Period 4a was composed of Pokemon created when Gold and Silver’s development was rebooted, sometime in late 1998 or early 1999. Period 4b were those Pokemon which were created soon after that, usually to fill spaces that had been vacated by problematic old designs. Both of these periods of design can be seen as the early half of the 1999 reboot. However, with Period 4c, we step into the very end of development. We're done with the brainstorming phase and now the team is working out anything that would need to be ironed out before the final build. On July 30th, 1999, the designers made a huge amount of changes to the Pokemon data. Lots of Pokemon got their stats updated to their final numbers. Sudowoodo and Gligar, from Period 4a, are two examples. Some, like Yanma, got placeholder stats replaced with real stats, and some, like Sunkern, got temporarily deleted. Lots of Pokemon finally got a moveset, and the Magnemite line, after a brief flirtation with the Electric/Normal typing, finally became Steel. In addition, at least nine Pokemon probably made their first appearance on July 30th, or thereabouts. Those are the Pokemon we’re looking at here in Period 4c. Why was the team so prolific on July 30th? One option is that the team spent June mostly figuring out the new map, the new plot scenarios, trainer teams, and other nuts and bolts of the gameplay of Gold and Silver. When they had time, they made a note about what they’d like to fix about Pokemon stats, and those notes accumulated, until July 30th was deemed the date to put all this accumulated balancing into the build at once. Another possibility is that, after a June focusing on other things, the team focused their attention on balancing in late July and spent a week or so going over the stats of all the new Pokemon with an eye towards getting rid of placeholders and finalizing everything. Either way, this seems to be a point in development focused on tweaking and on finalizing. In Periods 4a and 4b, it might have been alright for them to design a sprite and leave the moveset or stats until later, but by 4c the games were getting close enough to release that they didn’t have that luxury anymore. Whatever was happening, July 30th—followed by two more waves on August 14th and August 22nd—was a turning point in design where everything needed to be done and playable. So why add nine Pokemon at this late point? Simple: there were still empty spots in the Pokedex which needed to be filled in. At this point, though these spots were at once filled with old Spaceworld ’97 designs, they hadn’t been used for awhile. Period 4b was mostly characterized by a need to replace Pokemon that weren’t working with ones that were. This included switching out the evolutions of starter Pokemon for ones that made sense, replacing old baby Pokemon with original ideas, and getting rid of the Kotora line—which probably had problems filling a unique niche—with a new family meant to show off the honey mechanic. But the Pokemon in Period 4c weren’t replacing anything in particular: instead, they were just filling empty spaces that had long been vacated, or they were finally fleshing out Pokemon slots that had been marked with a "To Be Determined" sign for months. Importantly, this is also where the important Pokemon designer, Hironobu Yoshida, appears for the very first time. Yoshida seems to have been brought on to Generation II’s development towards the very end; it’s possible he was originally there to touch up other people’s sprites. But by Period 4c, Yoshida was asked to fill two empty slots (and in 4d he was asked to fill at least one more, which became Wobbuffet). At least in Period 4c, both of Yoshida’s Pokemon—Dunsparce and Celebi—took inspiration from designs found and discarded in the Korean Index. It’s likely that he went back to the designs of the rest of the team for inspiration (also, maybe because he had started just by fixing other people’s sprites, he went back to old sprites he could make his own) by searching through the Korean Index for ideas he could make work. My suspicions are that he made a lot more of the designs from here on out than the three we know: Girafarig-1, Misdreavus, Pineco, Corsola, and Kageboozu could also be his handiwork. Anyway, this late stage is significant because Yoshida is significant: he'll go on to have a huge effect on future Pokemon games. ID X28: Girafarig-1 And we’re back to the cursed, byzantine development history of the Girafarig family. We’ve already discussed, in detail, Girafarig, its pre-evolved form Twinz, and its short-lived final evolution, Girafarig-3. I’m going to rehash as little of that as possible here, so by all means go check out those articles for more details about the rest of Girafarig’s story. Here, though, I want to discuss only what we can about Girafarig-1. To briefly sum up the Girafarig story: Girafarig originally had a pre-evolution calls Twinz. Twinz was a Ghost Pokemon composed of two Haunter-like ghosts either connected to each other or swirling around each other, depending on the iteration. The idea was that these two conjoined ghosts would evolve into the palindromic Giraffe, and each ghost would become one of its heads. As cool as this idea was, it had real problems in practice: first, Girafarig didn’t really fit as a ghost type but Twinz felt wrong without the type; second, Girafarig and Twinz had very different body types, making it difficult to come up with a coherent moveset for them. The team obviously loved Twinz, because it got a reworked sprite, a shiny palette, and even an early Pokedex entry during the rebooted 1999 design. But clearly something wasn’t working. The first thing the team tried was to give Girafarig a third evolution. We don’t have a sprite for it, so we can only speculate, but they probably didn't get very far on it; instead, that Pokedex slot was probably set aside for the third Girafarig but the team never got around to actually designing it before the Big Overhaul. Another possibility is that the team never planned for it to be a completely new Pokemon. The plan could have been to make Girafarig into the third form and make the