1st, so either the team just flat out replaced Girafarig-3 with a Pokemon the team had decided to include but was waiting for an official slot, or the Pokedex slot was briefly empty for a week or two before it was updated again to add in Pineco’s sprite. The second possibility is that the creation of Girafarig-1 was the cause of Girafarig-3 deletion: Girafarig-1 made Girafarig-3 redundant. Under this line of thinking, once the team had decided that Twinz wasn’t working, they came up with a pre-Girafarig evolution much closer to Girafarig itself. Not only would this new evolution be thematically closer to Girafarig, but it would also help make creating a moveset much easier. This, I'm sure, would be a huge help for the team because even by August, they still didn’t have a moveset for pre-Girafarig (and presumably Girafarig)! But once they had this smaller Girafarig, a third evolution may not have been necessary for the concept anymore. Maybe the team only wanted that third slot to—like I suggested above—make a bridge evolution between Twinz and Girafarig, and without Twinz, a bridge was no longer needed. Maybe the team decided that two Giraffes that looked fundamentally the same were enough, and a third evolution wouldn’t have much to build on. Who knows. It could be as simple as what happened to Shuckle-2: once the team finally realized what they wanted to do with Girafarig, they realized a third evolution wasn’t necessary for that concept. Whatever the reason, Girafarig-3 disappeared by the end of July or early August, replaced with Pineco. It never got a moveset, and as far as we can tell, never got a sprite at all. It barely lived. I can say one thing to speculate about it. Girafarig-3 was a Ghost-type, and would probably have reflected that Ghost-type more fully that Girafarig, which was only a Ghost-type because that fit Twinz better. My guess is that had the team gone forward with a Girafarig-3 concept, the goal may have been to justify its Ghost-typing more than Girafarig did, and make it spooky. Certainly, it wouldn’t have ended up looking anything like Farigiraf. We can't know what it would have looked like, but here are a few Mega-Girafarig concepts I love. Maybe you can headcanon one of them into Girafarig-3.
11 Comments
ID #X11: LugiaThree years after Ho-Oh was first designed, its Silver-version counterpart finally made its way into the second generation of Pokemon games. Lugia’s a strange Pokemon, with an odd origin story that isn’t entirely clear. Let’s go through the steps, and see what we can find out about this very odd legendary…bird? Beluga whale? Before we start, by the way, it’s important to note that Dr. Lava wrote an excellent youtube video about the origins of Lugia, featuring a bunch of interviews with Takeshi Shudo, the head writer for the Pokemon anime, that he paid to translate. That’s where a lot of my source work for this entry is coming from, and I’d advise you to check it out! First, let’s talk about where I'm placing Lugia in Period 4. Lugia has finalized sprites by June 1999, and semi-placeholder stats two months earlier, in April (its stats are just Mewtwo's). This implies that it had been created close to the beginning of the reboot, but it was still being worked on; hence why it's here. It’s very likely that Lugia was put into the Pokedex at nearly the same time as Tyranitar (coming up next): both of them have nearly finalized stats in April 1999, and both of them have nearly final (but not completely final) sprites in the June 1999 version of the build. In fact, given how their timelines seem to match, and given that Sugimori likely drew both their sprites, that I wonder if he input their sprites into the games back-to-back. There is, however, one piece of evidence that suggests Lugia was in the game slightly before Tyranitar. On the scratchpads for Tyranitar, one of them has Lugia’s backsprite. There are a lot of reasons this could happen, but the most obvious one is that Lugia was originally in Pokedex slot #248 before it was moved to #249. If that was the case, then that has interesting implications for the planning being done in the early stages of the 1999 reboot. If Lugia was in #248, then why were their two (plus #251, so technically three) slots after it? Was the team planning for another legendary creature? If it was in slot #248, then Tyranitar might have originally been considered to be only a two-stage line, taking up #246 and #247, which is also interesting in and of itself. Why did they change that? I have no definite answers, and even this is very tenuous speculation (Lugia’s backsprite could be on Tyranitar’s scratchpad because Sugimori was using it as a way to judge proportions, for instance). But it’s interesting to think about what this could mean. Lugia is the only Pokemon from Generation II whose origins don’t come from the needs of the game itself. Instead, Lugia was created exclusively to be the star of the second Pokemon movie. The movie was written by Takeshi Shudo, who was the head writer for the Pokemon anime. His original idea for the film involved a real-world dinosaur fossil being discovered. It sounds as though this Tyrannosaurus Rex fossil would come to life and go on a rampage, and the Pokemon team would have to stop it. It’s a pretty nuts idea for a movie, and I’m glad it was eventually changed! Some people think that the discarded fossil Pokemon sprite from the Korean Index dates back to this strange idea, but that’s very unlikely. First of all, that sprite was probably made by someone who wasn’t a main designer of Pokemon, and thus were unlikely to be heavily involved in this sort of planning. Secondly, Shudo made a point of explaining that the fossil from the movie wasn’t a Pokemon, and therefore couldn’t be caught. (Not you) Anyway, after the "real world fossil attack" idea died, Shudo came up with the idea of a Pokemon god—like a mother spirit that would be the origin of all Pokemon. This Pokemon, initially dubbed “Pokemon X,” had a connection to the tides, and lived deep under the ocean. This is where things get really unclear. In interviews, Takeshi Shudo has said that he created “Pokemon X.” Or at the very least he came up with the idea for it; it isn't clear. Either way, I’m not sure what he meant. Shudo probably didn’t draw concept art for it—he was a writer, not an artist, after all—but he did have strong feelings about its characteristics: it should be related to ocean tides, it should be a mother spirit, etc. And certainly, when Shudo says he created it, it implies that he played a role in designing what it would look like. In fact, Shudo claimed that other people came up with the name for Lugia, and that he was surprised Lugia made it into Gold/Silver at all. Again, this implies that Shudo saw Pokemon X as his creation. However, the earliest concept art we have of Lugia was made by Ken Sugimori, and all the spritework for Lugia seems to be Ken Sugimori’s handiwork. Given that Shudo wasn’t an artist, it seems far more likely that Shudo came up with the idea for Lugia and then Sugimori made the initial concept art. Does that mean Shudo described to him an idea, or that Sugimori made it from scratch? I have no idea! Interestingly, the initial concept art has a prominent “X” on Lugia’s belly, which not only connects to Shudo's name for the Pokemon but also suggests that this concept art is very early. This leaves open a lot of questions, at least in my mind. How much of this initial concept was just Sugimori’s ideas? How much input did Shudo have? Did someone else make the initial concept for Shudo, and then Sugimori edited it further into this concept? None of that is very clear. Another thing that’s unclear is when, exactly, did Lugia lose the X on its tummy? My first reaction was that the X was an old piece of the design just seen in the concept art, and that it didn’t appear on the sprites. But it’s not that simple. Even though the sprites we have from June 1999 are almost identical to the final Lugia sprites (just lacking a few of the back spikes), you’ll notice that these sprites don’t actually show us Lugia’s belly. The first chance we get to see Lugia’s belly is in Lugia’s final Silver sprite, which make it look like it just has a darkened spot there, not an X (Thanks to FlowersBloom for pointing this out on twitter!). So even though the X looks strange in this concept art, it’s very possible that the team considered it a part of Lugia’s design all the way until a month before the final. It’s even a mystery what Lugia’s name even means. Shudo says that the name was chosen in a meeting without his input, so his ideas about Lugia probably don’t play into it’s creation. Dr. Lava thinks it might be related to a Latin word “Lugeo” which means “to lie dormant.” This would be referring to it’s role in the movie, sleeping under the surface of the ocean, but it’s a pretty obscure Latin reference and I’m not sure how much the Pokemon team knew about the movie’s plot when they named it. Bulbapedia thinks Lugia is related to “Beluga" Whale which is has a slight resemblance to, or the ancient Greek word for silver. But that word is “árguros” which sounds nothing like Lugia, so I don’t really know what Bulbapedia is trying to say here. What was Lugia’s visual design even inspired by? What, exactly, is Lugia? Beluga Whale fits as well as anything else; Bulbapedia also suggests that it’s based upon the ancient legend of Bahamut, which was sometimes described as a giant fish and sometimes as a dragon, but that seems like a huge stretch. To me, the hand-wings seem to come out of nowhere, though I admit Lugia's not the only video game character with that design. Still, why give a legendary bird digits? (Behold: Lugia) Some people think that Sugimori’s concept art of Lugia was a refined version of the viking boat Pokemon from early in the Korean Index (ID 344, to be exact). This has always been a strange Pokemon in the Index, since it is smack dab in the middle of a sequence of Sugimori-designed sprites, and it’s the only one of his sprites from that sequence that didn’t make it into Spaceworld ’97. That’s always given me the impression that it was a pretty important Pokemon design that got unfortunately cut. It’s even possible that it was a concept for a legendary Pokemon, given its proximity to Ho-Oh. It’s not completely out of the realm of possibility that Sugimori took that old design and reworked it into Lugia. On the other hand, I’m not sure I see the resemblance at all. So I think it’s up to you to make up your own mind about the connection (or lack thereof) between these two Pokemon. We do know one thing for certain. We have enough information from both Lugia's origins and Ho-Oh's origins to be sure that they were never designed together, and not originally intended to be counterparts. Originally, Ho-oh was designed as kind of the ultra-legendary for Pokemon 2. In much the same way as Mewtwo stands above Articuno, Zapdos, and Moltres in Generation I, Ho-oh was supposed to stand apart from Raikou, Entei, and Suicune. However, after the development reboot, these plans changed. It’s unclear whether the decision was made to pair Ho-Oh with something and Lugia was slotted in to fill that role, or whether Lugia was created first. It can't have been long since Lugia first appeared in the Pokedex, however: two legendary Flying-types would have begged to be thematically connected. Notably, from it’s first appearance in the Pokedex, Lugia was Psychic/Flying, even though as originally conceived by Takeshi Shudo, it should probably have been a Water/Flying Pokemon. According to an interview with Shigeki Morimoto, the Psychic type was chosen not because it necessarily fit Lugia’s design, but mostly for gameplay concerns. At that time in Pokemon history, Psychic-type was widely considered the strongest type: Not only were Alakazam and Mewtwo absolute beasts in Generation I, but Psychic-types essentially had no weaknesses in the first game, making them good for just about everything. Thus, the team probably believed that Lugia would be more impressive to players with the Psychic-type. Notably, in the earliest data about Lugia we have, it’s stats are nearly identical to Mewtwo's, suggesting that the team's first idea for Lugia was to fill the same role as Mewtwo as a super powerful end game Psychic-type. Later, they swapped some of these stats so Lugia would be more defensively oriented than Mewtwo, which at least gave it a bit of a unique identity. Interestingly, even though by this point Ho-oh and Lugia were probably mirrors of each other, it took Ho-oh an incredibly long time to get it’s own secondary typing. In Spaceworld 97, Ho-oh was just a basic Flying-type; even by June 1999, it was a Normal/Flying type, even while Lugia was already a Psychic/Flying type. It’s only in August 1999, in the Spaceworld ’99 build, that Ho-Oh finally becomes Flying/Fire, getting its second type to better match Lugia. It’s unclear why this took so long. However, we’ve seen a pattern in development in which older Pokemon designs were sometimes left in an incomplete "good enough" state until the last month or so of development, while the team was prioritizing newer creations that needed more work. Thus, the team could have known that Ho-Oh needed a second type, but just left giving it one until much later. But I’m unsure about this, since it's typing actually changed from pure-Flying to Normal/Flying while Lugia was already Psychic/Flying. Is this some evidence that the two weren't conceived of as a pairing until even later, potentially just three months before the end of development? Maybe the team just wasn’t very sure what Ho-Oh should be? …I also want to take a moment here to point out that Shudo’s initial idea for Lugia—that it was a Water-type—would have made it fit as Ho-Oh’s counterpart much better. Definitely a Psychic-Type In the final games, it’s easy to see Lugia as the representative of the moon and of the night, while Ho-Oh is the representative of the day. This works really well because of the time mechanic introduced in Gold and Silver: it creates a neat parallel for Ho-oh to look like the sun and for Lugia to be silver like the moon. Shudo has said that Lugia was related to ocean currents, so it seems as though Lugia’s connection to the moon may have been an early part of its concept. However, I’m not really sure when Ho-Oh and Lugia themselves got connected to the Night/Day cycle in Generation II. Certainly, it couldn’t have been before Ho-Oh got its fire typing, and that was in August 1999, really late in the day and only a couple of months before the final release. In a later interview, Shigeki Morimoto claimed "Ho-oh and Lugia were born in the world of Pokémon Gold and Silver for its new feature, the time system, giving us the day and night cycle. Thus, they represent the sun and the moon, day and night." But we know that he's exaggerating a little, since Ho-oh was created before Lugia, and Lugia was created for a different reason. By the end of the development, however, the team may have thought the way Morimoto is suggesting. However, now that I think of it, do the actual games ever make a connection between Ho-oh, Lugia, and the time of day? Or is this something the team thought of but never said in Gold and Silver? Finally, let’s talk about the signature moves of Ho-oh and Lugia. Lugia learns Aeroblast at level one, a Flying attack with 100 power and a high critical hit chance. It’s a really good move, albeit one that doesn’t really have anything to do with Lugia’s concept or with its Psychic typing. Aeroblast is interesting, because it didn’t appear in the Spaceworld ’97 build. It seems to have overwritten a move called “Coin Hurl,” another signature move for the Meowth line that would have done more damage the more money the player had (which seems terrible, incidentally, since it would incentivize competitive players to get the maximum amount of money they could, through grinding, making the competitive scene really unfun). Even though Aeroblast technically replaced Coin Hurl, however, it’s probably a rework of a different move, “Wind Ride.” Wind Ride was a move in the demo that was Flying, had 40 power, and also a high critical hit chance, so Aeroblast is more or less just a souped-up version of Wind Ride. But what’s most interesting about Wind Ride was that it was originally Generation II’s replacement for “Fly.” In the Spaceworld ’97 build of the game, there are a number of moves at the end of the movelist with field effects, meant to replace the old HMs: Brightmoss (Flash), Strong Arm (Strength), Wind Ride (Fly), Uproot (Cut), Watersport (Surf), and then Whirlpool and Bounce, presumably the two new HMs that became Whirlpool and Waterfall. We’ve talked about this before, but these were probably cut because of the confusion they’d create: if you traded a Pokemon from Generation I with Fly, would Fly function? If you then also taught it Wind Ride, would it be able to Fly twice? It just feels like an awful lot of complication to make moves that do the same thing as moves they’d already created. Anyway, Aeroblast no longer has the “Fly” effect in the final game, but it’s obviously a holdover from this idea. Given how late in the day Lugia was, when the team wanted a signature move to give it some identity, it probably made sense to repurpose an old idea and buff the power. In Ho-Oh’s case, the team spent even less effort creating a signature move. In Spaceworld ’97, Ho-Oh learned “Sacred Fire,” (Itself an oddity, since Ho-Oh was not fire type yet and wouldn’t be for almost two years) a new fire move also learned by the Arcanine family and by Moltres. It was already a rare move in conception, given only to really powerful Fire-types (and Ho-Oh). By the final, the team just took this move away from everyone except Ho-Oh, creating a signature move to contrast with Lugia’s. Personally, as a kid I was always a bit confounded by Lugia. Ho-Oh appeared in the first episode of the anime, and I had always thought of Ho-Oh as the star of Generation II. Lugia didn’t seem to resemble Ho-Oh at all, or have any reason to exist compared to Ho-Oh. Now, of course, that makes sense: Lugia was created very differently from Ho-Oh, and only welded to Ho-Oh late in development. My younger self would feel vindicated. ID #X12: TyranitarTyranitar is the king of the Johto region. Obviously meant to provide a cool endgame Pokemon, Tyranitar is definitely a fan favorite from Generation II. Feared by all, Tyranitar could defeat any Pokemon it confronted. I mean, theoretically. No trainer uses Tyranitar, and the only way to get one is by evolving a Larvitar from the final post-game area, at a point where most trainers have been defeated and experience isn’t easy to come by. Suffice to say, Tyranitar’s super cool, but another Generation II design almost completely wasted by the designers. Tyranitar’s clearly based upon a kaiju, in that it’s an imposing looking Godzilla-like lizard that looks as though it could stomp Goldenrod flat (though it’s only about six feet tall, so it could probably only destroy a model). It’s name in Japanese in “Bangirasu” which Bulbapedia suggests means “Barbarian” or “Savage”; "Girasu" then means Hateful or Hated -Saurus. I’m not sure of that provenance, but it’s the best we’ve got. Before it’s pre-evolutions were created (Yogirasu and Sangirasu), Tyranitar’s name was just “Girasu” without the “Ban” appended to it; if “Girasu” is supposed to mean “Hated Lizard” or “Hated Saurus” then this feels like an obvious play on Dinosaur (“Terrible Lizard”) or Tyrannosaurus (Tyrant Lizard). The resemblance between the two, especially the spines on the back of both Godzilla and