Counterparts: Houndour & Houndoom versus Wolfman and Warwolf(Houndour and Houndoom are official art; Wolfman and Warwolf are courtesy of @Raciebeep) The next set of four Pokemon in Period 2b are maybe the ones most obviously designed to be counterparts. The Houdour line and the Wolfman line have similarities that makes it seem as though they were designed together: both are canines, both families have contrasting types (Fire vs Ice), and both have an "evil" sensibility: Houndour and Houndoom are clearly devil dogs, while Wolfman and Warwolf are monstrous, werewolf designs. It's even possible that they might even be wearing the fur of some other Pokemon they killed. Spooky! As an aside, given that Spinarak and Ariados have creepy faces on them and Ariados had a Jack O’Lantern design at one point, it's possible that the three families (Houndour, Wolfman, and Spinarak) were designed as a Halloween trio. Both the families made it into Spaceworld ’97 and that fact, combined with their contrasting typing, suggests to me that these were not counterparts competing for the same Pokedex slot, but instead mirrored counterparts meant to contrast each other, like a Scyther and Pinsir duo. However, by the final, Warwolf and Wolfman were removed, while Houndour and Houndoom made it (though in a very minor role as an almost exclusively post-game Pokemon). Whatever connections the team wanted to make originally with these two, those connections were gone by time the games were finished. ID 388: Deruberu (Devil or Delvil) (Houndour) First, lets take a look at Houndour. This mean mutt was originally named Debiru, which translates pretty directly to Devil (its final name, Deruberu, is just a slight tweak of that). Its motif is pretty clearly based off the idea of a Hellhound. It also has a skeleton motif, and its evolution, Houndoom, accentuates some of its traits to look even more like a devil. Houndour’s interesting because it went through a lot of minor changes over time. For whatever reason, how, it seems that the team wasn't quite sure what to do with this guy. He made it into the final, but barely any trainers have him, and he's only found in one particular place in the post-game. (Two other takes on the Hellhound, the left being a depiction of Cerberus, and the right from Magic: The Gathering) When he was first designed, Houndour was only a Fire-type Pokemon, not the Dark/Fire that it became later. This is strange, given that it and Houndoom's designs are clearly based on ideas that would fit the flavor of Dark types snuggly. It's unclear why it didn't start out as a Dark-Type, but it seems that by Spaceworld ’97, the designers already realized their mistake. Though in the Demo its still pure fire, there's an NPC in Silent Hills who introduces Houndoom by saying, “This is HOUNDOOM! It's a POKéMON of a brand new type!” Given what we know of the future, the NPC must already be referring to Houndoom as a Dark-type even though it wasn't true yet. It’s especially odd that Houndour is a pure Fire-type as late as Spaceworld '97. We know from the existence of Rinrin earlier in the Korean Index (ID #365) that the Dark-type had already been created by the time Houndour was made. Rinrin could have been made out of order and overwrote something in that slot; that's a possibility since Rinrin looks like a Nishida design and most of Pokemon surrounding it are Sugimori designs. It could also be the case that Rinrin and Girafarig weren’t originally designed as Dark Pokemon but the typing was added later, before Spaceworld ’97, but also before the team got around to changing Houndour’s typing. But both of those explanations seem like stretches to me. My guess is that Houndour was a vaguely late addition to the Pokedex and that they hadn’t paid much attention to it before they included it in the demo. The team knew it wouldn't be seen in the demo, so it probably wasn't a huge priority to update its typing in the internal data. This theory's not perfect: the fact that an NPC states to the player that Houndoom was supposed to be Dark-type seems like evidence that they had planned the hounds to be Dark-type much earlier in development, so it's still odd that it was entered into the data as a pure Fire-Type at all. On the other hand, there are indications that Houndour was a pretty rough draft by the time of Spaceworld 97, lending credence to the idea that the team saw it as a low priority. First, take a close look at the SW97 sprites of Houndour and Houndoom. They're really rough. The proportions are all over the place, the mouths are cartoony, and there’s very little perspective in either sprite. They looks like some of the worst Pokemon Green sprites, and honestly a lot like the sketchier sprites in the Korean Index of Pokemon that didn’t even make it to SW97. Honestly, if we didn’t know what Houndour and