second slot a transition evolution that would make the transformation from Twinz to Girafarig less jarring. During the “Big Overhaul” of July 30th, the team had another idea. Around this time (we can’t say it happened exactly on July 30th, but evidence points to that), the team decided to finally, once and for all, scrap Twinz. In its place, they created a baby version of Girafarig. Here’s its backsprite, with Girafarig's backsprites as comparison: We don’t have the front sprite, as I’ve noted before. It’s a shame, though I’m not sure we’re missing much: in reality, this was probably a basic looking, slightly cuter, version of Girafarig. Saying that, Farore on twitter made this absolutely wonderful Girafarig-1 sprite. I doubt the actual one was this ghostey, but it’s still a great interpretation: Another reason this sprite probably isn’t accurate is that, interestingly, Girafarig-1 probably wasn’t Ghost-type! As we’ve seen, Twinz was originally Dark/Normal, then Ghost-type. But at the exact same time as Girafarig-1 seems to come into existence (we can’t be sure because we don’t know exactly when the sprite was made), “Twinz” became Psychic/Normal type. Its stats also changed significantly, losing the physically frail aspects that the Ghost-type Twinz had. Fourteen days later, Girafarig-Classic also changed its stats to match (it previously had stats twenty higher than Twinz in every category). As you can see below, they just added thirty to all of Girafarig-1's stats to get the numbers that Girafarig uses in the final game. These stat and typing changes are significant: they prove show that Girafarig-1 was more than just a new sprite for Twinz. Instead, this was a completely new Pokemon. At this point, the team had finally given up on Twinz and decided to replace it with something that made more sense as a baby Girafarig. It's also interesting because Girafarig-1 got these stats before they changed regular Girafarig: this leads us to conclude that Girafarig was changed to coincide with its baby form first, before the baby was deleted. Interestingly, this was also probably (again, we can’t be exact about the dates) the moment when the team also got rid of Girafarig-3. For whatever reason, creating Girafarig-1 coincided with the team deciding that Girafarig should have only two stages and not three. Two weeks later (Girafarig-1 barely got to live!) Girafarig-1 got replaced by the beginnings of Wobbuffet and Girafarig became a one-stage Pokemon officially. It's unclear why Girafarig-3 would get replaced at the same time, but there are two good explanations. The first—and most likely—reason is because these two changes both just happened to coincide with the “Big Overhaul” of July 30th. Maybe the idea was just to get Girafarig's house in order before the final deadline and these were two changes they had already decided on. The second explanation—and the more interesting one—is that Girafarig-1 was actually an idea they had had earlier, in June 1999, as a transitory stage between Twinz and Girafarig. The three stages of the Girafarig family may have been first Twinz, then Girafarig-1, then regular Girafarig, which happened to be in the final slot. If this was the case, the changes in July might be less a deletion of Girafarig-3, and more a movement of all the stages backwards. Girafarig may have been in the 3 slot, Baby Girafarig in the 2 slot, and Twinz as the first stage; in the “Big Overhaul” of July 30th, the team may have moved Girafarig back to slot 2, moved Baby Girafarig to Twinz’s first-stage slot, and then deleted Girafarig-3’s spot. In this reading, it wasn't like Twinz and Girafarig-3 got removed at the same time; instead, Twinz got deleted, its spot overwritten by Girafarig-1, and a new spot opened. Around this time the team also gave Girafarig-1 the same moveset as Girafarig, learned at earlier levels. That's far more than Girafarig-3 ever got, since the team never made it a moveset at all, and it's another sign that Girafarig-1 was under much more consideration that Girafarig-3 ever was. This was probably a change made as part of the “Big Overhaul” as a way to finally get Girafarig into a playable state (it also had no moveset until this point). The fact they were able to assign a moveset to Girafarig and its pre-evolution at this late point might suggest that they hadn't made a moveset earlier because of the problem of Twinz. Potentially, that means that the impetus behind Girafarig-1 may have originated in the need to finally give Girafarig some sort of moveset. Realizing that any moveset wouldn’t work if Twinz also got it, and that they were running out of time, the designers quickly remade Twinz into something that could get an appropriate moveset. I’m still not sure why Girafarig and Girafarig-1 were made Psychic rather than Dark though. Obviously, neither of them were Ghosts any longer and that typing needed to change. But Girafarig’s tail is clearly its evil counterpart, which seems to suggest Dark-type. I have two theories. First, for whatever strange reason, the team had decided by the final game to wall all Dark-types (except Umbreon) behind the Elite Four and make them post-game only. While I think that an early Dark-type Girafarig would have given Girafarig a unique niche—not to mention done wonders for Gold and Silver’s diversity in the early game—they may have made Girafarig Psychic type purely because they wanted it in the Johto section of the game and they didn’t want players to have access to Dark-type Pokemon. A second reason may be because the team didn’t want the Dark-type to be defined as the “evil type.” The developers have stated before that they wanted every Pokemon to have the potential to be the player’s partner or friend, and so they didn’t want any type to be just “bad guys.” However, making Girafarig Dark-type would imply a good/evil binary, and maybe the team were just hesitant to encourage this association. Whatever the reasons for creating Girafarig-1, like I mentioned above it didn’t last long. By the Spaceworld 1999 build, Girafarig-1 was in a transitory state, on its way to deletion. It still has Twinz’s name (so it