Tyranitar, is very close. I'm not Kaiju expert though, so if you think there's a kaiju that Tyranitar is more closely based on, let me know! My guess is that Tyranitar was created out of an interest (maybe Masuda’s) in having another Pokemon like Dragonite: a powerful endgame dragon which could serve as a rare, powerful Pokemon which was a reward for leveling up a rare but weak early stage. The parallels between Tyranitar and Dragonite are obvious in the final game: Both are three stage evolutionary families, both evolve into their final form at level 55, both evolve from weak early stages that require huge amounts of experience for each level, and both have 600 BST. The fact that both are large kaiju-like dragons (or at least dinosaurs) is probably also not a coincidence. It’s also telling that Tyranitar and Dragonite occupy very similar Pokedex slots. Dragonite is #149, the last Pokemon before Mewtwo, the ultimate legendary Pokemon from Generation I, while Tyranitar is #248, the last Pokemon before the two ultimate legendary Pokemon from Generation II, Lugia and Ho-Oh. I see this as less of a parallel between the two and more of an indication of the intent in creating Tyranitar. It seems very likely that during Masuda’s time in charge of the team, he would designate certain Pokedex slots for certain types of Pokemon that the game needed, and then Sugimori and the rest of the designers would fill that slot with an idea that fit the needs of the games. In this case, I think what’s very likely is that slots #246, #247, and #248 were designated by Masuda to be filled with a Dragonite-like pseudolegendary (though the term wasn’t around yet), and Sugimori complied by looking through his old designs and choosing something that would fit. The main differences between Dragonite and Tyranitar are aesthetics and personality. Dragonite’s extremely goofy; it gives kind-hearted “Puff the Magic Dragon” vibes. On the other hand, Tyranitar’s much scarier and much more imposing. Before you get to their Pokedex entries, their coloring, or even their facial impressions, the difference is obvious just putting their sprites back to back. Dragonite is designed around soft curved shapes, while Tyranitar is covered in prickly spines. So where did Tyranitar come from? There are two common theories. The first is that Tyranitar’s a rework of Gyaoon, the discarded Kaiju Pokemon from the very beginning of Generation I’s development. There’s a few reasons to take this theory seriously. First, Gyaoon seems to have been directly based on Godzilla. Gyaoon’s name is almost a direct transliteration of Godzilla’s iconic cry, and we have evidence that in the very early development of Pokemon, the initial cries were named directly after the Kaiju that these Pokemon were inspired by. Arguably, this is probably why Gyaoon didn’t appear in the final games: it’s a very generic design that just looks too close to Godzilla. Rhydon, created right before Gyaoon, was also an extremely generic Kaiju-like design, as was Nidoking and to some extent Kangaskhan. Given how many early Pokemon were direct references to Ultraman and Godzilla kaiju, the team may have felt Gyaoon didn’t have enough originality to its design to compete with other generic Kaiju-mon. The theory a lot of people argue for is that Tyranitar was a rework of Gyaoon, tweaked to be less generic and fit into Generation II. I can see some of the logic here: Tyranitar has obvious inspirations from Godzilla, like Gyaoon. Given the similar inspirations, it’s possible to imagine that Sugimori went back to his first take on Godzilla, played up the spiky bits more, and drew a new sprite. However, there’s not much more evidence for this idea, and some evidence against. One reason I have to doubt the connection is that Tyranitar and Gyaoon is the difference in emphasis in both their designs. Gyaoon’s major unique design feature is its large, bellowing mouth. True to its name, it’s screaming it’s cry at full blast; I’m almost certain that if it had made it the final game, Gyaoon’s Pokedex entry would have made note of its huge maw. Meanwhile, Tyranitar has a very tiny mouth. It’s not even opening it’s mouth in the Gold sprites, and when they experimented with a both where it’s screaming, it’s gape is incredibly tiny. Instead, Tyranitar’s aesthetic focus is on its chunky, imposing body and on the spikes that go down its back. Even though they have similar inspirations, Gyaoon and Tyranitar took this same starting point in very different directions. The biggest reason to doubt the Gyaoon/Tyranitar connection, however, is that the second origin’s explanation is so much better. Let’s start by looking at the April, 1997 cover for the Japanese gaming magazine, Microgroup Game Review. This edition of the magazine contains an interview with Sugimori, and a cover designed by him: (This is a clean version of the cover, without the text. Credit to Dr. Lava for the high resolution scan of this artwork) Hold on. That’s Tyranitar. But it’s way too early for Tyranitar to appear. This magazine was released seven months before Spaceworld 1997, and a wopping two years and two months before the first data we have that features Tyranitar’s sprite. Tyranitar isn’t anywhere in Spaceworld ’97; not even something that could have been a early design for it. There’s nothing in the Korean Index that even has a chance of being related to Tyranitar. So what’s going on? We’ve talked about this magazine cover before, when we covered Hitmontop (which could have been inspired by the weird Clefairymontop in the bottom left of that artwork). This cover is infamous in Pokemon fandom, because of how it seems to offer us a look at beta Pokemon designs. In Tyranitar’s case, it seems to suggest that Tyranitar predates Spaceworld 1997, even though, for no explicable reason, Tyranitar doesn’t appear anywhere in that build. However, I think there’s an easy explanation. In the interview conducted with Sugimori inside the magazine, Sugimori denied that these were actually Pokemon on the cover. Instead, the cover showed “another world, like Pokemon, but not Pokemon.” In Sugimori’s mind, these weren’t Pokemon, just drawings of Pokemon-like creatures. Now, Sugimori could be lying here to hide that these are beta designs, but really, what’s the point? If he hadn’t wanted to show off beta designs, he could have just put something else in his cover sketch. Instead, I think that Sugimori was telling the truth: at the time he drew this cover, he was just doodling some designs without any thought that they might find their way into a Pokemon game. Tyranitar does exist prior to Spaceworld ’97, in other words, but at the time is was just something Sugimori drew on his own. It wasn’t a Pokemon yet. In that same interview, Sugimori explained that he got his ideas for Pokemon designs by constantly going back to his earlier ideas and iterating them, refining them more and more until he’s happy with them (another piece of evidence Gyaoon could be an early Tyranitar, admittedly). But I think the more important take-away from this interview is that Sugimori is expressing the likelihood that, though these not-Pokemon on the cover weren’t being considered at the time, he was open to going back to designs like these when looking for inspiration. And I think that’s exactly what happened here. After Masuda took over development and started demarcating certain Pokedex slots for certain needs, he asked Sugimori to design a counterpart to Dragonite which could fill the same role in-game. Sugimori, who was probably quite busy and needed something that would work in that role, went back to these Microgroup designs and took the one he liked most, named is “Girasu,” and added it into the Pokedex. Though he may not have intended to use that design for a Pokemon, when there was an obvious need for a dragon or dinosaur like imposing super-mon, this one fit the bill perfectly and Sugmori would have been a fool to ignore it. Could both of these explanations be true? That is, could the Microgroup cover feature a reworked Gyaoon? I just don't think that's likely. If Sugimori truly had no intentions for these to be Pokemon, why would he spend his time reworked an old discarded Pokemon design? If he were going back to old designs, why change Gyaoon so drastically rather than just use it here? Why include a real Pokemon next to two designs that were definitely not anywhere in Generation I? It seems implausible that Gyaoon factored into this cover at all. This also explains why the two earlier stages of Tyranitar don’t get conceived of until months later, and why the team is using random sprites as placeholders for them in the June 1999 collection of sprites we have, whereas Tyranitar’s sprite is mostly finished. If Tyranitar were just a repurposed older design, then presumably Sugimori could easily turn it into a sprite. Whereas the other two stages would still need to be work out, since they weren’t part of the deal initially. Obviously, Tyranitar was the main inspiration here, and the team fumbled for a few months trying to come up with something that fit alongside it. Notice that it took them until August to create Pupitar and just they only figured out Larvitar in the very final stages of development: In fact, one of the earliest sprites we have for Larvitar is a heavily modified version of the Tyranitar sprite, made smaller and more compact so as to look like an earlier evolution. Showing again that Tyranitar was the base, and the other two were meant to fit in to match it. If you look closely at this Larvitar sprite, you'll notice that it isn't even an original sprite. It's actually just different parts of the Tyranitar sprite, copied and pasted closer together, to make it more compact. The eye might be the only new part of the whole sprite. One thing that’s always bothered me about Tyranitar has been its typing. Obviously, Tyranitar’s a Dragon-type: I mean, just look at it! However, for whatever reason, the team went with a bizarre Rock/Dark typing. It doesn’t look like the team ever wavered about this typing or thought about alternatives: from the earliest data we have of Tyranitar (in April 1999) it was already a Rock/Dark type, and that never changed. It’s moveset also has plenty of Rock and Dark type moves, suggesting that, unlike a Pokemon like Gligar, the moveset was made to closely reflect this typing. Anyway, my guess is that the team chose the Rock/Dark typing for two reasons. First, though Tyranitar is obviously a Dragon to my eyes, in think in conception, Sugimori was considering Tyranitar to be some sort of dinosaur. Given that all the other ancient Pokemon (Omanyte, Kabuto, Aerodactyl) are part Rock-type, a dinosaur might naturally have been a Rock-type in his mind. Secondly, as we’ve talked about before, I think the team was intentionally trying to keep the new Dark and Steel types very rare in the pre-Elite Four sections of the game. Commenters have suggested that this was to make them feel special, and I’m inclined to agree; I also think that since the strengths and weaknesses of these types were in flux, the team didn’t want too many of them too readily available because of the effects they could inadvertently have on gameplay balance if the types’ weaknesses were changed at the last second. The team might have chosen Dark-type for Tyranitar because as a way for it to show off the new Dark-type on a very powerful Pokemon, and thus make the type feel even more alluring. Certainly, if you went up against this thing and it Crunched your ace into oblivion, it would teach you that Dark-type is something special. In the same way that the team made Lugia Psychic-type because of the association of Psychic-type with powerful Pokemon, I think they were also trying to associate Dark-types with power by giving that typing to this monster. Of course, this strategy would work more effectively if any trainers actually used Tyranitar. Imagine if Karen used this guy instead of a Vileplume during her Elite Four battle, for instance. There’s no evidence for it, but I’d even suggest that Tyranitar might be the reason “Bite” was repurposed into a Dark-type move in Generation II. We know that Bite was still normal in the Spaceworld ’97 build, and so the choice to change it to Dark had to have happened later, presumably during the 1999 rework. Given how early Tyranitar shows up in this rework, and given that a giant intimidating kaiju like Tyranitar might Bite its opponents, the team may have changed the move to better reflect Tyranitar’s flavor. Then again, lots of Pokemon learn Bite. If the team had decided to keep Houndour and Houndoom around about this same time, then those two Dark-types might also have been the impetus to change the typing of the move. The reason Tyranitar is mostly unused in Gold and Silver probably comes down to the lack of development time. That probably didn’t affect Tyranitar itself all that much: it had 600 BST stats as early as April 1999, and those stats only slightly changed throughout the rest of development. Instead, it was probably underused because Larvitar and Pupitar took so long to get finalized. We’ll talk about them when we get there. One interesting final piece of Tyranitar’s development is that you can tell the team was trying to work out what Tyranitar’s personality should be—and, by extension, what flavor the Dark-type should have—as development progressed over 1999. For instance, in early (that is, early August 1999, so pretty late in development) Pokedex entries, the writers characterized Tyranitar as a tank, impervious to damage, that basically didn’t care what happened to it: Early Gold Entry: Because its body won't yield to any attack, this Pokémon won't fight without good reason. Early Silver Entry: The forehead is said to be the hardest part of its body. There's nothing that it can't crush. Both of these entries really stress Tyranitar’s identity as a Rock-type Pokemon. Meanwhile, compare those entries to the final ones: Final Gold Entry: Because its body won't yield to any attack, it will pick one fight after another. Final Silver Entry: It's so powerful that it can alter nearby terrain without breaking a sweat. It is so brash and cocky that it pays no heed to anything around it. In comparison, these entries really stress Tyranitar’s angry or bullying nature. It suggests to me that as the team became more certain that Dark-types would be aggressive, mean-spirited tricksters, they changed Tyranitar’s personality to match. I really like Tyranitar, and I think it was a great idea for Sugimori to bring this design forward to serve as a Dragonite stand-in. I like that both these Pokemon look really cool and powerful, but they have completely different natures: making Tyranitar much more dangerous looking makes it feel unique compared to Dragonite without taking away from it’s design at all. Tyranitar has a great design, which I’m sure made it all the more difficult to create cool Pokemon to evolve into it. ID X13: SentretIt’s a bit odd to me that Sentret was such a late addition to the Generation II roster. A basic early-game Pokemon, Sentret feels like the sort of thing they would have settled upon first, so that the beginning of Gold and Silver was as polished as possible when showing it off to people. Still, this little guy didn't make his appearance until after the 1999 reboot of development. Let's figure out why. But first, let’s talk a little about ID 360, the Ninja Squirrel that we saw much earlier in the Korean Index. I said back when we discussed it that I suspected ID 360 might have been an early design for Sentret; I’m still not sure about that, but there are at least a few things that make me suspicious. First of all, consider that ID 360 first appeared in Period 2a. That section of the Korean Index was primarily designed by Nishida and Sugimori, and almost every sprite in that section ended up used in Spaceworld ’97, with the exception of Proto-Cyndaquil (which obvious got used soon after Spaceworld ’97) and another flying squirrel like Pokemon that could have been a precursor to Furret. The fact that ID 360 appeared in that sequence automatically makes me think it was more important to the designers than the more-or-less throw away designs of Period 1e, right before. In addition, though ID 360 has a strange style to it—especially if you consider its eyes and it’s weird front-on perspective—it is found right next to the Volbear line of starter Pokemon, which were created by Nishida. It's certainly a bad looking sprite compared to Nishida's normal work, and it's possible ID 360 was made by the designer of Period 1e and somehow input into the Korean Index out of order. Still, it's possible it was in Period 2a for a reason. Which is important, because Sentret also strikes me as an Atsuko Nishida design (albeit one that was probably touched up by Sugimori). Nishida didn’t have much role in the rebooted 1999 period of development, and wasn’t actually credited with any of the graphical designs in the final game, but it’s likely she put some work into these sprites but wasn’t available to add the finishing touches, which Sugimori did. In this case, if you look at the earliest sprites we have of Sentret, and look at its face. The minimalistic style of the face doesn’t look like many finalized sprites we have, but the mouth is quite reminiscent of Pikachu’s Green sprite, and the dot eyes Sentret uses remind me of Ditto. This is all a lot of speculation though; Sentret could have been made by anyone. If ID 360 really was an early Sentret, the team probably dropped the Ninja motif to make Sentret into a more simple design, better suited to the early game. A Ninja squirrel might be a cool idea, but it’d be weird to find a million ninjas all running around the first route of the game. They certainly wouldn’t be doing a very good job of sticking to the shadows, as ninjas should. If the team did decide to revisit the Ninja Squirrel design, these ninja sprites were probably not ever actually in the game. We know that Sentret took the spot in the Pokedex previously held by Monja, the Tangela pre-evolution found in Spaceworld ’97. Monja very likely did not survive the reboot of development, and by April 1st, 1999 (our earliest snapshot), Monja’s slot had been completely cleared, only have placeholder stats and the Normal typing. That means that there was nothing in that slot entirely, though Monja's sprite may have briefly persisted. Sentret appears soon after. Sentret (and Furret) gain real stats on April 13th and those get tweaked to their final stats ten days later, April 23rd. To my knowledge (so far), these are the only Pokemon we have which got updated stats in at this time: every other Pokemon added in Era IV either already had stats by April 1st, or got them later, in June or July. That means we can pretty definitively date Sentret and Furret as the dividing line between Pokemon which were mostly done before our first data from the development reboot, and those that were worked on during 1999. Note that Sentret was actively being worked on throughout this period: its sprite, for instance, went through a number of small tweaks. The team obviously wasn’t sure how wide it’s body should be, and they even experimented with a very squished-looking version of the sprite (though I’m not positive that was ever really the actual sprite, or if it was just a failed experiment). From the sprites, it appears that Sentret was based on a Flying Squirrel, but it’s early name implies a different origin. Initially, Sentret was called “Moruten,” which is both a pun on the animal “Marten,” and uses the kanji for Martens for the syllable “ten”. Martens, in case you aren’t up on