Houndoom would eventually become, I think these sprites wouldn't stand out as good designs in the Korean Index; rather, they'd probably look a lot like the other rough designs that were removed early on. All of this is to say, I don’t think Houndour was originally sprited by any of the main four designers: it likely was drawn by someone that did minor spritework on Generation I, or maybe by the same mystery designer who gave us a slew of unused designs in Periods 1b, 1e and 2d. Regardless, if this was created by a minor designer, it’s very possible the creator was out of the loop concerning other changes the team was planning for Generation II, including the inclusion of new types like Dark. What I’m trying to say is that Houndour could have been created as a pure Fire Pokemon by someone who didn’t know the Dark-type was available to design for, and it was added into the game under this original idea, before one of the other designers realized it could be a perfect candidate for the Dark type. This is backed up by looking at Houndour’s early moveset, which is notable for what it lacks: Dark moves. By SW’97, Rinrin had Faint Attack, a Dark move (Girafarig didn’t really have a moveset at that point), so there’s really no excuse for Houndour to lack Dark moves, unless it was still working from an outdated moveset made before the decision had been made to change its typing. Houndour does have Bite, but that wasn’t a Dark move yet in Spaceworld ’97; otherwise, it’s moveset pretty much follows what you’d expect from a pure-fire Pokemon: In fact, compare this moveset to Growlithe’s. Another hint that Houndour might have been only minimally worked on by SW97 is how similar these movesets are, suggesting that Houndour’s might have been tweaked from Growlithe’s: It's not perfect: a lot of the similarities could be chalked up to both having staples of Fire-type Pokemon; still these movesets are much more similar than they are to the SW97 movesets of Vulpix, Ponyta, or Honooguma. The main differences between the two are that Growlithe learns its moves a little more slowly than Houndour, and that Houndour has Bonemerang, a move that clearly references the skeleton parts of the dog’s design, while Growlithe has Sacred Fire, a new fire move created for Gen II and given only to Growlithe, Arcanine, Hooh, and Moltres. Another reason why Houndour might have been shifted to a Dark/Fire type was because of this similarity: Generation II really didn’t have room for two fire type doggo Pokemon. After Spaceworld ’97, Houndour and Houndoom got more attention than they had previously. The sprites they started with, which were so rough at first, were touched up multiple times, eventually adding shading, perspective, and more proportionate features. These sprites are more or less in the same poses and give off the same basic personality: my guess is that Sugimori took the sprites from Spaceworld and personally tweaked them, bringing them more in line with the quality of the rest of the sprites. The biggest difference between the early sprites and the final versions is that the Skeleton/Bone theme is subtly downplayed. First, Sugimori removed one rib bone from the back of Houndour, making the bones on its back look less obviously like a ribcage. Then, right before the final, the backsprite lost the spinal column section, again moving it away from a skeleton design. The effect of this on the final is to make Houndour’s inspiration a little less obvious in its design, but it also to make it not quite as dark. This is a continuing theme with the final sprites in Gold: Kyonpan and Norowara were removed, likely for being too scary, Girafarig lost its head for maybe the same reason, and Remoraid and Octillery lost their most obvious design elements in order to make them less objectional. It’s a shame so many sprites got neutered (or removed), but this seems to have been the goal of the team by the end of development. Notably, Houndour (and Houndoom)’s sprites also got significantly revised for Crystal. It seems that, even after launch, Sugimori still wasn’t happy with Houndour’s pose, so it got a slightly different one, one that collectively makes it look a little less flat. Houndoom’s sprite got a complete overhaul, probably based loosely on its original silver sprite. Most importantly, they completely changed their colors: rather than being fiery dogs, both Houndour and Houndoom because purple and black, further making them look more like Dark types and less like Fire types. A Pokemon that started out completely like a rehash of Growlithe by the end of design became significantly more unique. The lesson of Houndour is that sometimes, polish is what makes perfect. The original Houndour and the final Houndour aren’t that much different from each other, but the collective total of the type change, the moveset