probably never got its own unique name), it had the new Girafarig moveset, it still evolved into Girafarig (at level 24) and it had the backsprite we’ve seen above. But its front sprite is a goofy drawing of what would become Wobbuffet, which clearly means Girafarig-1 was on its way out. The fact that so much of its data was still intact suggests to me that the Wobbuffet idea was very recent as of that build, maybe even a couple of days. Still, that means that Girafarig-1 probably only existed as an idea from July 30th to August 17th, or three weeks, at most. It isn’t clear why the team so swiftly decided against Girafarig-1, but we have at least one small clue. Hironobu Yoshida designed Wobbuffet, and clearly he’d been given Girafarig-1’s spot in the Pokedex by August 17th with the mission to create something new. Given that his other confirmed Pokemon—Celebi and Dunsparce—were very late additions, it seems likely to be that Yoshida was a fixer: he was allowed to create Pokemon in slots that needed to be filled very quickly before release. If that’s true, that tells us that for whatever reason, the team had decided something about Girafarig-1 didn’t work. What that is is up to speculation: maybe a tiny Giraffe was odd because Giraffe’s are known to be large, maybe it didn’t fit into any of the Johto environments, maybe the designers disliked the front sprite. Who knows. But Girafarig-1 wasn’t scrapped because they needed to make room for more new Pokemon, and it wasn’t scrapped for any other external reason. It was given to Yoshida because the team decided Girafarig would be better as a single-stage Pokemon and not a dual stage. Maybe they just missed Twinz and thought the baby giraffe theme wasn’t compelling on its own? If that’s the case, I disagree. Girafarig has such an interesting design and yet was so undercooked in the final games. Giving it a cute evolution would have made it so much more memorable. Girafarig deserved better. ID X29: Pineco When Girafarig-3 was deleted to make way for the baby Giraffe, what did it get replaced with? Turns out it was Pineco, the world’s cutest pine cone. Pineco went through a number of changes in the short time between its development and the final, so it’s worth looking into. The first place to look when researching Pineco is back in the Korean Index. Remember ID 382, the bagworm Pokemon? For reference, here he is, right next to Pineco: Palette for 382 done by @OrangeFrench! One of the most popular theories is that ID 382 was an early design for Pineco. In the article on 382, I argued against this idea, mainly by pointing out that Pineco was originally designed as a single stage Grass-type Pokemon, while the Bagworm guy was definitely a dual-stage Bug-type. I mean, ID 383, the moth, is obviously an evolution of it and they even share some similarities in their sprite. More likely, this idea was either directly revisited in Generation IV to create Burmy and its evolution, or coincidentally the idea of a bagworm Pokemon was revisited to create Burmy, which was superficially similar to 382 and 383. There’s a middle ground explanation as well, though. As I stated above, Yoshida was brought aboard the Pokemon development team very late in Gold and Silver’s development, mostly to quickly generate Pokemon ideas for holes in the roster. The confirmed Pokemon he created were Celebi, Wobbuffet, and Dunsparce. But I think he might also have designed Pineco, and maybe its evolution, Forretress. One reason to think this is that Pineco took over the slot opened by the deletion (or whatever) of Girafarig-3. Pineco's name first appears in the data in Girafarig-3's slot on August 1st, during the Big Overhaul. Though it isn't entirely clear if Girafarig-1 showed up here or slightly later, the evidence suggests it replaced Twinz at almost the exact same time. Now, again, this could be a symptom of so many things happening around the “Big Overhaul.” But this could also be a sign that Pineco and Girafarig-1 were created at the same time: Pineco to take over the unneeded slot that Girafarig-3 used and Girafarig-1 to fix the train wreck that was Twinz. And if Yoshida was asked to quickly generate a Pokemon to fill the empty Girafarig-1 slot three weeks later, it is very plausible that the team had also thrown him the Girafarig-3 slot as well, at the same time (If we postulate that Yoshida also created Girafarig-1 too before scrapping it to produce Wobbuffet two weeks later, which is absolutely possible, then this timing becomes even more plausible). After all, they were both two empty Pokedex slots, opened at around the same time, within two slots of each other, and it's plausible that Yoshida played a role in both of them. (There's also some evidence that Pineco may have made an appearance even earlier, on July 20th, but we'll get to that in the next entry). Furthermore, we know one key signature of Yoshida’s designs. Of the three Pokemon we know he created, two of them—Dunsparce and Celebi—took obvious inspiration from unused Pokemon found in the Korean Index. Celebi’s first sprite (which we’ll talk about) draws from the Kokopelli sprite at the very beginning of the Korean Index, while we’ve already talked about the Proto-Dunsparce that had to have inspired the final Dunsparce design. So we know that Yoshida—facing a very short deadline and probably used to, by this point, fixing up other people’s sprites—turned back to the unused designs when he was looking to quickly fill a slot. Thus, it’s not farfetched to me that Yoshida may have also created Pineco after being inspired by ID 382 and changing it to be his own design. That’s not for certain, and I still think that Pineco’s original Grass-typing is good reason to doubt any direct connection between the two. But it’s also worth noting that Pineco is called the “bagworm” Pokemon in the final and there’s good reason to believe it was also the bagworm Pokemon from the moment it was created. And that’s before we even look at Pineco’s Pokedex entries. By August 17th, Pineco already had a Pokedex entry that suggested it was a bagworm. Sort of: to get into the weeds, Forretress originally had