your furry animal taxonomy, are thin, brown or black furred mammals that are closely related to weasels, to ferrets, or to stoats. They’re all pretty cute, but are only rarely domesticated. The name was probably changed because, while Furret looks a lot like a Marten (and of course its English name is a play on “Ferret”), Sentret looks nothing like one. (A Marten on the left, and a Ferret on the right) The “Moru” part of the name is more interesting, because “Moru” can mean “to guard,” or to protect. It’s final Japanese name, Otachi (Tail weasel) doesn’t make any reference to Sentret’s nature as a guard, but interestingly, the name “Sentret” derives from the word “Sentry,” which has a similar meaning. Sentret’s final Pokedex entries make note of this quirk of it’s personality: Gold Final: A very cautious Pokémon, it raises itself up using its tail to get a better view of its surroundings. (Note: TCRF translates “cautious” to “highly vigilant,” which is likely a better translation). Silver Final: It stands on its tail so it can see a long way. If it spots an enemy, it cries loudly to warn its kind. Crystal Final: When acting as a lookout, it warns others of danger by screeching and hitting the ground with its tail. Interestingly, an early Pokedex entry gives it a different personality, claiming that Sentret is cowardly: July 20th: This cowardly Pokémon raises its tail to see as far away as it can. Of course, when I read these Pokedex entries, and the fact that Sentret is “guarding” something or is a “sentry,” I wonder what it’s guarding. And I think I know the answer: Berries. Ever since Spaceworld ’97, Generation II had added held items as a mechanic. There are numerous items in the data of Spaceworld ’97 that either increase the damage moves of a certain type do, or decrease the damage done by moves of a certain type, so long as your Pokemon is holding the item. However, the held item mechanic wasn’t very developed in that build: no Pokemon are actually holding any of the items, for one, and there are no items that activate in battle. Both of those important dimensions of the held item mechanic were added later. Pokemon actually holding items in the wild only first appeared on July 30th, 1999, when 41 different Pokemon are given held items. Importantly, as TCRF has noted, Sentret and Furret are the only Pokemon holding berries at first, though they are eventually given to other Pokemon. They hold these berries rarely, but for at least one period of development, Sentret had a berry in its common held-item slot, meaning that it had a 25% chance to have a berry when you fought it. My theory is that Sentret and Furret were added into the game as early game Pokemon precisely because they were Pokemon which would hold and eat berries in battle. If the team wanted to teach the player how to use these new held items, have a very early Pokemon consume a berry while you fought it would be an obvious tutorial on how they were used; the fact that catching a Sentret would often lead you to obtaining your first berry would also encourage the player to make use of them, even before the first gym. It also makes sense of Sentret’s lore: if Sentret is really a guard or a sentry, then it was probably just trying to protect its berry stash from trainers encroaching on its territory. By the final game, this doesn’t really come through, given that berries are more commonly held by all sorts of Pokemon. But if this was the original intent of Sentret, then it makes sense of why an early game Pokemon was dropped into the Pokedex so late: the team needed one last mechanic explained, and the old Ninja Squirrel was heavily repurposed to fill that niche. Or not. Sentret could have just been added because the team realized there wasn’t enough diversity on early routes: after all, in the final you can still find Pidgeys and Ratattas, and without Sentret it’d be hard to even know you’re playing a game with new Pokemon. Or it could be the case that the team knew they didn’t want so many baby Pokemon, and Monja’s excision opened up a new slot that got filled with the first idea someone had. It is definitely a hole in my berry theory that Sentret and Furret got added in April but Pokemon didn’t gain held items in the wild until July 30th; Sentret and Furret could have just been convenient already-existing creations to show off the berry mechanic, not created on purpose to show it off. So there are lots of other ways to read the data we have. But enough about Sentret. Let's turn to its adult form, Furret! (Fan art by Amy Walek!) ID X14: FurretThere isn’t actually a lot new when we turn our eyes to the grown up Weasel/Ferret thing. Furret’s design story is basically identical to Sentret’s; they both came into the world at the same time, and were modified at the same time. I do think Furret is pretty cute! If that counts for anything. First of all, Furret, like Sentret, got it’s first set of stats on April 13th, and then got them updated on April 23rd. Its stats are, for the most part, just twenty or thirty point improvements on Sentret’s, though it strangely heavily improves on Sentret’s two worst stats: HP and Speed. In HP’s case, it gets a fifty point boost, going from a pathetic 35 to a respectable 85; in Speed, the ferret goes from an awful 20 in speed (I had no idea flying squirrels were so sluggish!) to 90! I'm not sure why there's this drastic increase in stats. Maybe they wanted to make Sentret especially weak so it wouldn't be a threat on early routes, but wanted to make it viable in the late game if you put work into it. It's early stats, before they were revised on April 23rd, by the way, are a bit odd. While Sentret’s are kind of generic (Mostly 50s and 40s in all stats except speed, which is 55), Furret has an odd 64 in Attack and 58 in defense. I have no idea why they would choose those numbers. If there’s a theory that Sentret is a rework of ID 360, the Ninja Squirrel, then there’s also a reasonable chance that Furret is a rework of ID 377, or the Windsock Fox sprite also from Period 2a. While the Windsock Fox was probably made originally by Sugimori and not by Nishida, everything else I said about ID 360 applies here as well. ID 377 is a remarkably good sprite (unlike 360), and it’s the only other sprite other than ID 360 from Period 2a to not make it into any of the builds of the games. It also has the long, tube-like appearance of Furret, though its face doesn’t quite look like the same sort of woodland creature (most commonly, people read it as a fox). We have no way of knowing for sure if the Windsock Fox is related to Furret; I’m personally less sure of this connection myself, which is why I didn’t include it on the graph I put at the beginning of this entry. But it is possible that, when the team needed a new early game ‘mon, Sugimori looked back at this design and decided to redraw it as something that had a more obvious species connection to Nishida’s flying squirrel sprites she used for Sentret. Like Sentret, Furret went through a couple name changes. It’s first name was “Ninten” which uses the same “ten” that Sentret used that denotes a Marten. However, I have no idea what the “Nin” part means and I can’t help but wonder why it happens to have the same name as the protagonist of Earthbound Zero, who was named after Nintendo. That name was quickly changed to Dakuten—I have no idea what that name means—and finally to “Ootachi” which is similar to Sentret’s final “Otachi” but the long “O” denotes the Japanese word for “Big”. So essentially the two are “Tail Weasel” and “Big Tail Weasel.” The most interesting thing about Furret is its early sprites. First, it looks completely different! This early Furret doesn’t quite look like the tube of fur it’s final design looks like: it’s got a more distinct tail that is thin at the base and resembles a bit of a pom-pom. It’s fur is also not as detailed, and lacks the stripes across its whole body that the final version has. It does, however, have the exact same poses as the final design. My guess is that Sugimori touched up the design right before the end of development, to make it look a little more like a Pokemon and a little less like a generic animal that happens to be in a Pokemon game. I have no idea who did the earlier sprite: it could be Sugimori, who then touched it up himself, or it might, like Sentret, be a rough Nishida design that Sugimori finalized since Nishida was busy with other projects. Obviously, the thing that stands out most about these early sprites is the palette: Furret’s bright pink! That’s usually the thing everyone comments on, but it’s probably not as exciting as you might think. As FrenchOrange was the first to note, Furret’s pink palette may have been a leftover from Gyopin (the Goldeen Pre-evolution). Given that Furret replaced Gyopin in the Pokedex (we see this from the scratchpad data we have), and how the pink palette matches Goldeen, it's likely that the palette comes from a palette given to Gyopin before its removal. Thus, Furret was probably never intended to be pink. The team just didn’t get around to giving it a proper palette until the last couple months of development. This could also explain why the sprite got a pretty heavy rework: Sugimori may have felt that a brown palette didn’t work very well with the old sprite, and may have resprited it to specifically work better with a palette like the one Sentret had. (Thanks to OrangeFrench for the recolored sprite on the left!) The early sprite does look pretty bare and textureless in that palette. In any case, something about the accidental pink palette made an impression, because it did make it into the final games as Furret’s shiny palette! Sort of; the palette isn't exact, but you can see the similarity. If you were ever wondering why Furret had such an out there coloring when you found a shiny one, now you know! Someone on the design team clearly got used to the pink. Anyway, that’s all there is to say about Furret! Onto our next Pokemon! ID X15: SudowoodoSurprise! It’s Sudowoodo! I honestly love Sudowoodo. I loved it as a kid, I love it now, and it’s always one of the first Pokemon I use when it becomes available (It’s an early game beast in Scarlet as long as you make sure to make it remember its level one moves). But what’s up with Sudowoodo? It’s a weird concept and a weird design, and I’m not completely sure how the design team came up with it. First off, the name’s quite clever. In English, Sudowoodo's name is a portmanteau of “Pseudo” and “Wood,” which literally means it’s not wood. A fitting name for a Rock Pokemon that pretends to be a tree! The Japanese name, however, is even better. “Usokkie” is a combination of “Uso,” meaning “false” and “ki,” the word for tree. Bulbapedia also adds that the full name is also a pun: Usokkie looks similar to Usotsuki, which means “liar.” So the name works on multiple levels, and pretty accurately describes what Sudowoodo does. We know that Ken Sugimori designed Sudowoodo, because he said so in an interview about Generation II’s development (even without this interview, the sprites have tons of hallmarks that make this an obvious Sugimori design). Sugimori explained that Sudowoodo was created purposefully as a roadblock Pokemon, and that he asked the scenario designers to add in a part into the story where a Pokemon would block the player’s path. As a tangent, Dr. Lava has pointed out that this makes Sudowoodo the exact opposite of Snorlax design-wise: Snorlax was initially created because the team needed a way to block a path in the map they had already created, while Sudowoodo was made and then the designers found a reason to add in a blocked path. Sudowoodo’s creation—as a new Pokemon that served the same function as Snorlax—strikes me as part of a larger trend in Generation II’s development. As I discussed above, initially the team wanted entirely new HM moves that served the same function as the original HMs; Spaceworld ’97 also had a skateboard the player could use, presumably as a fast-travel option that would replace the bicycle from the first games but serve the same purpose. While most of these ideas got discarded, Sudowoodo seems like an attempt at the same idea. The team could keep things fresh by not reusing Snorlax (though, of course, in the end they did, in the Kanto region of the games), but they could still have the same gameplay of locking locations behind plot progression, and keeping the overall feeling of the gameplay the same. I have no idea why Sugimori settled on the idea of a Pokemon that looks like a Grass-type tree but is actually, secretly, made of stone. It’s a strange concept, stranger still when it turns out you need a Squirtbottle of all things to make it move (I guess since it's weak to water and thus even a squirt hurts it?). There’s some resemblance between Sudowoodo and Bonsai trees (tiny, carefully maintained trees that function as miniature models of much larger trees); in fact, when Sudowoodo got a baby version in Generation IV, it was named Bonsly and designed specifically to look like a Bonsai tree. And of course, one of Sudowoodo's main traits is that it's a false mimicry of something else, which is fitting for a Bonsai tree, which isn't actually the type of tree its mimicking. So I wonder if Sugimori drew inspiration from the idea of a miniature tree that was actually artificially created to resemble something it was not. While all of the above is pretty straightforward, there is something off about Sudowoodo's development. It’s pretty certain that Sudowoodo replaced Kyonpan, the Panda/Vampire Ghost Pokemon from Spaceworld ’97 which had evolved from Norowara, the voodoo doll. Though we don’t have sprites for Kyonpan after Spaceworld ’97, Kyonpan’s sprites do appear on the scratchpads for Sudowoodo, and most interestingly, until July 1999 (even more than a month after it got a sprite!) Sudowoodo is a Ghost-type Pokemon with drastically different stats. The most likely explanation is that Sudowoodo inherited some stats from Kyonpan and the team just didn't have time to give Sudowoodo its own unique stats and typing until July. Still our earliest sprites come from June 1999, so we have no idea if this Pokedex slot used updated Kyonpan's sprites or there was something else there entirely. If this is evidence that Kyonpan survived into the reboot, there’s something suspicious about the data we have. Interestingly, if you compare the stats of this ghostly echo of Kyonpan to Norowara—which, importantly, still existed in the game until at least Spaceworld ’99—then the stats don’t match up at all. Take a look: (These stat tables taken from TCRF) These aren’t the stats of an evolutionary family; Kyonpan’s stats are completely different from Norowara's, and they share BSTs (both have 400, which is low for an evolved form like Kyonpan anyway). In addition, Norowara was, by this point, changed to Ghost/Dark typing, and no Pokemon goes from a dual-type to a single-type upon evolving. There’s also no evolutionary data connecting the two Pokemon. Which means that if Kyonpan survived the 1999 reboot and survived in the slot Sudowoodo took for awhile, it must have undergone quite a bit of changes. Even before Kyonpan was slotted to be removed from the games (as it was, clearly, by June; it’s stats and typing were still in the data, but Sudowoodo’s sprites show that it’s replacement was just a matter of time), it had been severed from Norowara, making them two unique Ghost-type lines. This makes some sense; afterall, Norowara’s design and inspiration had almost nothing to do with Kyonpan's. But that means that for a few months in development, the team was considering having two unique, unrelated, new Ghost Pokemon: one pure Ghost type, one Ghost/Dark. Given that Twinz was also still in the mix at this point, there’s an alternate world where Generation II introduced three original Ghost-types (Four if Girafarig or its evolution had become Ghost). That’s really cool! I would have loved more Ghosts in Generation II. It’s also possible that the Ghost which preceded Sudowoodo wasn’t Kyonpan at all, but a completely original Ghost which existed for a few months after replacing Kyonpan. Given that the pre-Sudowoodo Ghost didn’t share typing or evolution data with Norowara, it could have been so completely redesigned away from Kyonpan that it was an original creation by that point. Who knows; we don’t have sprites going back early enough to check. I’d really love to get a hold of a December 1998 or January 1999 build of the game; alas, we’ve probably gotten everything that’s going to turn up. Of course, Sudowoodo could have been originally designed as a Ghost, but I somehow very much doubt that. After all, how is an incorporeal spectre supposed to block your path? One thing that could help us get a picture of what the pre-Sudowoodo ghost was like would be a glimpse into Sudowoodo’s moveset around May or June, back when it still had the Ghost-typing. I wonder, for instance, if it still had a moveset that was more suited to something like Kyonpan, or if it had already gotten a hold of Rock-type moves. If you have access to this data please let me know what you find! Anyway, by July 18th, Sudowoodo got it’s Rock type, and by July 30th, it got original stats to replace the Ghostly ones it had before. These new stats emphasized Sudowoodo as a physical Pokemon: it had low speed, but high attack and defense. Fitting for a Rock-type. Even better, Sudowoodo became the very first pure Rock-type in Pokemon history! All the Rock-types from Generation I were either dual-type Ground types or water-types, or Aerodactyl. This is almost assuredly because there was no Rock-type through most of Generation I's development, as it only got split from Ground-type very close to release. As a result, any Pokemon who gained Rock-type probably kept their original typings from before Rock existed as well. So it's cool that Generation II finally had a chance to give us a Rock-type with its own identity! Given how much of this period in development was driven by filling existing holes the in Pokemon roster, I wonder if the brief for Sudowoodo was initially to design a pure-Rock type Pokemon and Sugimori came up with the Bonsai aspect of it later. If so, this may have been a funny joke of Sugimori’s: rebelling against Masuda’s directive to create a Rock-type, Sugimori gave him an obvious Grass Pokemon design and then told him, “Don’t worry, it’s actually a Rock-type just pretending.” I wish Sudowoodo was a slightly better Pokemon; unfortunately, I almost always end up replacing it by the end of the game, as its stats are just not good enough to fight the Elite Four. But I adore that it exists. I’m the world’s #1 Kyonpan fan, and I’ll always mourn that we never got that awesome hopping-panda-vampire. But if we had to lose Kyonpan, I’m glad it was replaced by such a winner. (The artist who made this awesome piece is Mazee, or まぜう, and can be found here: https://www.pixiv.net/en/artworks/60077051)
ID #X05: TogeticTogepi was designed to be a cute baby Pokemon. Thus, from the very beginning, the designers must have known that it would have to hatch into something. There’s not indication that they worked on a Togepi evolution before the development hiatus, but upon Masuda taking over development, I’m sure it was one of their biggest priorities. Togetic—better translated as “Togechick” to indicate that it’s a young hatched bird—seems to have had a spot in the rebooted Pokedex reserved from the minute they reordered it, even if the designers took their time to figure out exactly what Togetic would be like. Our earliest data from the reboot begins in April, 1999, though we only have name data, stats, and typing from that point, not sprites. In the first collection of sprites we have—from June 1999—Togetic’s sprites are already complete, and would remain exactly the same until the final versions of Gold and Silver (In Crystal, they tweaked its palette to better reflect the white color of Sugimori’s artwork, and the backsprite got tweaked to make clear that Togetic’s crown went all the way around its head). It’s got to be said: Togetic’s got a strange design. Like Togepi, it has a yellow palette, though its body is a darker shade of yellowey orange than Togepi’s. It’s got a strange, hunched over pose that looks a little like a chick, but it’s neck is far too long to make this look natural. And though Togetic no longer has the eggshell that its baby form had, it still had the blue and red symbols on its belly that were on the eggshell. That last part’s a curious choice: it makes no sense biologically for Togetic to keep the design on the eggshell on its skin, except that the design serves as a visual identifier to link it to Togepi. (Sugimori's official Togetic artwork for Generation II) Its body is also a strange shape: its more or less the exact same shape as the eggshell was, as if the eggshell was exactly covering Togetic’s torso like a glove, and the limbs and head just sprouted out of it. Something about Togetic’s design feels off and a little bit strange to me, though it’s difficult to point to exactly what. I do have a guess at the reason though. By the time the team was designing Togetic, Togepi’s design was already locked in. Togepi had already shown up in the anime already, and had appeared on some of the promotional materials. Made after the fact, Togetic had to grow naturally from a design that already existed, both being a logical evolution of Togepi's aesthetic as well as keeping visual reminders that made it look enough like Togepi. Thus, I think the designers just took the most obvious route. Taking Togepi’s basic body shape as a starting place, they just extended each of Togepi’s limbs, to give it a more grown up look. In fact, on the scratchpads, there’s an early sketch of Togetic that seems to be exactly this. In the sketch, Togetic’s basic silhouette is drawn out from a rough oval about the size of Togepi’s original body. Notice how the things they really wanted to get right in this initial sketch was Togetic's face, which is basically Togepi's but now on a neck and with a nose, and the shape of the body. Though these sprites are completely finished by June 1999, other indicators suggest that they weren’t done much earlier than this, unlike say Heracross and Totodile. Mainly, if you look at the data in the slot Togetic takes up circa April 1999, what you find is that Togetic exists, but its stats and typing are just copied over entirely from Togepi (it does, at this point, have its final name). So it seems likely that the origins of Togetic went something like this: --(Late 1998/Early 1999) The reboot team reorders the Pokedex, and determines that Togepi needed an evolution. They designate Pokedex slot #210 for this evolution, and copies over Togepi’s stats into the new slot. --(Early 1999, up to April) The team comes up with the name “Togetic.” --(Between March and May 1999) Someone (Probably Sugimori) designs and edits Togetic’s sprites. --(July 30th) Togetic’s is given unique stats. It is also given Normal/Flying typing for the first time. This might seem straightforward, but I bring it up because (as I alluded to in the last article), it seems to be a very different style of creation than the early development of Pokemon 2. During 1996 and 1997, the development of new Pokemon seems to go like this: --Artist brainstorms common animals, and comes up with a Pokemon-esque twist. They design sprites of this and put it in the communal sprite bank. In some of the latter stages of 1997, these may have been created after a vague design need from Sugimori or Tajiri: “Design an ice-type” or “design an evolution for x” --These sprites are put into a developing Pokedex. Similarities between different designs lead to the developers connecting them evolutionarily. --Types are determined by the look of the sprites; movesets are created to reflect that typing. --Names are created. It’s a completely different process. Early in development, the sprites came first, everything else second; by 1999, the Pokedex slot was designated to a particular type of ‘mon, with the sprites being one of the very last things finished. Part of this may have been how late they were in development and much the goal was now to fill in gameplay niches they needed. Part of it might be that Masuda’s style was different than Tajiri’s. Probably a bit of both. One’s not worse than the other; however, I do find the drastically different approach in Era IV interesting. One regret I have is that no one has systematically documented the moveset data we have between Spaceworld ’97 and the final build. As a result, I have no way of tracking the thinking of the designers as they tweaked the movesets, so the best we can do is speculate (see the Gligar entry, below). But one thing worth mentioning here is that at some point in this period, Togepi and Togetic got a drastic moveset change. In Spaceworld ’97, Togepi got a pretty decent moveset. It got two Flying moves, Peck and Mirror Move (Japanese: Parrot Mimicry) (which makes a good case that the name “Togechick” would have been a much better translation) along with Protect, Recover, Spike Cannon, and Skull Bash. It’s a balanced and pretty unique moveset. Few other Pokemon use Spike Cannon, and Recover is rare among Pokemon that aren’t Psychic-type. At some point—probably after the creation of Togetic—the team decided that the Togepi line should be almost purely defensive, and the team took away all of the family’s offensive options, though they gave it the incredibly inconsistent Metronome, which could technically do damage. My guess is that this was in response to the anime: Togepi first appeared on the anime during the development hiatus, on June 25th, 1998. That meant that by the time development restarted in early 1999, the anime had had months to establish Togepi’s personality. Given that Togepi never attacks in the anime, except to wave its fingers, cast metronome, and cause a giant explosion, the development team probably chose to align the new moveset with what the anime had built. However, this also meant that Togepi lost the Flying-type bird moves that probably inspired Togetic’s design in the first place. Sure, Togepi may not look like a Pokemon that can Peck, but Togetic certainly feels incomplete without at least one Flying move! I think in modern Pokemon, this is the sort of thing the team would fix by giving Togetic a move it learns upon evolution (And that’s exactly what they do, now that I check—it gets Fairy Wind, a damaging move, upon evolution). But it very much does feel to me like Togetic got the short end of the stick when they changed Togepi’s moveset. The only other notable thing about Togetic’s development is a slight change in it’s Pokedex entries. It’s original (or at least, earlier) Silver entry was sort of melancholy: "It softly flies through the air, searching for a kind person that will make it happy.” I honestly kind of love that entry: it’s cute and evocative of a Pokemon searching for a friend. The team downplayed the sad parts of this in the final entry: “If it's not beside a kind person, it loses its pep. It floats in the air without moving its wings.” This version is less likely to be a downer, but I think the first one was much more evocative and elegant. Togetic’s the kind of Pokemon that’s no one’s favorite. If you like it, you'll like Togepi more; if you don’t, you probably still like Togepi more. It’s not surprising that Togetic was designed to an existing need: there’s very little passion in this design, and a huge feeling of “well, this baby needs an adult version, let’s see what we can come up with.” I have a feeling that the design of Togekiss, in Generation IV, was an attempt to revise this, and give Togepi an evolution that feels much more like it has an idea behind it and wasn’t just made to a menu. As much as I don’t like Generation IV designs, at the very least Togekiss looks like a design the team was excited to create. (Fanart by Loadez) ID #X06: GranbullNext up is Granbull. Granbull’s got a lot of similarities to Togetic, and was probably designed for much the same reason, and as part of the same process. As a result, my guess is that these two were probably made very close together. In the end, they were probably both made so that Snubbull and Togepi--the two mascots of Generation II--would be useful in the later parts of Gold and Silver. It's not clear which of these evolutions came first. On the one hand Togetic’s sprites were completed by June 1999, while Granbull was still missing its backsprite by that point. Still, Granbull’s front sprite went through no changes until the final, so the team may have already iterated that a few times by June. On the other hand, Granbull’s stats were complete by April 1999, while Togetic’s were still just copies of Togepi’s by that point (it didn’t get its real stats and its typing until July 1999, three months later). So who knows which came first—the bulldog or the egg. Take these two entries as more or less simultaneous. Togepi needed an adult form, given that it was the poster child for the new baby mechanic. But there was something else going on as well. If you’ll remember back to Era III, Togepi and Snubbull were added to the Spaceworld ’97 Pokedex at the last second; we can be pretty sure of this because they both are in an unorganized section at the very end of the Pokedex. This final section, after the Legendary Pokemon, contained a bunch of 'mons that clearly didn't belong there: in addition to Togepi and Snubbull, Sneasel, Aipom, and Leafeon were in this late section. This suggested that this whole group were very late additions to that build: they were probably thrown in so late that the team hadn’t yet had time to reorder them into more logical positions. While I'm sure these late additions are all there for varying reasons, I'm pretty confident that Togepi and Snubbull were added because of their cuteness factor. By Spaceworld 1997, when the game was first being shown off, the team probably planned for to show off the new games with these guys: they probably already knew that Togepi would appear in the anime as a recurring character, and at least according to one of my commenters, Snubbull was used prominently in the promotional materials for the new games at Spaceworld. Thus, while Snubbull appeared pretty early in the Korean Index—in Period 2a, a large chunk of Sugimori’s earlier designs—it had probably been scrapped for a decent period of development, only to be added in at the last second when the team needed a few more Pokemon to fill out the Spaceworld ’97 roster.* Note as well that Snubbull’s moveset (below) is a pretty generic at the time: it has Sing, Safeguard, Double Edge, and Sweet Kiss, all moves that were given pretty liberally to pink puffball Pokemon at the time and an indication that its moveset wasn't deeply thought out. Thus, even though Snubbull had been around for a long time, it never really had a chance to get an evolution. Given that the team had already used it in promotional materials and it had appeared prominently in “Pikachu’s Vacation” which debuted in July 1998, Snubbull was a must include even after Masuda’s reboot of development. So they probably decided it needed an evolution then, and Granbull was one of the first Pokemon they worked on. Granbull faced the same sort of restrictions on it’s design as Togepi, in that since Snubbull was already released to the public, Granbull had to fit awkwardly around Snubbull’s design, rather than the two of them being made in concert or Snubbull being edited to better match its evolution. Despite this restriction, I think Granbull came out a lot better than Togetic. Granbull obviously builds on Snubbull’s design, but not in an awkward or crowbarred in way like Togetic: Snubbull has big lower teeth, so they grow in giant tusks for Granbull; Snubbull is pink, so Granbull is a shade of purple; Snubbull is grouchy, so Granbull is irate. Though Granbull is a very simple “larger, fiercer version of the original” type evolution, it works quite effectively here. *Strangely, though, even though Snubbull looks like a last-minute addition, it was one of the few Generation II Pokemon without placeholder stats. Though its stats at that point aren’t much different from the placeholder 50s across the board (70 attack and special attack, 65 HP, and 60 defense), you’d think the team wouldn’t have bothered to update these in the limited time they had. My guess is that the team may have considered giving one of the demo trainers a Snubbull, to tease its existence in the demo, but eventually decided against it. Beyond these discussions of Granbull's initial origins, Granbull's a pretty straightforward Pokemon. The team already had its final stat distribution by April, 1999: they had decided it would be a massive, 150 point boost over Snubbull, which included the huge 40 point boost to its attack stat, which was already a respectable 80 under Snubbull. Clearly, the team knew what they wanted Snubbull to grow into and stuck to it. The moveset did change by the final, but nothing interesting got added: the final version of Snubbull and Granbull got Roar and Rage, which fit their personality better than Double Edge, but otherwise the moves are about what you’d expect for a Bulldog Pokemon. The team originally gave Granbull Pokedex entries which stressed its timidity and cautiousness (or, as TCRF puts it, anti-social nature): It's usually unsociable. However, it seems that it can smile in the presence of its trainer. When it picks up an odor that it's never smelled before, it emits a low-pitched growl. In the final, the team went all out and made Granbull and outright coward, which is honestly much cuter and gives a much more vivid sense of its personality: It's actually a scaredy-cat. When attacked, it frantically flails its limbs around in an attempt to fend off foes. Honestly, who wouldn’t want to raise a cutey like that? Finally, there’s only two trainers in the entire final game with a Granbull: a Lass on Route 24, and the rematch with Pokefan Beverly, who only had a Snubbull in her first fight. Once again, this seems like an incredible waste: Granbull feels like a Pokemon a ton of late game Lasses should be using! One more example of Gold and Silver not using their own roster. Anyway, that’s all there is to Granbull: Snubbull needed an evolution, and it got one. Gligar, on the other hand, is much more interesting… ID #X07: GligarThere’s something odd going on with Gligar. We don’t have enough evidence to know exactly what, but there’s something very off about its constituent parts. I honestly don't know if these oddities will lead to anything at all, but I'll show them all off and you can judge it yourself. Before we get there, let’s go over Gligar’s design. Gligar didn’t appear at all in Spaceworld ’97, but it popped up at some time during the reboot. We don’t know exactly when it was first created, and honestly, I wonder if Gligar’s first design showed up before the development reboot, but after the end of the Korean Index. Part of the reason I think that is because we have an early sprite of Gligar that seems to be before even its June 1999 sprites; this could indicate that I should have put it even earlier in Era IV. This early version of the sprite is facing directly towards the camera, and lacks any sort of shading. It's also mirrored directly down its center (except for its tail), suggesting it was mocked up quickly and was quite rough. It doesn’t look at all like a Sugimori or Nishida sprite, and makes me wonder if it was created by one of the new designers. It’s mouth also bears a slight resemblance to Animon’s, which was the Pokemon it replaced, and I can’t help but wonder if that was intentional. This version of Gligar was replaced by June 1999 with a new sprite that kept the same design but was rotated to a 3/4s view, a camera angle much more naturalistic and more typical for Pokemon sprites. Here, we can see that it has a red and grey palette. It’s a much more polished sprite—notice that the designer put some shading on its belly, and there's a sense of perspective now—but it still doesn’t look much like a Pokemon, especially the cute and rounded designs that were common to Generation II. What exactly is this version of Gligar supposed to look like, anyway? I mean, it definitely has a scorpion tail and what kind of looks like pincers on its wings. Not to mention a mean-looking face. Gligar’s early name was “Sasoriga,” and “Sasori” is scorpion in Japanese, so there was definitely scorpion influences in this early design. Some people have theorized that the unused scorpion design from Period 1b in the Korean Index was actually an early design for Gligar, and it’s vaguely possible to see that strange design evolving into this one after it got iterated enough. On the other hand, I really doubt there was any connection. These sprites, for instance, were definitely made by different people; their styles are completely different. Furthermore, I don’t really see any similarities between these two Pokemon except for a common scorpion inspiration. Bulbapedia also suggests a pretty likely origin story. The Gazu Hyakki Yagyou, an illustrated bestiary of Japanese yokai written in 1776 by Toriyama Sekien (kind of ancient ancestor of the idea of a Pokedex) features a brief entry on a yaokai called “Amikiri” with an illustration making it look like a flying crab or lobster. There’s some suggestion that “Amikiri” derives from a corruption of the yokai “Kamikiri,” a better known mischevious Yokai which looks a bit like a kappa and secretly cuts people’s hair. Whatever the case, “Amikiri” seems to be a pretty rare yokai, and it isn’t actually clear they come from any old Japanese legends, but made have just been imagined because Kami sounds a but like "Ami" which is a type of shrimp with pincers. There’s another, more interesting, explanation for Gligar’s design. Doesn’t this early Gligar look like…well, an Alien? In the movie Alien, the eponymous Aliens hatch from eggs into a flying monster with finger-like limbs. It latches onto someone's face (hence their nickname, "facehuggers") and injects new Alien DNA into the body of the host, which eventually leads to an Alien bursting out of the stomach of the aforementioned host. It's an incredibly scary concept that plays on ideas like sexual assault and pregnancy. And I'm not the first to point out that a facehugger has an incredible similarity to the initial design of Gligar. Yeah, that’s really dark. It certainly feels like an inspiration that’s just a bit too adult for a Pokemon game. But let’s sit with it for a second. Do you know the name of the famous horror designer who came up with the concept art for the movie Alien? Oh yeah, that guy’s name is H.R. Giger. What if “Gligar” was just a portmanteau of “Glide” and “Giger”? It certainly matches that early design. That’s the point at which the coincidences get too much for me. As weird as it is, I’m pretty confident that this must absolutely be the origin for, at the very least, Gligar's name. My guess is that maybe one designer created “Sasoriga” from the idea of an amikiri, and then later someone else noticed that the original design looked a lot like a facehugger, and changed the name to be a pun on Giger as an in-joke. With an origin like this, the