Pineco’s entries, and they were swapped by the final game: my guess, which we'll get into later, is that these two Pokemon were originally the same concept before being split into two evolutionary stages. Anyway, Forretress’s (soon to be Pineco’s) entries are telling. Its Gold entry (below) was used for Pineco in the final, and its silver entry—very similar to the Gold entry—was deleted before the final: Gold: It likes to make its shell thicker by adding layers of tree bark. The additional weight doesn't bother it. Silver: It will consume any kind of tree bark. It plasters any leftover bark to its shell to make it stronger. While I could plausible see either as being traits of a pine cone-mon, both entries sound more like something a bagworm would do. Even the mention of “shell” makes me think of a Bug rather than a pine-cone. Anyway, even if Pineco was originally conceived of as a Grass-type, it didn’t take long to fix that. It was Grass-type on August 14th, and a week after Spaceworld ’99 (August 22nd) it was officially made into a Bug-type Pokemon. At the same time, it was given an evolutionary link to Forretress, ending its 1-3 week reign as a single-stage Pokemon. Its stats were also lowered to make room for Forretress: a decision I doubt was playtested much before the final release, as Pineco and Forretress are remarkably weak in the final games. As an aside, I wonder how much the headbutting mechanic influenced Yoshida’s choice to make a bagworm/pine cone Pokemon? It seems very likely that Ursaring and Tediursa were inserted into the game to show off the Honey mechanic, and that was just a few months earlier than Pineco. It’s plausible to me that Yoshida chose the bagworm idea because it could hang from trees, and thus it could fit well into the headbutting mechanic that was probably being developed or finished around this point in development. Certainly, in the final games, Pineco feels like the most exciting Pokemon you can find in a headbutt tree: everything else is more or less the bland bug Pokemon you can already find around the Ilex forest. Pineco’s moveset is also fascinating: The first thing to note is how unique it is. Bide was of course a TM in Generation I, so none of the Gen I Pokemon learned the move. Pineco is, then, the first Pokemon to learn it naturally, and the only Pokemon in Gen II that learns it naturally. Pineco also learns Rapid Spin, a new move limited to the Squirtle, Staryu, and Donphan lines. It has a very low base power in Generation II but is the only way to remove things like Leech Seed and Spikes. Spikes, funnily enough, is actually the signature move of the Pineco family in Generation II! While Spikes had been designed all the way back during the Spaceworld ’97 days, the design team hadn't been sure what to do with it. In Spaceworld ’97 it was assigned to Murkrow, probably because the move felt like the kind of dirty trick a Dark-type Pokemon would employ. But Spikes didn’t make a lot of sense on Murkrow, and so during the reboot Murkrow dropped it and for a time, no Pokemon learned it. Pineco didn’t even learn the move originally: it was added to Pineco’s move list very late in the game, on August 23rd. Given that Misdreavus gained Pain Split, another previously unassigned move (originally given to Norowara in SW97) on August 23rd, and Sneasel got Beat Up on August 30th, it seems that Game Freak was, last minute, looking for places to put moves they had created but had gotten forgotten in the shuffle. Pineco seems to have learned Spikes because it was the most plausible Pokemon to put this move on. All those moves are nice and interesting, but Pineco lacks one critical part of a good moveset: a good move that does damage! For a regular damage-dealing move, Pineco is limited to Tackle, Rapid Spin, Take Down, and Double Edge; after that, it gets the afore-mentioned Bide, Self-Destruct, and Explosion, which all technically do damage but which are very hard to use. Given the recoil effect of Double-Edge and Takedown, it seems the team liked the idea of Pineco having a "hurt itself to hurt the opponent" theme. Which isn't very pine coneish but whatever (I guess a pine cone hurts when you step on it?). But glaringly, Pineco learns no STAB moves! It learns no Bug moves at all (Ironically, it learns Giga Drain and Solarbeam from TM, which would be STAB if it remained a Grass-type). You could teach it Return, but Pineco’s probably one of the worst Return users in the entire game. It's not hard to figure out why the team did this. There aren’t many good Bug moves in Generation II, though obviously Pineco should have gotten access to at least Fury Cutter. And it seems they wanted Pineco to be Generation II’s Voltorb: a Pokemon which is prone to exploding and ruining your day. The problem is, the moveset they gave it makes Pineco pretty unusable in actual gameplay. I suspect the team may have relooked at this moveset had they had more development time, but I don’t know: Generation II made some strange choices. Again, it’s another case of Gen II making all its new Pokemon either unobtainable or really bad for seemingly no reason at all. It’s just especially bad in Pineco’s case, given that Forretress is one of the only new Steel Pokemon a trainer can capture without using the link cable. It’s too bad: Pineco has a cool design, and I’d love to be able to use it and feel like it contributed to the team… ID X30: ForretressI wasn’t sure quite where to put Forretress in this list, because its origins are so murky. We’ve got evidence that Forretress was created on July 20th, at least two weeks before Pineco; we’ve got evidence that it wasn’t made until two weeks after Pineco. We’re not even sure which slot Forretress originally started in, though we know it probably was designed to take the role of Plux, the deleted Steel-type Pinsir. Forretress has an odd history: let’s walk through it. First, some general thoughts about Forretress. Like Pineco, Forretress is “the bagworm Pokemon.” Unlike Pineco, Forretress isn’t obviously based on a pine cone or any particular type of plant. The closest thing it resembles is a bug that’s crawled inside a Walnut or Chestnut or something, though Forrestress’s shell is made of metal. It’s a weird coincidence that Shuckle was also designed in the same generation, and that Shuckle is also a strange bug-guy who hides within something else (no to mention Magcargo, another bug hiding in a shell!). One of the designers must have had hidden bugs on their mind; maybe they had a roach situation at home while designing this game. Forretress is also unique in that it’s one of the few Steel-type Pokemon available to players in a normal playthrough of Gold and Silver. Magnemite and Magneton, of course, were made Steel/Electric types for Generation II, and they’re the easiest Steel types to acquire in the games. But other than that, Steelix and Scizor are locked behind trade evolutions, and Skarmory is a late game find, exclusive only to Silver version. For the vast majority of the game, Magneton and Forretress are the only Steel-types available to a trainer. Which makes it a shame, because Forretress is awful. As I stated above, the designers balanced Forretress and Pineco with an abundance of caution, which made them more or less unusable. Forretress is an excellent idea for a Pokemon, but it’s unusable in a normal playthrough. In some important ways, Forretress was a spiritual successor to Pinsir’s evolution, Plux. Plux was one of the very last Pokemon to get replaced before the final game, and my guess is that Plux and Tsubomitto (the Weepinbell evolution) were very close to becoming official Pokemon. The problem for Plux was that it was too similar to its counterpart, Scizor, and to Heracross, which also served as part of a buff-bug-type triad with the other two. It seems like Forretress was made, in part, to continue the idea of having another Bug/Steel Pokemon, but give it a different niche as a more defensive type so it felt distinct from Scizor and Heracross. Of course, they didn’t do a great job of it, because—as I’m sure I’m impressed upon you—those Pokemon are good, and Forretress is not. Image credit to Shulky, on Deviantart (As an aside, it occurs to me that the team may have been trying to make Bug into a more formidable type throughout Generation II’s development. Giving Scyther and Pinsir an evolution, making a high stat Bug-type like Heracross, and giving them additional typings like Fighting or Steel all seem a good way to make bugs worth using). Anyway, Forretress is weird. Our first evidence of Forretress appears on July 20th. But not where you'd expect: for some reason, there are some Forretress fingerprints on Pokedex slot #246, the slot set aside for Larvitar! On July 20th, the slot that will become Larvitar was in a strange position. We don’t know what sprites it was using, but it was likely using the same sprites as it was in June: an early Cyndaquil sprite and a Rinrin backsprite, both used as placeholders. It was probably also still using a Green/Purple palette, one that became Larvitar's palette but obviously doesn't fit the sprites that were there. It still has the name “Igaru,” which, as we’ve said before, is a bit of a mystery: is Igaru an early name for Larvitar, a name for the early Cyndaquil before it became Cyndaquil, or something else entirely? Lastly, “Igaru” evolves at level 28 into the Pokemon in slot 247 (which is probably a placeholder for Pupitar) and then that Pokemon evolves in Tyranitar at level 43. So #246 was just an early slot meant for Larvitar, right? Probably, but something’s weird. First, at this stage, “Igaru” is known as the “bagworm” Pokemon, the identification both Pineco and Forretress have in the final version. Now, if this were the only oddity, I’d just say that this identification was a mistake, but that’s not the only thing. You see, we also have an early Pokedex entries for Larvitar dating back to July 20th. Here it is: Gold: Its whole body is covered with a hard, steel-like shell. No one knows what it looks like on the inside. Uhh…that’s not Larvitar, that’s Forretress. About two weeks before Forretress should exist: Pineco doesn't even exist yet! What the heck is going on? To add to the mystery, this isn’t the Pokedex slot that Forretress otherwise uses. In Spaceworld 1999, Forretress replaces Plux in Pokedex slot 213. This is a natural space for it, as it shares its Gold/Silver palette with Plux, and both Forretress and Plux were Bug/Steel Pokemon. In the final games, Forretress was found in slot #205, right after Pineco. Neither of those are anywhere close to #246. This is strange! Here’s what I think. First, my guess is that Forretress and Pineco started out with the exact same concept of a hard, steel-shelled Pokemon. Remember that I mentioned how in the earliest Pokedex data, Forretress and Pineco’s Pokedex entries were switched? I think this is because Pineco, before it was a two-stage Pokemon, was initially conceived of as the Steel-type bagworm Pokemon. Once Pineco developed, they decided it worked better as a Grass pine cone, then as a Bug. And once Pineco diverged enough from its initial concept, the team went back to the drawing board and created an evolution to Pineco nearer to that initial concept. What I’m saying is that this Pokedex entry—again, when Pineco formally first appears in the data ten days later, it has this same Pokedex entry—was probably meant for Pineco, not Forretress originally. Thus, this is technically early evidence of Pineco, not Forretress, though they may have been the same thing in early stages. But why does this Pokedex entry appear in slot 246? The obvious reason is that the team had decided to make Pinecorretress by July 20th, and they had decided the most disposable slot in the Pokedex was Tyranitar's first stage. After all, they still didn't have a design for Tyranitar-1, and did it really need three stages and not two? However, when the team abruptly decided to change Girafarig from a three stage family into a two-stage one on July 30th, it inadvertently opened up an empty slot in the Pokedex. Since there was no more need to find a place for Pineco by deleting Tyranitar-1, the team moved their initial ideas for Pineco over to Pokedex slot #204, allowing another designer to