design might have looked a little too close to Alien. Thus, by Spaceworld 99, Gligar has a completely new sprite (and palette). What’s interesting about this new sprite is that nothing about the body of Gligar has changed all that much. It still has the wings, with the same six prongs—the pincers have been deemphasized on the middle two arms, so that it now just looks more like part of the wing—the same scorpion tail, and a head that looks remarkably similar. However, the entire body, from tip to tail, has had a makeover. Rather than looking aggressive, Gligar is now sticking its tongue out. Rather than a scary bat, Gligar’s entire body now has cartoonish proportions, giving it a silly appearance. Rather than growl, it now smiles. Gligar’s been completely domesticated by its final design. The weirdest thing, and the main issue with Gligar I want to discuss, is its typing and its moveset. Gligar’s unique in that it’s the first Ground/Flying type Pokemon. To this day, it’s basically unique in this typing: the only other Pokemon who share the typing are its evolution, Gliscor, and some weird thing from Gen V called Landorus. There’s no proof, but my guess is that it was originally designed to fill the hole left by Bombushikaa, which was the unique Fire/Water typing. My hypothesis is that Bombushikaa was initially created as part of a sequence in the Korean Index where the team was trying to create unique typings, and I’ve speculated that the strange megaphone-bird next to it in the Index might have originally been an attempt to create a Flying/Ground Pokemon to compete with the “opposites attract” theme of Bombushikaa. If I was right in that case, then Gligar could have been a second attempt at a Pokemon of that typing. Which presents us with a mystery. Have a look at Gligar’s moveset: Uhhhh…is something missing here? Gligar doesn’t learn any Flying moves, nor does it learn any damaging Ground moves (In Generation II, Sand-Attack is technically a Ground move, but many things learn it). Instead, it has one Poison move (Poison Sting), one Dark move (Faint Attack), and then a bunch of Normal moves. Why give Gligar this typing if it learned no moves of its types? Okay, but in Generation I, there were lots of Pokemon like this. Voltorb and Electrode, for instance, couldn’t learn any Electric moves naturally; the Rhyhorn family learned no Rock or Ground moves; the Sandshrew line learned no Ground moves either. All three of these Generation I Pokemon, however, learned a bunch of same-type attacks through TM; in fact, it seems like the Generation I design philosophy was to make most of the strongest moves only available through TM, in the same way that in the early Final Fantasy games, you needed to find or buy the strongest spells. Sandshrew and Rhydon, for instance, both learned Earthquake and Dig, the best (in fact, two out of the only three) Ground moves by TM. So what about Gligar? Gligar learns two more Poison moves (Toxic and Sludge Bomb) a Fighting Move (Rock Smash), one more Dark move (Thief) and even a Bug move (Fury Cutter). But not a single Flying or Ground move! What's going on? As the youtuber Scottsthoughts has made me especially aware, Gligar can’t even learn the TM for Mud Slap, an early game Ground move given by the first gym leader that just about every other Pokemon in Gold and Silver learns. In fact, Gligar is the only Ground type Pokemon out of all 252 existing Pokemon that can’t learn Earthquake or Mudslap. Not to mention its inability to learn the HM for Fly. Okay but maybe Gligar learns Flying or Ground moves from breeding? Well, it learns Counter, Razor Wind (though it's impossible for it to learn this move without trading from Generation I), Metal Claw, and Wing Attack. So yes, it finally learns one single Flying move, if you go through the trouble of breeding a new Gligar. Through a convoluted method, it learns one STAB move. What's weird about that, though, is that it turns out Wing Attack was a late addition to its Egg Moves--it didn’t learn this egg move prior to August 30th, 1999, only a few months before the final. Instead, prior to the end of August, it learned Psybeam and Giga Drain; earlier than that, Psybeam, False Swipe, and Leech Life. Instead of learning Ground and Flying moves, the designers gave it indirect access to a Psychic move, a Grass Move, and a Bug move. What the heck is going on? Unfortunately, though the data’s out there, I don’t have a record of Gligar’s moveset from earlier builds; no one on The Cutting Room Floor has compiled this data and I don't know a thing about coding to check myself. Because Gligar didn’t yet exist in Spaceworld ’97, the only version I have to go off is the final moveset. But I’d kill to take a look at versions of the moveset going back to earlier in 1999. Did it have any Flying and/or Ground moves in earlier iterations? Did they, for some reason, remove these moves from Gligar’s roster? Or did it have a bunch of moves from a different typing originally, as if the team had actually planned for Gligar to be some other typing entirely? What we do know is that Gligar was a Ground/Flying Pokemon from the earliest data we have: by its first appearance in April 1999, it already had these types. So it's very unlikely that Gligar was original some other typing and it got changed to Gound/Flying late in the day. It could be the case that, late in development, the team decided to change Gligar to a new typing (Say Poison/Dark or Poison/Flying) and got a new moveset ready, but forgot to update its actual typing. The team was under lots of pressure towards the end of development, and the people who were working on movesets could have been working at a different speed than everyone else. Likewise, it’s absolutely possible that there was a miscommunication between the person in charge of Gligar’s moveset and the design team: one group could have planned for Gligar to be Ground/Flying, and the other could have thought from Gligar’s palette and design that it was obviously a Poison-type or Poison/Dark—hence why it learns Sludge Bomb by TM and starts with Poison Sting and Faint Attack. UPDATE: GoldS_TCRF, in the comments, checked this moveset out for me, and it turns out that Gligar's moveset was identical to the final starting all the way back in April 1999, the first snapshot of the rebooted development we have! So unless Gligar's typing and/or moveset predates April, it seems that it was always designed to be a Ground/Flying type without Ground or Flying moves. Again, I'm not sure what to make of this. Was this intentional? Did the team just forget about Gligar and overlook this glaring problem? Did Gligar just get accidentally lost in the shuffle? I have no idea. If some mix-up like that happened, it may have been hard to catch. Gligar is a version exclusive to Pokemon Gold, and it's only found late in the game on the last route of the Johto region, a route that many players ignore entirely because it just connects the final town to the beginning town. Thus, it would have been easy for the team to overlook it during playtesting and just not notice the strangeness of its moveset. Or, as odd as it sounds, maybe Game Freak intended Gligar to have such a bad and unusable moveset. After all, in Pokemon Stadium 2 (released the same day as Pokemon Crystal), one of the rewards you could get from beating the game was a Gligar that had Earthquake already learned, even though it could not normally learn it. Another prize was a Gligar that knew Wing Attack. Depending on how you view it, the strangeness of gating these two moves behind Pokemon Stadium 2 could mean two things. First, it could be the team issuing a corrective after they made a mistake with Gligar: by giving it access to Earthquake and Wing Attack here, they could be implicitly admitting that they overlooked Gligar in the original release. Of course, remember that Crystal version released the same day as Stadium II, and that Crystal had already tweaked a few Pokemon’s movesets—notably, for some reason the developers thought that Sneasel was substantially incomplete without learning Metal Claw. If the team thought Gligar’s inability to get Earthquake was a mistake, why didn’t they just add Earthquake to its moveset in Crystal? But what if leaving off these moves was the point? A few commenters have argued that the developers saw Pokemon Stadium II as the true endgame of Generation II, and if Stadium was designed way in this fashion, maybe an Earthquake-using Gligar was an incentive to play Pokemon Stadium? It's possible that the designers left Gligar without any STAB moves on purpose, so that a decent Gligar could be the reward for beating your rival on Pokemon Stadium. Gligar, thus, would be a Pokemon designed to sell the functionality of Pokemon Stadium II. Personally, I find the design philosophy of “make something unusable on purpose, just to give a competent version of it as a reward later” to be completely baffling, but who knows what Game Freak was thinking here. I wish we had more information here to know exactly what they were going for with Gligar. I really like the theory that the moveset was mistakenly designed with a Poison/Dark Pokemon in mind, but I admit that with what we have, there’s very little evidence for it. On a different subject, since I mentioned Sneasel, it's worth lightly discussing the relationship between Gligar and Sneasel. In Generation II, there was no relationship: they were simply two new Pokemon introduced. The two Pokemon have little in common in Generation II—they both have odd typings and the same incredibly low Special Attack stat (35), but that’s about it. In fact, it appears that the version exclusive counterpart of Gligar was Delibird (potentially Mantine, it’s a bit unclear) in Gold and Silver. I’ve always felt a little bit bad for the kids who got Silver and got the useless Santa Claus penguin instead of this guy, though now that I’m analyzing Gligar I’m coming to the realization that it isn’t much more useful than Delibird in a typical game. However, it appears that in Generation IV, the designers decided that Sneasel and Gligar should be counterparts. Both Gligar and Sneasel got evolutions in Pearl and Diamon: Gligar got Gliscor, and Sneasel got Weavile. They were both bad Pokemon held back by terrible stats, so these additional evolutions made a lot of sense to help them gain new life in a new Generation. To drive home how similar they were, the designers gave them almost the exact same evolutionary method: Gligar needs to level up at night while holding a Razor Fang to evolve, while Sneasel needs to be holding a Razor Fang while doing the same. Given that Gligar and Sneasel have some of the most memorable designs in Generation II, I understand why Game Freak would have retroactively made them counterparts. Both are sneaking, mischievous Pokemon with something to prove. (Image by LucasPireArts!) By this point, a pattern should be emerging. Pokemon designed or refined (such as Sneasel or Dunsparce) during Era IV are often pretty unusable in the final game, due to either weird stats, horrible movesets, or being available exclusively in Kanto. Gligar’s a great example of this, but Yanma, Slugma, Pineco, and Shuckle are all upcoming. Gligar’s a cool design; it’s a shame the team made it so unusable, by choice or by mistake. ID #X08: PiloswineThe next Pokemon, Piloswine, had completely finished sprites by June 1999 and has drafted stats as early as April 1999; both lead me to confidently place it here as one of the earlier Pokemon of Era IV. While there’s much more to say about Piloswine’s evolutions (that’s right, plural), there’s still a bunch that’s interesting to cover with the OG Ice Pig Going to be honest here, I love Piloswine’s dopey look. It’s one of my absolute favorites designs, to me a bit like the Totoro of the Pokeverse. On the other hand, my wife thinks it looks like a highland cow and is much less positive about the design. Piloswine, as you may have guessed, was based on a pig, or on a boar. We know for certain it was designed by Shigeki Morimoto, who was probably busy working on other games for most of Eras I, II, and III; in fact, this might be one of the first Pokemon he worked on for Gold and Silver, depending on if his other work completely precluded designing Pokemon for the Korean Index era. He explained in an interview that he liked the idea of having a Pokemon for each of the animals in the Chinese Zodiac; Piloswine, obviously is for the year of the boar. Which was 1995, not 1999, so obviously Morimoto was a little late. We know that Piloswine was created first, before Swinub, and that in fact, Swinub took a long time for the team to design. However, it's clear from the very beginning that Morimoto had been asked to make Piloswine part of an evolutionary family: two more slots in the Pokedex, both before and after Piloswine, had already been carved out and were using placeholder graphics by June of 1999. There’s a few theories that link Piloswine back to an earlier moment in development. Some people think that the boar Pokemon found at the beginning of the Korean Index (ID 319) could have been a very early precursor to Piloswine, but I doubt it. In addition to having none of Piloswine’s distinctive traits—they both have tusks I guess—the most distinctive part of 319 were its antlers, which have no continuity with Piloswine. In addition, as I said, Morimoto was absent, working on other projects for most of this early part of development, so it’s unlikely he designed this boar (I may have said differently earlier on in the Cryptodex, but since then, I’ve talked to people with more knowledge of Game Freak’s products than I had). More plausible is that there was a tenuous connection between the Wolfman and Warwolf line from Spaceworld ’97, which were replaced very soon after the 1999 reboot. They seem to have survived at least for a small time into the reboot, given that post-SW'97 updated versions of their sprites were being used as placeholder graphics as late as June 1999. But they were cut by at least April 1999, for some reason. My guess was that they were counterparts of the Houndour line and the team liked Houndour more, but others have suggested that their design, which depicts a Pokemon hiding in the pelt of another killed Pokemon, might have been too dark for the games. In any case, there’s a decent chance that at least Wolfman was repurposed into Snorunt by the next generation. But when Wolfman and Warwolf were cut from the lineup, it left a huge Ice-type hole in the entire roster. Sure, the team had Delibird, Smoochum, and Sneasel, but Sneasel naturally learned no Ice moves (and was conceptually more of a Dark-type), Smoochum doesn’t exist except if you get deep into the breeding system, and while Delibird hadn’t yet been made into the useless gimmick Pokemon it became, it still wasn’t, shall we say, good. They knew by this point that the game would have an Ice-type gym leader who needed a new Pokemon to show off the type, and they knew there’d be a Dragon-type gym leader and so the player would need something besides Dewgong to counter it with. So I’m sure when Warwolf and Wolfman got the chopping block, an immediate concern of Masuda’s was to make sure an Ice-type Pokemon family replaced them. In this case, three Pokedex slots were set to the side for a new design, and Morimoto was likely told to design a new Ice-type three-stage evolutionary line. Morimoto started with Piloswine, though by June 1999, Piloswine seems to be the only one of the three complete. The slot before Piloswine, #221, has an empty placeholder sprite, while the one after it, #223, uses Warwolf’s sprite, but was likely also being used as a placeholder given the absence of Wolfman. Importantly, all three slots already have stats penciled in by April 1999, the Ice/Ground type, and evolution data that says they evolve into each other. They had placeholder movesets: each of the three forms just learned Powder Snow and nothing else until August, suggesting that initially the line was defined only by the need for an Ice-type evolutionary line. Beyond just the missing movesets, these Pokemon were substantially complete if you ignored their missing sprites. The stats of the Piloswine family, by the way, are a bit interesting: (Data and formatting taken from The Cutting Room Floor) Notably, all three have higher defense and special defense in the early stages of design, and much lower Speed, Attack, and Special Attack. This would have made them slow tanks, like Slowbro, but honestly, with those stat distributions the Piloswine family would have been a dud. As it is, the Ice/Ground type combination is weak to five different types—Fighting, Steel, Fire, Water, and Grass; with that incredibly low Speed, the Piloswines would be taking huge hits before they were even able to move, regardless of how high their defensive stats were. At least when the stats were modified to make them faster and with higher attack, they’d have a better chance of sweeping the opponent after they took a hit. My guess is these stats were improved so that Piloswine could serve as a decent counter to Claire, especially as Swinub only appears in the Ice Path, which just so happens to be right before her gym. But let's move on. We've still got two more pigs to get through... ID #X09: SwinubBy June 1999, Morimoto had created his Piloswine design. The team was clearly happy with it, because it existed unchanged until the final game; even the Crystal sprites only had minor changes to the palette and back sprite. However, Morimoto had no clue what the other two stages of the Piloswine family would look like, and they remained without sprites, possible as last as August 1999. Eventually, one of those Pokemon became Swinub. The other was lost to time. What’s especially interesting about the Piloswine line was that we have evidence that suggests that Morimoto wasn’t sure what stage Piloswine would actually be in the family. In fact, Morimoto drew multiple iterations of Piloswine as a first stage, a second stage, and a third stage Pokemon, probably assuming that they’d slot it in at whatever level it worked best, and then design the other two evolutions later. He very likely never meant for all three of these sprites to be used, but it does mean we have this funny proto-Piloswine family lineup of a smol Piloswine, a regular one, and an XXL Piloswine: They're all adorable and I love them so much. Smol Piloswine is like a tubby little piglet, and XXL Piloswine is roaring its heart out. I know it wouldn't have made sense to have all three in the game. But think about the possibilities! One thing that I really like here is that if you look close, there’s been a number of subtle changes to the base Piloswine design. The third-stage sprite is the most obvious: the tusks are much larger, and it’s now roaring angrily with its mouth agape. But the smaller first stage version is interesting too: though its feet aren’t much different from Piloswine’s, its smaller stature and nose makes them more prominent on the sprite and really makes it stand out as a four legged creature, whereas vanilla-Piloswine looks more like a mass of fur. Notably, I see a lot of similarities between this sprite and the eventual sprite Morimoto settled on for Swinub, but the biggest difference is he did away completely with the feet, making Swinub almost look like a pig-caterpillar. Again, this is a really good example of just how development of Pokemon designs was different under Masuda than it had been under Tajiri. By this stage, Morimoto was designing a Pokemon to fit the Ice-type needs of the game; his Piloswine design wasn’t being forced into the roster, like earlier designed Pokemon, but instead he’s actively redesigning it to better match the design needs of the game. In addition, the team seemed unsure about whether the Piloswine family even deserved three slots in the Pokedex. While we know that the third evolution was the one eventually cut—was this potentially because Morimoto couldn’t come up with a third sprite he was happy with?