actually develop an idea for Larvitar and preserving Tyranitar as a three-stage line. Strange to think that we almost had a two stage Tyranitar line and a three stage Girafarig family in the final game! I know this is all a little convoluted, but I’m pretty confident that this is what happened. At least at first, the team initially considered Pineco to be a single-stage Pokemon: its stats were initially much higher and it didn't have evolutionary data. I'll bet ongoing misgivings about Plux made the team decide to remake Plux into an evolution for Pineco. This seems to have happened a little bit later. While we know that Pineco got its own name by August 1st and stats by 8/14, Forretress was still only a glimmer in the developers’ eyes by Spaceworld 1999, on August 17th. Pokedex slot #213 still had Plux’s name, stats, and moveset (the typing and palette were also still those of Plux, but Forretress ended up inheriting them). The only indication that Plux was going away was that its front sprite had been replaced by a very sketchy (and very angry!) approximation of what Forretress would become. (Tangentially, this is the argument for putting Forretress later, in Period 4d, which I almost did. The ordering comes down to whether you think the evidence indicates Forretress was going to replace Larvitar on July 20th, whether you think post-Spaceworld 99 is when they actually made Forretress, and whether you buy the idea that Pineco and Forretress were originally one and the same). In between Spaceworld ’99 and the final, there were a few attempts at Forretress’s sprite, luckily captured for us on the scratchpads. Here they are, properly colored and lined up next to its final sprite: Honestly, I like these sprites a lot! They have a lot more personality than the final Forretress; I think giving the eyes pupils like that makes them more expressive and interesting. So I kind of wish the team had gone with the earlier designs, even though the final eyes give it more of a Bug-like look to it. As you can imagine, Forretress' sprites were probably finished at the very last moment: after all, the team only had a sketch by late August, just two months before release! I’m sure Sugimori wasn’t particularly happy with them. I mean, forget about the front sprites for a second: those backsprites are some of the worst sprites in all of Pokemon! As a result, you can see a very distinct difference in the Crystal front sprites (though the backsprites aren’t really fixed). In the first place, the team got rid of the old Plux palette. Given how that dark red palette fits so oddly for the Steel-type Forretress in the first place, I’d bet that it had that palette simply because they ran out of time to design a sprite with a better palette. Of course in Crystal version one of the first things Sugimori did was make its shell more metallic and give it the palette Forretress deserved. Secondly, the shell itself has gotten significant revisions. No longer does it look like a walnut or even like a natural wooden thing; instead, Sugimori’s replaced its wooden, winding and twisted exterior with uniform silver bumps. You’d be forgiven for not noticing Gold and Silver Forretress’s Steel-typing, but the Crystal version heavily emphasizes this part of its design. Overall, these are very good changes. That’s it for Forretress. Next up, another mystery dude trapped inside a hard casing. It’s Pupitar! ID X31: Pupitar Tyranitar was one of the first ideas implemented when Masuda took over control of development in early 1999. By repurposing an unrelated design Sugimori had made for a magazine cover, the team came up with an inverted mirror to Dragonite: a terrible, mean looking Dragon Pokemon ready to mess up the opponent. It was obviously a homerun: Tyranitar ended up being a great design. But to make the parallel with Dragonite work, the team couldn’t just have Tyranitar in isolation: it would need two pre-evolutions that would grow into it. As was common in the reboot period of Gen II’s development, the person in charge of the Pokemon roster (either Masuda or Sugimori) set aside two slots in the Pokedex directly before Tyranitar for its pre-evolutions. And then…they sat there. For months, the team just left those slots alone. Both Larvitar and Pupitar got their names—and thus were probably first added in any meaningful way—on August 1st, during the “Big Overhaul” of Period 4c. Tyranitar got a name change here as well. Originally Tyranitar was named “Girasu” which Bulbapedia suggests probably means something like “Hated Dinosaur;” when Pupitar and Larvitar got created, the team decided to use this suffix for all three to link them together. Thus, the first evolution became Yogirasu (baby dinosaur), the second evolution became Sangirasu (pupal dinosaur), and Tyranitar become Bangirasu (maybe barbarian dinosaur). The data on Cutting Room Floor is unclear, but it appears that both Larvitar and Pupitar got their final stats during the Big Overhaul as well, which were just calculated by taking away stats from Tyranitar. They got their final moveset (minus Rock Slide, which was added later) on August 16th, right before the Spaceworld ’99 build. Notably, Tyranitar had gotten its moveset about a month earlier on July 20th, which suggests that the delay in Pupitar and Larvitar’s movesets was because the team hadn’t yet gotten around to populating those two yet. Maybe it’s a coincidence, but it is worth remarking on the fact that Tyranitar got its moveset on the exact same update that Larvitar got some of the traits of Forretress/Pineco. This could suggest that the act of working on Tyranitar made the team question whether it really needed two pre-evolutions. Saying all that, it appears that Pupitar may have been completed just a little bit earlier than Larvitar. Purely off of the sprites, Pupitar’s sprites were complete in Spaceworld 1999: it still needed a distinct Silver sprite, but its front and back sprites matched what the final would use. Larvitar, on the other hand, only has a goofy sketch sprite at that stage of development. Saying that, the team clearly was on a time crunch getting Pupitar’s sprite ready, as it was still using the palette it inherited from the Wolfman sprites the team was using as a placeholder. By Crystal, Sugimori completely redid Pupitar’s palette, hinting at how he’d have refined it if he’d had the time before release. Because its sprite was completed earlier, it seems like the team came up with Pupitar’s concept prior to Larvitar. And what a weird concept it is. Pupitar’s got to be one of the strangest design choices to me. While Tyranitar is a badass dinosaur Pokemon that destroys everything in its path, Pupitar…is a cocoon? Dinosaurs don’t grow from cocoons! I get that the team wanted to do something different, and Pupitar certainly is unique. And just like Dratini is just a little guy who eventually evolves into a giant monster, Pupitar has the same “he’s just a little guy” energy. So I get that part somewhat. But a dinosaur pupa? If anyone out there has a theory on why Game Freak decided on a design like Pupitar’s, I’m all ears. I will say that the design of Pupitar does seem to be in conflict with the flavor of the Tyranitar line, in ways that become obvious once we examine its Pokedex entries. Its early entries aren’t much different from its final ones, so here are the final entries, for reference: Gold: Its shell is as hard as sheet rock, and it is also very strong. Its thrashing can topple a mountain. Silver: Even sealed in its shell, it can move freely. Hard and fast, it has outstanding destructive power. First of all, Pupitar’s a little guy. There’s no way this tiny thing has “outstanding destructive power” or could “topple a mountain.” I’d definitely buy that Tyranitar can do those things, but Pupitar is literally just an angry looking Metapod. I think there’s a tension here in that Pupitar’s design is supposed to make it look unassuming, so that Tyranitar is all that more surprising when you finally get it, but the team also simultaneously wanted the player to know how destructive and powerful the Tyranitar family is. The effect is that these two things clash, leaving the player with the impression that a tiny cocoon somehow has the power of the Gods inside its shell. Second, the Silver entry makes a point of saying that Pupitar is “fast” and it “can move freely.” Yet, it’s not that fast: its 51 base speed makes Pupitar just slightly faster than Swinub, Piloswine, and Donphan, all Pokemon known for their lightning fast reflexes. On top of that, why is the Pokedex entry even boasting about Pupitar’s speed? The whole point of its design is that it’s a Pokemon of potential: by hiding in a cocoon, eventually something amazing would come out. If the designers really wanted a Pokemon that is speedy and destructive, why make Pupitar that Pokemon? It just strikes me that the idea of Pupitar and the sprites of Pupitar diverge a bit more than is probably reasonable. I feel as though if the team had a little more time, they may have amended one or the other, to makes its concept more coherent as a singular design. That’s not to say I dislike Pupitar. Like I said, it’s unique and imaginative, something that should count for a lot in Pokemon. I think it’s more fair to say I’m perplexed by Pupitar. Why this design? That’s all I want to know. ID X32: Larvitar Of course, let’s also address Larvitar, the first stage baby counterpart to Tyranitar. As we’ve discussed above, Larvitar needed to exist: since the very beginning of the reboot, Larvitar was penciled into the Pokedex as a first-evolution of Tyranitar. However, despite being preordained, Larvitar took a long time for Game Freak to actually figure out. So what was going on with this guy? Before we even get to Larvitar, we first need to deal with a mystery surrounding the sprites that predate it, because they don’t make a lot of sense. Before Larvitar was anything, the spot was using a green/purple palette, the early Cyndaquil sprite from the Korean Index as a front sprite, and Rinrin’s old backsprite. To top it all off, the slot had the name “Igaru.” None of these things make much sense. First, we’ve talked about the significance of the name “Igaru” before when discussing Cyndaquil, because it’s reasonable that maybe “Igaru” was the name that Cyndaquil had prior to becoming a starter Pokemon. That just might be possible, since “Igaru” is found so near the end of the Pokedex. Back in Spaceworld ’97, we noted that the last few slots of that Pokedex were taken up by strange Pokemon like Togepi, Sneasel, Aipom, and Snubbull. These didn’t fit at the end of the Pokedex: they weren't legendary, they weren't rare, etc. The explanation I gave (and I’m pretty confident in) was that these Pokemon were probably added to the very end of the ‘dex as a kind of last-minute decision, and the rest of the Pokedex hadn’t yet been reordered to fit them into earlier slots where they’d make more sense. It’s possible that something like this happened at the very beginning of the reboot: the team added in Igaru and possibly a few others to the very end of the Pokedex, and then eventually reordered the Pokedex to fit them in more logical places. This early Cyndaquil sprite was probably added to the end of the roster at the very beginning of the reboot, as they were constructing a new roster. Then it was quickly moved to the beginning of the Pokedex once it was edited into a fire echidna, and the old sprite was left here until something else could be found to replace it. That’s possibly why the name Igaru persisted here. But as we’ve said before, “Igaru” does not have any obvious meaning in Japanese: commenters have suggested it means spines and some people have argued it was a transliteration of “Igloo” but none of those explanations are entirely plausible. So maybe the name “Igaru” found in slot #246 relates to early Cyndaquil, but without knowing what the name was supposed to mean there’s no proof that they go together. Beyond that, even if “Igaru” was originally in this slot briefly, there are some other oddities. The green/purple palette, for instance, was clearly not designed for “Igaru;” that sprite looks awful with it. I’d almost guess it was the palette for Rinrin, but the green doesn’t fit Rinrin’s backsprite