—for about two weeks in August 1999, the team almost cut the first form, which eventually became Swinub. Notice how Swinub gets reverted to placeholder stats for two weeks if I bring up the stat charts again: From July 30th to August 14th, Swinub’s stats were completely wiped, though it didn’t quite lose it’s typing. After the fourteenth, the team realized they had made a mistake, and reinstated the stats; at the same time, they gave Swinub and Piloswine movesets very close to their final ones. Then, by Spaceworld 1999 (compiled three days later, on August 17th), Swinub appears in the Pokedex, still without a name (It was Mitei 09, or “Pending 09”) but with its final sprites already complete. Notably, it has the same sprites in both Gold and Silver, even in the final. This heavily suggests that the team was too rushed by this stage to come up with two separate designs for Swinub. Swinub's incomplete in a few more subtle ways in Spaceworld '99 as well: for instance, it still has the generic Rhydon menu sprite, showing that they hadn’t yet finished transitioning the Pokedex slot over to Swinub. Meanwhile, Slugma’s sprite had replaced the third evolution of Piloswine by this point, demonstrating that the team had conclusively decided on Swinub only by this late point in development. From our standpoint in the present, it seems odd that Swinub was ever on the chopping block. After all, Piloswine just doesn’t look like a first form, even in the mini-version that Morimoto created. But clearly that just wasn’t the case for the designers. Maybe this is a case of hindsight being 20/20, because clearly the imposing bulk of Piloswine didn’t immediately convince Game Freak that this guy should be a fully evolved dude. Saying that, by the time they had created Swinub’s final sprites, it was much more clear. Swinub, the sleepy, unimpressive pig-caterpillar was a perfect design for the first stage of this family, immediately drawing a similarity with Piloswine while also being cute and adorable in its own right. They took awhile to figure this out, but it was definitely a winner that helped focus the designers while they thought through Piloswine’s place in Gold and Silver. Adorable. ID #010: Piloswine-3But if Swinub was completed, and the line was scaled back to two evolutions, what about the last boar? What about the one that got away? Of all the unused Pokemon in Era IV, Piloswine-3 is the one that was the closest to completion before it was deleted. Most of the others—only Kageboozu is an exception, and that’s a weird case—do not have any surviving sprites we can access, and I’m doubtful that most of them ever had sprites at all. While the XXL Piloswine sprite was probably not intended to be a final sprite, at this guy has something. In addition, Piloswine-3 got a fully established moveset and TM list (though it’s moveset is just Piloswine’s, of course, but later moves are learned at higher levels) and it got its own set of stats. If they team did create a unique Piloswine-3 sprite at any point that we just don't have, then Piloswine-3 was completely done before Game Freak decided to delete it. Like I said above, in June 1999, when only Vanilla-Piloswine’s sprite was finished, Piloswine-3 had stats and a typing, but was using Warwolf’s sprites. It’s unclear if the sprite above was ever assigned to it, or if that sprite was just a sketch Morimoto did to get ideas for what Piloswine-3 might have looked like. Personally, I love the ferocious roar that Piloswine-3 is diplaying, and I hope that would have been carried over into whatever unique sprite they gave it. Piloswine-3 existed until at least the middle of July 1999 before the team started to erase it. By Spaceworld ’99, there are substantial portions of Piloswine-3 that are still unchanged, but it’s clear the team has already started transitioning away from Piloswine-3. By SW'99, early Slugma sprites had replaced whatever was there for Piloswine-3. However, almost everything else still screams Piloswine. The typing is still Ice/Ground, it still retained the same moveset, and Piloswine still evolved into Piloswine-3 at this stage (a strange transition, from a furry boar into a puddle of slime). In addition, the faux-Slugma still has a footprint that looks like a larger version of Piloswine’s, and it has a unique cry that resembles that of Piloswine’s. I think it’s very likely that if we just had a slightly earlier build of the game—maybe just three days earlier—we’d see some sort of unused sprite for Piloswine-3 that we don’t have access to now. So what happened to it? Why get rid of a Pokemon that was so substantially complete? I think the answer is just that the team decided three stages to this evolutionary family were taking up too much space. They’d already created Magcargo by this point, and I think there was a realization setting in that Magcargo deserved an earlier form. Piloswine-3 was the easiest thing to cut, because the team didn’t lose a lot of Pokemon diversity in the roster by cutting it. While, like I said above, they considered cutting the Swinub instead, Piloswine-3 obviously made more sense to them. Probably because, echoing above, Piloswine just didn’t look like a first stage Pokemon on its own, but it certainly could play the role of a final stage. After that, Slugma was moved to earlier in the Pokedex, in order to be next to Magcargo for the final. Corsola was moved into slot #222, but by this point it no longer had anything in common with its fluffy boar predecessor beyond taking its old Pokedex slot. When they cut Piloswine-3, Game Freak also significantly raised Piloswine’s base stat total, to make it more formidable as a final stage evolution. While Piloswine had a 380 Base Stat Total as a second stage—pretty low and unimpressive, honestly—it gets buffed to 450 after Piloswine-3 was deleted, making it more attack oriented and giving it some speed. As welcome as this is though, this feels a bit like it was robbed, given that Piloswine-3 had a 500 BST, which was substantially better. Swinub was rebalanced after Piloswine-3 was removed, mostly to make it less defensive and speedier, but it didn’t get a similar stat total buff. In fact, it went from 260 BST to 250, making it quite weak, even for a first stage Pokemon. Given how much I love Piloswine, I’m honestly crushed that Piloswine-3 made it so close to the final but was cut at the last minute. A third Piloswine evolution would be amazing. Now, obviously, the team probably agreed with me, because in Generation IV, they brought the idea back. Behold: Mamoswine. On the bright side, Mamoswine is an absolute beast. If Piloswine-3 was going to have 500 BST, Mamoswine exceeds it, at 530 BST. It’s hard to compare their stats, since we know Piloswine and Swinub were made more less defense oriented and more aggressive after Piloswine-3 was removed. Presumably, a final version of Piloswine-3 would have lost some defense stats and gained some Attack and Speed. Still, comparing the two shows just how much better Mamoswine is: On the negative side, well, Mamoswine is a Generation IV design. And I know this is just personal preference, but man, I do not like the choices the designers made in creating new evolutions for Generation IV. Swinub and Piloswine are adorable because they’re giant balls of fluff: you can barely make out limbs of any sort, nor can you read their facial expressions. So of course they ruined all that in Mamoswine’s design. Why give it weird creepy eyes? Why make it less a ball of fur?
Interestingly, Mamoswine was made a counterpart of Tangrowth and Yanmega in Generation IV: all three of them only evolve if they know Ancient Power, and all three take on traits of pre-historic species when they evolve. Tangela becomes a cave-man, Yanma becomes a giant prehistoric insect, and Mamoswine becomes a Wooly Mammoth. Alright, this is vaguely neat that the designers came up with theme. I just wish it didn’t completely ruin Piloswine’s look in the process. I doubt Piloswine-3 ever shared any aesthetic similarities with Mamoswine. First of all, like I always say, these were developed about six years apart, and probably by different designers. Any similarities would be odd coincidences. On top of that, Mamoswine looks the way it does because of it’s inclusion in the Ancient Power trio; since that didn’t exist when Piloswine-3 was being invented, I doubt the Wooly Mammoth inspiration was present in its design. Thirdly, though its tenuous, the Piloswine-3 sprite we do have shows it with its mouth bellowing a roar. While that could easily have changed, it hints at an entirely different personality than Mamoswine: one that was meaner and more aggressive. Unfortunately, Slugma got created, and Mamoswine took away any chances for Piloswine to have an evolution befitting its greatness. Piloswine-3: What could we have had? Era IV OverviewAnd now we’ve made it. To the final section of Pokemon Gold and Silver’s development. Pokemon Gold was originally slated to be released in 1998; however, that was not to be. On the one hand, the infamous Porygon-seizure-incident put Pokemon in the spotlight in the worst way; it seemed, for awhile, as if the outrage over that event may have led to the death of the franchise before it started. But there were other issues as well. The success of the first game in the series had led to demand for stop-gap games to keep the public interest in the franchise high. Developers were tasked to work on Pokemon Yellow, for instance, and the development team quickly repurposed most of the sprites they had made for Gold and Silver into special-edition sprites for the release of Pokemon Blue, a Japanese exclusive deluxe release of the first games. These projects likely siphoned away many of the developers from Gold and slowed progress for those games. However, the biggest reason Gold was delayed was probably due to internal issues. In interviews, the development team has said that the first approach to this game—showcased in the SW97—had major problems that the team was finding difficult to address. The main issue the designers pointed to was the map: at the time, Spaceworld ’97 was designed to encompass the entirety of Japan, with the original setting of the first games, Kanto, relegated to one small town. This map, apparently, didn’t work; it was trying to do too much. Due to its design, and possibly to cartridge limitations, there were too many cities that didn’t contribute, and the cities were so close together that there wasn’t room for interesting routes in between. While the team doesn’t mention anything about problems with the Pokedex roster, I find it likely that they were also having trouble with it. My number one and two guesses are that there were too many single stage new ‘mons, and that the babies didn’t add enough interesting variety. Whether or not I'm right, something wasn’t working. And so, at some point, Satoshi Tajiri, the creator of Pokemon and the director of Generation I, stepped down as director of Generation II. There was some sort of gap (probably most of 1998) in the development of Gold Silver around this time; it’s unclear if that gap was while the team decided on a new director, or if the gap was motivated by letting the team work on Yellow, Blue, and a few other projects. But sometimes in late 1998 or early 1999, =Junichi Masuda took over as director. Masuda's a big name in Pokemon: he has since worked on many of the later games in this role. Masuda seems to have immediately made massive changes to the games, to the point that Era IV—which starts in 1999 and covers the entirety of Masuda’s tenure as director—is almost a completely new game that happened to draw from some existing assets from Tajiri’s time as director. First, Masuda completely redesigned the map. The exact changes to the map are beyond the scope of the Cryptodex, but suffice to say, the plotline and locations of the game after the reboot are completely new. The Pokedex roster, on the other, is harder to sum up in terms of how it evolved, and one of the ongoing questions we’ll be looking at in Era IV is what Masuda’s overarching philosophy in regards to the roster was. At this time, I can’t see very many clear patterns. I think Masuda was less interested in featuring Pokemon that linked to the original generations that Tajiri was, explaining the culling of the babies and of the numerous third-stage Pokemon that were dropped after Spaceworld ’97. I think, generally, Masuda was interested in creating cute Pokemon, and I think, in general, his goal was to make the games more friendly to international audiences by toning down some of the more oblique references to Japanese culture in the Spaceworld ’97 roster. I do think that the massive changes to the roster, at a point in development where Masuda had less than a year to deliver a completed product, were probably an overambitious mistake. We’ve already seen how so many of the newly introduced Gold and Silver Pokemon are underused or almost completely absent from the final games, and that becomes even more pronounced in Era IV. It took Generation I six years to build up its roster and properly balance all of them against each other, and even then there are a lot of Gen I Pokemon that are bad or absent from those games. And unfortunately, a lot of the elements added or significantly changed during Masuda's time in charge are the most rushed parts of the final games. If Masuda decided to throw out a good portion of the roster that had already been worked on, then that meant he was tasked with creating around forty new Pokemon in less than a year, in addition to balancing the old ones that he’d inherited. My speculation is that this is why so many Johto Pokemon feel badly balanced—better to keep Yanma and Ledian weak and unusable rather than accidentally break the game by making their stats too high. It's probably also why so many of the new additions to the roster (especially things like Slugma, Larvitar, and Misdreavus that were only created after the hiatus) are so hard to find or locked resigned to the post-game sections of the story. The less these new Pokemon impacted the difficulty curve of the main game, the better, because the team could tweak and test them all the way to the very last moment. If Morty never bothered with a Misdreavus, there's no chance that giving it 20 more Speed might accidentally make the fourth gym unbeatable. There’s also an important technological change in this era which affects our ability to analyze it. The Korean Index was last updated on May 6th, 1998. We know, however, that the roster wasn’t worked on much after Spaceworld ’97, because almost everything in the Korean Index matches the SW’97 demo. There were probably a few changes to the Korean Index between November 1997 (the SW’97 demo) and May 6th: Bombushikaa’s and Ho-oh’s sprite were updated, Stantler’s was probably updated or overwrote something, Elekid’s was at the very least touched up, and Cyndaquil could have overwrote something already there. But that’s it. So why did the designers stop using the Korean Index? It appears that the team found a way to more directly program in the sprites into the game. As far as we can tell, up until May 1998, the team’s process seems to have been to draw the sprites on a scratchpad, decide on the design they liked, and then manually update an Index number in the Korean Index with the sprite they wanted to use. When the games built from the source code, the instructions were to pull these sprites directly from the Korean Index into the data of the game. After May 1998, however, they seemed to have found a more efficient way to update the sprites. It looks like, instead of manually dragging their sprites to the Korean Index, they told the game to build directly from the scratchpads that they were using to work on the sprites: the game’s programming told it to take whatever was in the top left of the scratchpad, allowing them to leave drafts and experiments in the space to the right and below the sprite. As far as I can tell, this would have been much easier for the team, because every time they updated a sprite on the scratchpads, it would automatically update itself in the game, so long at it was built again from the source code. Which means that the team no longer needed the intermediate step of the Korean Index. This leaves us with several disadvantages. First, we no longer have a hint at the creation order of any of the Pokemon in Era IV. We can reconstruct a vague order in which these Pokemon were created by checking when they first got stats, when they first got sprites, etc, but it’s a bit of a tragedy, because this only gets an approximate ordering. That means that any patterns in the different Periods of development in this era are obscured, because we don’t know which Pokemon were created in batches. I’ll do my best to explain why Pokemon are where they are in my listing, but know that in Era IV, the ordering inside each Period is more of an educated guess. The other downside to losing the Korean Index is that we no longer have any sprites for Pokemon which were created but then abandoned before June 1999. The Korean Index was full of oddball sprites that were never present in any of the builds of the games we have, but we just don't have the same insight into Era IV. We know from the stats of certain Pokemon and by traces left in the data of a couple of Pokemon that were considered but ultimately rejected; however, because we don’t have the Korean Index, we usually don’t have sprites for these guys (with a few exceptions). So we’ll be talking about a few Pokemon in this era that are more or less mysteries, without even a fun visual to get our imagination going. One other technical change is that, after Masuda took over, the development shifted to making Gold and Silver not for the Super Gameboy, but for the new Gameboy Color. As a result, the sprites designed for this era of development are designed with a richer palette in mind and completely different shading. On the upside, these sprites are a lot more expressive, and give the games a more distinctive look! They also show us the exact moment when a Spaceworld ’97 Pokemon was updated (because the new sprites would now be shaded for the Gameboy Color, not the Super Gameboy), and for how long it just stuck in the roster as dead weight before the team found a replacement for it. On the downside, however, these new sprites are much less distinctive in how they’re shaded, and because of that, it’s much harder to determine which sprite artist drew them in the first place. We have a lot of other ways to determine the author of a ‘mon, but guessing at the authorship of a sprite is much less likely to work in this stage of development. Alright, that was a lot. With that in mind, let me outline the ways Era IV will work a little differently from the previous three eras on the Cryptodex:
Period 4a: Development Restarted Like I said above, Period 4a is going to cover the first year of Masuda’s time as development lead, from approximately May 1998 to just before the collection we have dating to June 1999. That isn’t to say that development was active for this entire year; I expect that the team was still on hiatus for most of 1998. It also seems like there was a period of development post-Korean Index (so after May 1998) where the team was still iterating the SW'97 designs: the scratchpad files we have have examples of Kotora, Hoppip, Akueria, and Borubea that all seem updated from their SW97 designs. So there might have been a chunk of 1998 that was still under Tajiri, before Masuda took over as director. At some point though, there seems to be a sharp break. My guess is that Masuda became director late 1998 or January 1999 and started to drastically reshape the game. Whenever that hand off actually was, by the time of our first snapshot of development—April 3rd, 1998, which gives us a look at the stats of each Pokemon—the game had already been substantially changed. By April 1999, there have been a bunch of changes to the line up, mostly replacing the baby Pokemon from Tajiri’s time in the lead. In total, 16 Pokemon had already been completely deleted from the roster by April 1999: Almost all the babies had been deleted. Mikon, Monja, Gyopin (though Furret shares its palette after June 1999, suggesting it may have still existed until recently), Para, Koonya, Puchikoon, Betobebii, Pudi, and Baririina were all gone. In addition, Taaban, Bombushikaa, Shibirefugu, Jaranra, Riifi, Animon, Manbo1, and Ikari left no traces in the updated roster. In addition to these 16, there were eight more Pokemon from earlier stages of development whose's names, data, and types had been deleted, but whose sprites still remained in some form. It had probably been decided to deleted these Pokemon by April 3rd, but that decision may have been made later than the Pokemon above. Alternatively, some of these may have been deleted at the same time as those above, but their sprites survived purely because the designers hadn't yet figured out what they wanted to overwrite these sprites with. For instance, Gurotesu exists as a placeholder sprite, and its shading was touched up, suggesting that it survived at least for a small amount of time after the reboot. Likewise, Berunrun's sprite are used in as placeholders in a slot which eventually became Dunsparce's. Rinrin's backsprite was used as a placeholder for Larvitar's backsprite, and Nameru's back was used as a placeholder backsprite for Tyranitar. Likewise, Wolfman's sprite was being used as a placeholder for what became Larvitar, and its evolution, Warwolf, seems to have been used as a placeholder for a third Piloswine evolution that was later cut. Hinaazu had already lost its name, but existed at least as a palette and backsprite for Azumarill. There’s not a lot of clear patterns to the Pokemon added during this period from what I can see initially (and I’m very interested in whether readers see one that I’m missing!). One thing we do see is that it seems that Masuda’s period of development started by first culling all the Pokemon from Spaceworld ’97 that they didn’t want, and then allocating that newly freed space in the roster to new two-stage or three stage lines. So, for instance, the three slots that became the Piloswine line (one was later cut) were probably allocated as “a three stage ice-type Pokemon” long before Piloswine had been designed; it was probably created to fill those slots, not the other way around. Likewise, most of the other new Pokemon seem to have been designed to fit gaps created by the earlier culling. Very likely, the team didn't create more Pokemon in this era and then delete Pokemon to find room for them; instead, they deleted first, and created new ones after. From the patterns we see, it seems as though this approach to creating new Pokemon during this period was quite different that the time during Tajiri’s directorship. During the earlier stages of design, the Pokemon developers seem to just brainstorm a ton of new designs and fit them into the new line-up; they would then create evolutionary relatives for the most popular of these designs. However, here it seems more like the team was filling holes: parts of the roster already seemed earmarked for a two-stage line before the first Pokemon was even created. While, obviously, I’m sure that was done to some extent in the previous parts of development we’ve discussed—especially Era III—it appears that the default beginning point in the creation of Pokemon in 1999 was to start with a hole to fill, and to design an appropriate Pokemon to fill that hole. In the case of Period 4a, it’s clear that one of Masuda’s goals was make Generation II feel more distinct by adding two and three-stage lines wholly unconnected to Generation I. Given that most of the Pokemon deleted to make room in the roster were evolutions of Gen I Pokemon, it seems like an intentional choice to replace them with Swinub, Yanma, Teddiursa, Shuckle, and the like. The team did delete existing unique lines from earlier in development—for instance, getting rid of the Manbo1/Ikari/Gorutesu line and the Wolfman/Warwolf line—those were probably already lines that had serious problems with them and needed to be deleted or reworked (The Ikari line, for instance, was stitched together from three unrelated Pokemon, and the Warwolf line was already created in competition with the Houndour line in the first place). But regardless, the intent here seems to be to create a more diverse set of stars for the new roster. It's easy to extrapolate this too far, but this could also suggest that this stage of development was very focused on creating Pokemon to fill a gameplay purpose. If Masuda started by allocating certain slots in the roster as “ice-type three stage line” or “two-stage early game Pokemon” than the designers might have been designing to a role. Of the Pokemon in Era IV, those in 4a are the ones most likely to appear commonly in the final game, which means there was not only probably time to find them a place in Johto, but also that they were probably made with a place to fit in mind. ID #X01: Heracross (Before June 1999 sprite coloration is hypothetical, and was colorized for me by OrangeFrench!) It’s very likely that Heracross was the first brand new Pokemon to be created in the new era of Pokemon Gold and Silver's development. First of all, Heracross is one of only two Pokemon from Era IV for which we have a pre-June 1999 sprite; we’ll talk about it in detail below, but the tone of the sprite is so different from the rest of Era IV I almost wonder if it was created before Masuda even became the director. In addition to this very early sprite, Heracross already had stats, a sprite, and typing by April 1999; though the sprite needed polish and its typing changed before the final, it was a Pokemon that was pretty much finished by the time of our first timestamp (April 1999) of the rebooted development. Thus, I’m pretty confident putting it at the beginning of Period 4a. Tajiri’s stated in interviews that he was fascinated by bug catching; the beetles he collected while hunting through the forests as a kid inspired the idea of hunting for Pokemon. While we don't exactly know his relationship to the practice, it's also very popular in Japan to watch beetles fight against each other, and it's likely the sport of "beetle fighting" inspired Pokemon battles. Pinsir was likely created out of the inspiration of a fighting stag beetle, while Heracross was designed after a Rhinoceros beetle, which could also be used to fight. Fighting beetles perfectly play tribute to the initial inspirations for Pokemon. But it’s a bit weird to do it twice. Why do we need both Heracross and Pinsir? Back in 2018, when the only data we had was of Spaceworld ’97, a leading theory was that Plux, the Pinsir evolution found in Spaceworld ’97, was an early design for Heracross. According to this theory, Heracross was later separated from Pinsir to become its own Pokemon and given the Fighting-type to differentiate it. And to this day, there are a lot of people who still buy into some form of that theory. I mean, look at it: both Pokemon are designed around fighting beetles: the stag beetle in Pinsir’s case and the Rhinoceros beetle in Heracross’s. Both of them have similar moveset, both learning fighting moves, and both even have a similar pose. The two fit the same niche very well, raising the question of why they both exist at all. Okay, one looks like a Jason-type serial killer and the other one's your buddy, but other than that, same niche, certainly. Unfortunately, once the Gigaleak of 2020 provided us with more data on Gold and Silver’s development, we now know that there never was an evolutionary link between Pinsir and Heracross. In fact, for at least five months, Heracross and Plux coexisted with each other in the same roster. What seems to be much more likely is that Heracross was created to be part of a bug trio alongside Plux and Scizor. Originally, none of them had a second typing—Plux and Scizor became Bug/Steel later in development, and Heracross hadn’t yet gained its fighting type—and they appeared one after another in the early Pokedex: Scizor’s 212, Plux was 213, and Heracross was 214. Together, all of them formed something of a "fearsome bug trio": Furthermore, though their stats would be tweaked--and in Scizor's case almost completely redistributed--as of April 1999, their stats were all remarkably similar: Let's quickly get the most obvious thing out of the way: all three are absolute beasts and could easily dominate any Pokemon of their levels. But also note the similarities: they all have the same stat total, at 540, and they each have high attack: 130, 130, and 125. Plux and Scizor have one more stat they excel in—Plux has defense, Scizor has Speed—while Heracross has two stats its also good in—HP and DEF—and trades that with numbers radically lower in the rest of its stats. Of the three, Plux and Heracross’s stats the closest, but Heracross is much more one-sided towards physical stats, while Plux was a bit more of an all-rounder. I guess, conceivably, you could argue that that Heracross’s stats were just updated versions of Plux’s, and an attempt to make a new beetle that filled its role, but I feel there’s enough variation to demonstrate that the Pokemon team were trying to make all three feel similar enough to be a trio but different enough to feel distinct. I think the team liked the idea of making the Bug type more formidable in Generation II, and the reboot had already inherited Scizor and Plux, which were already contenders. I’m not sure exactly why the idea was to make a trio of terrifying bugs when Scizor and Plux already existed; Heracross feels very unnecessary as an addition, and I’m at a loss for why the team designed it. But again, like I’ve noted before, they continued to work on Plux all the way until a few months from the final. It even got a Pokedex entry. The evidence heavily suggests that Heracross was not there to replace it in any fashion. Maybe someone on the team liked to idea of two different beetles fighting it out? Of course, eventually someone must have realized that Heracross and Plux were just too similar. There were a number of attempts to differentiate them: the team tweaked their stats to be more different, then they gave Plux the Steel type and Heracross the Fighting type. But it seemingly was not enough to justify both, and Plux was eventually discarded. Another problem undoubtedly had to do with Scizor. Scizor was originally differentiated from the other two by being much faster than them, but the addition of the Steel type put a wrench in that plan. The team had decided that Steel-type Pokemon were supposed to be slow, which means that Scizor’s stats were modified such that it was even slower than Scyther. This meant that not only were Plux and Heracross too similar, but the new rules about Steel-types made Scizor's stats even closer to the other two. By Spaceworld ’99, all three of the bugs were very, very similar, and something needed to give. Eventually, the team decided to replace Plux with Forretress. This accomplished a few goals all at once. First, it replaced a Pokemon related to a Generation I ‘mon with one related to a new Generation II ‘mon, Pineco. Not only did this make Pineco a bit more useful and interesting (though not really enough, which we’ll discuss when we get there), but it also served the goal of giving Generation II an identity less tied to the first games. Furthermore, Plux’s design is too aggressive to drastically change its stats to fit into a new defensive niche. By switching to Forretress, the team could preserve the Bug/Steel typing while also creating a more defensively minded bug. Cute! They even let Forretress keep Plux's palette! The biggest discontinuity between Heracross and the other two is that Plux and Scizor were of course second-evolutions, while Heracross didn’t evolve from anything. Was that always the case, or did the team ever envision Heracross as a dual-stage Pokemon? It’s possible, given how early Heracross was created, and how spotty our information on late 1998/early 1999 is, that the team may have briefly set aside a slot for a first stage Heracross. As we’ll see throughout the rest of Era IV, that seems to have been the default during this period: two or three slots would be set aside for a yet-created family, and then later some of those slots would go to other Pokemon if the designs didn’t need that extra member of the evolutionary family. Shuckle, Piloswine, and Yanma, for instance, were all designed with an extra slot available for an additional evolution, which was later repurposed into an unrelated Pokemon. It’s very possible Heracross was in the same boat, and its extra family member was given a slot for a month or so before being cut. If this did happen, mind you, the extra evolution probably didn’t last for very long; I doubt it even had a sprite or stats before being cut. There’s no real evidence for this except for Heracross’s relationship to Scizor and Plux. Furthermore, the Pokedex slots on either side of Heracross held Sneasel and Plux as far as our data goes back, so if there was a Heracross pre-evolution at one point or another, it would have had to fill a slot far away from Heracross, which is not something that typically happened when the team was designing Pokemon in Era IV. Still, it’s at least worth considering, even if the idea didn’t get very far. Plux, Scizor, and Heracross all first got their secondary typings on July 30th, 1999, suggesting that these secondary typings—in addition to being a way to make Plux and Scizor feel like more interesting upgrades for Pinsir and Scyther—were also done at the same time to help Heracross fit a niche unique from these other two. Heracross doesn’t actually learn many Fighting type moves, even from TM; it learns Counter, Reversal, and Detect (from TM), all situational attacks that depend upon what the state of the battlefield looks at the time. Making Heracross only able to use fighting moves to counter its opponents was probably partially a way to give Heracross a bit of identity: it feels very fitting for a Rhinoceros beetle, which can use its horn to lift its opponents in the air, would only use its fighting prowess to use its opponent's strength against them. However, the lack of Fighting moves was also probably a result of how late Heracross’s typing changed; the team likely didn’t want to mess with the moveset too much lest they unbalanced Heracross at the last second. Or they simply didn’t have enough time to mess with the moveset too much, so they just threw a couple fighting moves on him and called it a day. The most interesting move Heracross learns, though, is not a fighting move but a new Bug-type move: Megahorn. Megahorn’s a great attack: it’s a base 120 power Bug-type move that only Heracross learns. Signature moves are always cool, and I’m always glad to see them because they add a lot of flavor and identity. Even though only Heracross learns it, it’s also sorely needed in Generation II, especially because of how little diversity there was in Bug-type moves in Generation I. What’s interesting about the move is that it didn’t exist in the Spaceworld ’97 iteration of the game. Instead, Megahorn seems to have overwritten Megaphone, a signature move for Hoothoot that lowered Special Attack. Megaphone wasn’t nearly as flashy as Megahorn, and it isn’t a huge loss. However, my favorite thing is that Megahorn was probably created because of a misunderstanding about Megaphone. Both of them are spelled the exact same way in Japanese: there's not a sound for "fo" and so the the sound is often spelled as "ho" in Japanese words. Thus, Megahorn and Megaphone were spelled the same way! As a result, it’s likely that playtesters misread Megaphone as Megahorn while they were testing, mistakedly assuming that somehow Noctowl had a huge horn on its head! The move was probably changed to fix that funny misunderstanding. Who knows, maybe Heracross was created entirely so that the team would have a horned Pokemon that could make use of the new move! The last interesting thing about Heracross is its early sprite. This sprite dates from before June 1999, making it one of the few sprites we have from between May 1998 and June 1999, a period that is otherwise quite a mystery. We know that it’s an earlier sprite, by the way, and not a later experiment, because it appears that the backsprite for the June 1999 Heracross was based upon this early design: check out the arms on the back sprite and how they match the earlier pose: If we look at the shading of the sprite, it’s clear it was made for the GameBoy Color, dating it to after the rebooted development of Gold and Silver. Given that, there’s a lot that’s strange about this sprite. Final Heracross’s face is fun-loving, kind of goofy, but most importantly, anthropomorphic. Pretty much all of the Generation II Pokemon that made it to the final have expressive faces with big eyes or humanoid facial features (even the exceptions, like Piloswine, still look like they could use their body to be expressive); in general, I think that Masuda preferred that Pokemon had a way to emote or that looked cute. This could be another reason Plux didn’t make it to the final: Forretress’s eyes are big and expressive, while Plux has a creepy looking blank stare. But early Heracross, on the other hand, has a strange, bug-like face, much more akin to Pinsir’s face from Generation I. It’s a lot less cute and appealing than the final Heracross, and it’s pretty obvious why the face was amended. But it’s also a sign of the big differences between the design aesthetics of early Gold and Silver, and the aesthetics of the final game. In the Korean Index, there are a lot of discarded designs that have strange, unexpressive faces and would look right at home next to this early Heracross. For instance, the early Scorpion sprite, the two unused flies, the strange seed plant, and even Hanamogura all have strange faces that show a lot more diversity than the final game. The differences between early and late Heracross shows just how much design sensibilities had changed: we’ve gone from Pokemon being able to embody any type of strange monster, to the idea that Pokemon have to be, to some extent, relatable and able to express emotions with their faces. Heracross was a great addition! It has personality, it has appeal, and it shows that bugs are no pushovers this time around! I’m not sure I like trading it for Plux; I liked the idea of Pinsir evolving into something greater. Admittedly, though, Heracross fills a better niche. I’m glad we got something like it in the final, even if Pinsir was left high and dry. ID X02: Totodile (Colorization of early Totodile sprite is also done by @OrangeFrench!) Moving on from Heracross, it's time for Totodile! Probably one of the most beloved starters in any of the Pokemon games, Totodile has a ton of fans, in part because of how good it is. However, as we know, Totodile (probably a portmanteau of "Tot" and Crocodile) wasn't in the first iteration of Pokemon Gold. Even so, all of our evidence suggests it was one of Masuda's main priorities. Even if there's no sign of the Totodile line before the hiatus, by April 1st, 1999, all three of the Totodile family had stats, though these stats were barely more than placeholders, stolen directly from the Squirtle line. They also had name data, and by June 1999, each of them had both a front and backsprite; these front and backsprites were well fleshed out and used all the way to the final, suggesting that a significant amount of work had already been put into them. Totodile, like Heracross, even has a very early sprite that probably pre-dates the June 1999 sprite bank we have. This sprite (seen above) notably has a lot less personality, a brown nose, and has a different design for its stomach, more akin to the design of Feraligatr. While this sprite could have just been an experiment while the team was concepting what Feraligatr would look like (it was found on Feraligatr's Scratchpads), I think it's more likely that it was a first draft of Totodile, right at the beginning of the rebooted development. From all of this, it’s very likely that one of the first things that changed after the reboot was a redesign of the original water starter line, Kurusu, Akua, and Akueria. Though their stats (and presumably movesets, though I don’t have access to that data) would need a bit of tweaking, Totodile and its family were already substantially finished by June 1999. I’m on the record as being a huge Kurusu fan. It’s super cute, I love the Pleisosaur design that it has, and I think the evolutions do a great job building on the design that’s already there. Saying that, I get why Masuda might have first turned towards the Kurusu line as something that needed fixing. The line is very lacking in design hooks. If you look at Bulbasaur, the bulb on its back not only is a cool and memorable part of its design, but suggests how the Pokemon will change as it evolves. Charmander, likewise, has the flame coming from its tail, a distinctive trait. On top of that, Charmander becomes an iconic dragon; Squirtle is clearly a turtle, and Bulbasaur, while a little more vague, looks enough like a four legged dinosaur that it’s easy to grok what its supposed to act like. The Kurusu line doesn’t really have any of those elements in their designs. Kurusu doesn’t really hint at what its evolutions are going to look like, beyond bigger sea-dragons. It doesn’t have any distinctive design hooks in its design; it’s mostly smooth and light colored. And it isn’t even that clear what animal it’s supposed to be inspired by: is Kurusu a seal? As cute as it is, it doesn’t do the job a starter is supposed to do, and I think the design team recognized that. Totodile, on the other hand, succeeds at all of these design requirements. While I personally think Totodile is much less cute that Kurusu, it immediately pops out to the player. Its face is extremely expressive, telling you that Totodile is excited and fun-loving. It’s teeth and snout immediately let people know that it’s an alligator (or crocodile) and there are all sorts of aspects to its design that could potentially grow as it evolves. Will it get a bigger, fiercer jaw? Will the spikes on its back become more pronounced? Spoilers: all of the above. Another way that Totodile’s design feels fresher than Kurusu’s is that Totodile’s sprite is filled with sharp angles: the pointiness of its snout, its teeth, and its back spikes all give Totodile a prickly, triangular look to it. This is a notable departure from most of the designs in Gold and Silver: notice how many Generation II Pokemon are circles or at the very least spheroid and squishy. Natu, Igglybuff, Octillery, Forretress—they’re all just circles. Kurusu fell into this trap; it wasn’t circular per se, but it did have a soft, huggable look to its body that didn’t help it stand out. Even if you don’t immediately notice why, Totodile looks distinctive amongst the rest of Generation II’s Pokemon, helping make it far more memorable. Totodile is also a triumph because it gave Game Freak a way to resolve a puzzle they’d been trying to solve since the initial stages of Gold and Silver’s design. The problem was that new starter Pokemon were tricky to design: two different gameplay needs pulled their designs in opposite directions. On the one hand, the designers want to make Gold and Silver new experiences for players: they don’t want the game to feel exactly like what they had done before in Pokemon Red. Thus, the designers would want to make the starters as different from Charmander, Bulbasaur, and Squirtle as possible. On the other hand, the starter Pokemon fill a very specific role in the gameplay of the Pokemon games. The starters are supposed to be flexible, all-around strong Pokemon that are good no matter what team you construct with them. They’re also supposed to have a huge movepool, so that they offer maximum flexibility to the player, and good enough stats that they can carry an inexperienced player through hard sections of the game. Thus, the designers would want to make the starters all pretty similar, because the stronger they are and the larger the movepool, the more choices a player would have. So one gameplay need wants the starters to be different from the first generation of starters; the second wants them to be more or less the same. What are the designers to do? When we discussed Chikorita and Cyndaquil, we discussed one way the designers were able to square this circle. Chikorita and Cyndaquil have very similar stats to their counterparts (in fact, Cyndaquil’s stats are identical to Charmander’s) in Generation I, and their movesets aren’t very different either. The way to make them feel unique comes from the design of the games. In Generation I, Bulbasaur was the easiest starter to use, while Charmander was the hardest. However, in Gold and Silver, Chikorita immediately faces difficulties and throughout the game has a harder time than the others, while Cyndaquil blazes through Bugsy and has a decent time against most of the other gym leaders. Thus, even though Chikorita’s really similar to Bulbasaur, the experience of using Chikorita is quite different. Totodile, on the other hand, tries to solve this tension by providing a drastically different experience in a Water starter, while still being a flexible, all-around good Pokemon to use. While Squirtle was a defensively oriented Special Attacker, Totodile’s stats are distributed much more towards Attack and HP, meaning that it’ll excel with physical moves, not the Water moves that Squirtle’s more or less balanced around. As a result, Totodile learns much more physical attacks than Squirtle: it learns Rage and Slash, which Squirtle does not, and it only learns one single Water move before level 52. As a result, Totodile is going to spend most of the game using its physical stats, rather than even bothering with same-type moves. Notably, Kurusu has a third distinct playstyle. While we don’t have stats for Kurusu, it learned Aurora Beam and Body Slam, both strong sweeping moves, alongside Safeguard and Mist to make sure it wasn’t debuffed. If Squirtle was a defensive special user, and Totodile a physical dps, Kurusu seems to have been designed to be a versatile offensive Pokemon. So while I love Kurusu, I can definitely see the argument for replacing it with the Totodile line. And given how popular Totodile is with the Pokemon community, that decision has been borne out by time. Totodile’s an interesting, effective design, one of the biggest successes of the rebooted 1999 development. ID X03: Croconaw Croconaw is Totodile’s grown up brother. A little bit bigger, and sporting what almost looks like a cave-man fur across its body, Croconaw is still cute, but notably looks far more grown up and a bit more dangerous than Totodile. There’s not much that’s new with this guy, but let’s jump in and see what we can find. My gut feeling is that Croconaw and Feraligatr were designed after Totodile. Croconaw, for instance, looks like the stereotypical second-stage evolution, clearly awkward looking on its own but acting as a design bridge between the aesthetics of Feraligatr and those of Totodile. Croconaw has the white, pronounced jaw of Feraligatr, but its pose and the spikes on its back look a lot more similar to Totodile. If I had to guess, however, I’d say all three were designed more or less at the same time. Their stats were all changed at the same time in the data we have, and Totodile, Croconaw, and Feraligatr are some of the few new Pokemon that have both a finalized backsprite and frontsprite by June 1999. This suggests the team thought of them as a group to be worked on together (as opposed to the Cyndaquil or Chikorita lines, which lack one or two key members of the family until much later). If I had to guess, I’d posit that Totodile was designed by a different creator than Croconaw and Feraligatr. Notably, the spiky, hard angles of Totodile are gone in Croconaw’s design, replaced with more curves, most notable in Croconaw’s large snout. The eyes are slightly different as well, but most strangely, the palette for Croconaw (and Feraligatr) is a green/turquoise color, whereas Totodile is a bright blue. If you look closely, the palette is actually different for all three, with Croconaw’s being just a slightly lighter shade of green than Feraligatr. This is just another example of how Croconaw feels like it was purposefully designed as a bridge: even though the difference is barely noticeable, the designer of Croconaw wanted even its color scheme to be a transition from Totodile to Feraligatr. Obviously, this palette difference could be for a million reasons, but one possibility is just that someone else was working on Croconaw and Feraligatr and settled on a different shade of blue they liked more. Obviously it wasn’t a palette change that the team liked overall, because in Crystal version they changed all the sprites of Croconaw and Feraligatr to just share the same blue that Totodile uses. One thing that is especially notable is how the Crystal sprite brings Croconaw much more into line with Totodile's aesthetics. If you examine it closely, not only does Crystal Croconaw have the same blue palette as Totodile, but the floppy bangs it has on its head in the initial sprite have been replaced with spikes more in line with Totodile. For some reason, the back sprite has also switched which shoulder the cave man furs tie over, even though that doesn't match the front sprite. Who knows why. We know that Crystal's sprites were tweaked to better match Sugimori's official art that was released after Gold came out. Sugimori based his Croconaw art more on Totodile's coloring and design, so these tweaks to Croconaw's sprite were probably to bring it in line with those. Does that mean that Sugimori was the one who designed Totodile? Does that mean he's the one who edited the Crystal sprites? It's hard to say--the shading of GBC sprites makes it hard to say anything definitive. If Croconaw was made by a different designer than Totodile, it was probably designed by the same person who created the original three Water starters for Generation II. We’ve discussed this before, but it’s worth bringing up again here. For whatever reason, the designer decided to leave an easter egg in Croconaw’s design by leaving it with the exact same pose of its predecessor, Akua: Alright, alright, the pose isn’t completely definitive. Their heads are angled up at the exact same angle, but of course, they have completely different bodies and their tails are in different positions, so I’m willing to grant that it could be a coincidence. But man, I look at these two sprites and their similarity feels too obvious to ignore. On top of that, their faces also share a likeness: just look at the ovular shape of the mouth and snout. It seems to me pretty much beyond doubt that the designer of Croconaw put just enough of Akua into the new creation so that it kept some of the personality and was a subtle reminder of what came before. The only other notable thing about Croconaw is that its Silver Pokedex entry changed between the end of Spaceworld ’99 and the final. In the early silver entry, the Pokedex makes reference to a behavior of real-life crocodiles: It spends most of the day underwater, with only its eyes and nose sticking above the surface. I don’t think this entry really works; given the pose Croconaw has in its spritework, it doesn’t feel like the sort of creature that lurks below the waters’ surface. I mean, look at that defiant pose it’s striking! This is a dude who doesn’t hide, but faces down his enemy. Game Freak might have thought something similar, so by the final, they gave Croconaw a new entry, which was actually just reworked from an early Totodile’s entry that was substantially the same: It opens its huge mouth to attack. When it bites something, any fangs that fall out will grow back one after another. Meanwhile, Totodile got this entry, expressing a key feature of its excitable personality: It is very rowdy despite its small size. It will bite anything that moves in front of it, no matter what. Honestly, it doesn’t feel like the Pokedex team worked very hard on Croconaw. Its early entry and its final just reference fun crocodile facts: Crocodiles spend most of their time lurking beneath the water’s surface, and when they lose their teeth chomping on things, crocodiles are also known to grow them back. Which is fine and all, but I don’t really get a sense of Croconaw’s personality from its revised Pokedex entry. Honestly, it seems like the team just opened an encyclopedia to the crocodile entry and wrote down whatever seemed the most neat. Croconaw’s not a bad design, by any means, but it certainly is the least memorable of the Totodile family. Like I said about Totodile, I personally prefer the original Water starter line, especially the middle evolution, Akua, that Croconaw replaced. Still, I understand why they did it, and Croconaw retains a lot of the mischievous energy that the team crafted into Totodile. I’m certain that even Croconaw, the overshadowed middle stage evolution, has more fans than Akua ever would have. ID X04: FeraligatrAnd finally, we make it to Feraligatr. Feraligatr’s ferocious: Our boy Totodile has completely grown up, leaving this absolute monster in his place. While Meganium’s my personal favorite Generation II starter, Feraligatr is a great example of how to transition a first stage Pokemon from cute to fearsome but still keep it feeling like the same ‘mon. Feraligatr feels like a natural extension of Totodile—well, minus the strange turquoise color that they thankfully got rid of by Crystal—in that it keeps all of the traits of Totodile but just ramps them up to eleven. It is worth noting that Feraligatr, according to its name, officially changes real-life species when it evolves: Totodile and Croconaw are both crocodiles, and Feraligatr is an alligator! Obviously, the difference between these two species is small, and so this isn’t as big a deal as, say, a Pokemon transforming from a fish into an Octopus (what kind of design would that be?). What’s the difference between Alligators and Crocodiles, you might ask? Well, there are three main ones: Crocodiles are greener in color typically, Crocodiles live in freshwater while alligators live in salt-water, and Crocodiles have a pointed snout while alligators have a rounded snout. Feraligatr breaks this first rule, by being greener than the other two (maybe Sugimori changed its palette to address this inconsistency!). But if you look at its sprite, it absolutely has a more rounded snout, a clear contrast to Totodile’s. Croconaw, the odd one out, unfortunately has an alligator like snout. (Alligator on left, and Crocodile on right) As for the second difference, I have no idea whether Feralgatr prefers salt water or fresh water. It’d be pretty awkward if, on evolving, a Croconaw suddenly couldn’t breath in the fresh water environment in which it lived its whole life. Everything that we said about Totodile goes for Feralgatr, so there’s unfortunately not much more to add here. Feraligatr’s stats started out as a reflection of Blastoise’s, before becoming focused towards HP and attack; likewise, Feraligatr specializes in biting and slashing moves, as opposed to the heavier dependence on water moves that Blastoise and Akueria had. Feraligatr, like its two relatives, was also substantially finished by the first sprites we have, in June 1999. In fact, the only things that got changed around before the final of Gold and Silver were the Silver sprites for each member of the Totodile family. Totodile’s early Silver sprite was the same pose and concept, just cleaned up a little for the final game; Croconaw’s early and final both show it growling at an enemy, but the pose is pretty different between the two, and substantially cleaned up by the final. Feraligatr, however, has the Silver sprite that went through the most drastic transformation. In the original Silver sprite, which we first see in Spaceworld ’99, Feraligatr is on all fours! It’s a pose that Feraligatr has never been associated with, even on the anime, so it took a lot of people by surprise. I, certainly, think it’s a really cool pose for Feraligatr, and it gives it a lot of personality. The designers may have thought it diverged from the Gold design a bit too much, however, because the final changed the Silver pose. Its head is still pointing downwards, and we’re viewing Feraligatr from an angle pointing almost directly at the camera, but now Feraligatr is standing up on its hind legs, just like the Gold sprite. It very much feels like this revision was made to bring the Silver sprite more in line with the official design, not because they disliked the previous Silver sprite. There is one thing I find fascinating about this final Feraligatr sprite. Look at its pose, and compare that pose with Akueria, the water starter that Feraligatr replaced. Don’t those poses look similar? Obviously it looks different because Akueria has such a long neck, but the angle facing the camera is almost the same, and though Feraligatr’s tail is hard to see, it appears to be in the same position as Akueria’s. Certainly the similarities in poses are not as definitive as Croconaw’s. But I also have a hard time ignoring these similarities either. That’s it for the Feraligatr line. From what we have, their design history seems pretty simple. However, that early Totodile sprite suggests that there was a lot of work done on these guys in the period of development between the Korean Index and the June 1999 sprite sheet. Unfortunately, we have so little information from that period that any interesting experiments concerning the development of the Totodile/Croconaw/Feraligatr sprites are lost.
Saying that, what we know seems pretty conclusive. It appears that Masuda wanted to replace the Akua line, and that’s exactly what happened. It seems as though the design team was very decisive about the replacement Water starters, and quickly came up with a winning concept and strong sprites. However much time the team spent perfecting the sprites of Totodile, Croconaw, and Feraligatr, they were very happy with them by June 1999. The team stuck it out with these sprites all the way until Crystal, demonstrating just how good these designs were. |
AuthorMy name's Aaron George, and I'm both a historian and a fan of Pokemon, especially of development. Reach me at @Asmoranomardic Archives
February 2024
Categories |