at all, so those are probably not related. And it isn’t the same palette as Tyranitar, so it wasn’t just put in this spot because the team knew a Tyranitar pre-evolution would eventually fill it. Larvitar does end up using this palette, but the palette exists from at least mid-June 1999, if not earlier, far before they knew what Larvitar would look like. The question is obvious: where did the palette come from? There’s no obvious answer here. It could be that something even earlier than Igaru held the purple/green palette and Igaru inherited it, just as Larvitar later inherited it from Igaru. But going through the old SW97 Pokemon, I’m at a loss to guess any Pokemon which would work with the palette. As far as I’m concerned, it’s a mystery. Furthermore, if this slot was used for Igaru, why does it have Rinrin’s backsprite and not the backsprite it has on the Korean Index? Maybe the team was lazy and just imported the front sprite; after all, if they knew they were going to redo Igaru into Cyndaquil, maybe they knew they were going to remake the backsprite and decided to start from scratch. But even if this was the case, it’s weird why Rinrin was even in this space in the Pokedex at all. Its evolution, Berunrun, had filled slot 206 prior to Dunsparce taking that slot and so logic would dictate that Rinrin was probably around there. Rinrin could have been used previous to Igaru as a random placeholder sprite, but again, if that were the case, where did the palette come from? So I don’t really know what to make of these placeholder sprites. There’s something weird going on with them and they don’t entirely add up. The palette, especially, seems strange to me. On to Larvitar though. Larvitar makes its first formal entry into the development data on July 30th, as part of the Big Overhaul of Period 4c. It gets its name as well as stats on July 30th, though its moveset comes a bit later. Saying that, as I mentioned in the Forretress entry, it did become “The bagworm” Pokemon ten days earlier and got a Pokedex entry clearly meant for Forretress (or Pineco), which strongly suggests that the team was considering amending Tyranitar to a two stage line. It was only saved because Girafarig-3 got deleted, which opened up another slot for Pineco to be built in, and preserved the Larvitar spot in the Pokedex for Tyranitar’s baby. Which is exactly what Larvitar was meant to be: a baby. Unlike Pupitar’s perplexing presence, it is pretty clear that the underlying idea for Larvitar was simply “baby Tyranitar.” You can see this everywhere: “Yogirasu” could be short for “yooji,” (infant) or "yousei" (larva) in Japanese. After mistakenly being called the bagworm Pokemon, Larvitar was changed to the “baby” Pokemon before the final “Rock Skin” Pokemon. Furthermore, one of Larvitar’s earliest sprites (seen on the scratchpads) was just an amalgamation of different parts of Tyranitar’s sprite, squished together to make it look smaller and like a child version. Even the sketchy Larvitar sprite that has been made almost into a meme in the Pokemon community gives off “tyranitar, but smaller and goofier” vibes. As easy as this concept was, either the team didn’t make it a priority or it proved a tough nut to crack; as we’ve discussed, Larvitar only had the sketchiest of sprites by Spaceworld ’97, and even after that the designer felt the need to make a model of it out of the Tyranitar sprite to get a sense how it would look. But given how late in the day the final sprite was drawn, I’m impressed by how good it turned out: This is clearly a sprite done by Sugimori, as it shows his skill with the craft of sprite art. You can also tell it’s his because Larvitar underwent no changes in Crystal version; usually, it was Sugimori making those changes and the Pokemon which underwent the least changes were usually his own (since he got them right the first time). That’s not definitive, but it’s evidence to suppose that Sugimori was the man behind Larvitar.
What I like best about the sprite is how Sugimori captured the mood of Tyranitar so well. Larvitar is grumpy, with a mad “I don’t care about you or your family” sort of energy. The sprite conveys very neatly that although Larvitar doesn’t have the sheer “destructive power” of Tyranitar, and may not be filled with the same amount of rage, it still clearly has the same outlook on life. If Larvitar could level a mountain, it definitely would, just to mess with you. We’ve discussed Larvitar obliquely in so many other entries that there’s not much more new to say here. But one thing we haven’t covered is Larvitar’s typing. I’ve already mentioned that I think Tyranitar should be a Dragon Pokemon, and clearly it’s not. But the Dark-type, at the very least, does fit it well. It’s strange, then, that Larvitar and Pupitar aren’t Dark-type, but Rock/Ground! I’ve never really understood the flavor behind making Rock-types also Ground (besides the fact that Rock was a late addition to Generation I) so I’m not really sure what Ground-type adds to these guys, flavorwise. But it is really odd that Tyranitar loses the Ground-typing to gain Dark, rather than just make the entire line the same typing. My guess is that the team wanted to make a drastic statement with Tyranitar and show that it’s somehow special. Since Dark is so rare in Generation II, gaining Dark at the end of its evolutionary line certainly makes Tyranitar distinctive. But if it were me, I think making the whole line Dark-type would have made it stronger overall. It's weird, for instance, that Pupitar gets STAB with Earthquake but Tyranitar loses that, especially when Tyranitar is supposed to be the strongest of the three. Thinking about it, are there any other Pokemon that replace their second type like this when they evolve (not just add a type, but lose one to gain another)? To sum up, I’m surprised that Larvitar turned out as good as it did: it’s a good sprite, it’s still a very popular Pokemon even nowadays, and it works really well as part of the evolutionary family. While “baby Tyranitar” could have turned out a very mediocre concept, Larvitar managed to make it work quite well.
8 Comments
|