change, and the sprite revisions made a much stronger Pokemon. A complete overhaul isn’t always needed; sometimes a design gets better with just a few changes. ID 389: HoundoomHoundoom and Houndour were clearly designed in concert. The skeleton motif of Houndour is further accentuated in its adult form, and smaller details, like its open, angry mouth, now lead to a payoff, like Houndoom breathing fire. As a result, most of the things you can see in Houndour’s development are echoed in its evolution, Houndoom. That doesn’t mean there isn’t anything left to talk about, however. Houndoom veers a little bit away from the skeleton design to instead embrace the properties of a pop culture devil. It's lost it's skull helmet, but gained devil horns and a pointed tail. But similar to Houndour, you can also see how it’s first sprite is extremely rough, and probably made by the same designer. Sugimori clearly edited the sprite for Houndoom just like he did for Houndour: not drastically redesigning it, and keeping mostly the same pose, but increasing the quality of the sprite, by adding shading, better perspective, and a cleaned up face. Sugimori even added a more devil-like tail to Houndoom, a rare case of the design inspiration being strengthened in the final Gold/Silver sprites and not being deemphasized instead. Also Like Houndour, Houndoom also started as a pure Fire-type Pokemon, though it was clearly meant to be Dark-type by the time Spaceworld ’97 was finished. Although the final sprite actually gained more Devil-like aspects, you can still see how Houndoom's concept was softened by the final if you glance at the Pokedex entries of Houndour and Houndoom. For instance, here’s Houndoom’s Spaceworld ’99 Pokedex entry: “The living embodiment of aggressive instincts, it will not stop fighting even after its opponent has been ripped apart.” “Ripped Apart” is…quite graphic for a Pokemon game, and it makes sense why it was changed. In general, the designers have stated that they don't want to paint any particular Pokemon as purely good or evil, since all could all be your friend. This idea that Houndoom is naturally aggressive contradicts that ideal. So by the final, Houndoom was given Houndour’s draft Pokedex entry, and Houndour was given a much more peaceful one, like so: Houndour (Draft), Houndoom (Final): “If you get burned by the fire it spews from its mouth, the pain from the wound will never go away, no matter how much time passes.” Houndour (Final): “It uses different kinds of cries for when it's contacting others in its pack than when it's chasing down prey.” Overall these are both improvements, and don’t paint either in an evil or malevolent light. The main thing I want to talk about concerning Houndoom is how virtually unused it and Houndour are in the final games. Obviously, the only way to get Houndoom is to evolve Houndour, but Houndour is almost entirely absent. It can only be found on one route in the Kanto section of the map, deep into the post-game and far after where this Pokemon would be useful. In addition, both Houndour and Houndoom are almost entirely unused by trainers as well! In Spaceworld ’99, there was only one appearance of Houndoom in the entire game, in Karen’s elite four team. In the final, one of the Rocket Executive’s was given a single Houndour and a single Houndoom, upping their total appearances to three in the entire game. Why create an entire Pokemon, just to leave it as almost unused content? I’d get it if Houndoom was a legendary or gimmicky Pokemon, but it’s really not. It fits the same niche as Growlithe does in terms of party composition, and Dark-type Pokemon are overall pretty rare in the games. Outside of Houndour and Houndoom, the only options a player has for Dark types before the Elite Four are Murkrow and Umbreon (Crystal added Sneasel to the Ice Path, but originally it was also post-game). Houndour would have added some needed type diversity to the early-to-mid game, making it even more of a headscratcher for why it wasn’t included. Sadly, this is weirdly the case with a lot of the unique and interesting Pokemon developed for Generation II. Misdreavus, the only new Ghost type created for the generation, is only found in the final, post-game dungeon. Not even Morty, the Ghost-type gym leader, owns one! Slugma and Magcargo are also only found in Kanto after the Elite Four, the Larvitar family is only found in the final dungeon, and Pichu can only be attained after the Elite Four, since Pikachus are only found in Kanto. It’s a really weird choice made by the development team, one that continues to bewilder me. Why hide your hard work somewhere where no one would see it? I have a few explanations for this, all centering on the hurried end of development. Misdreavus wasn't finalized until just before the final, having previously been Norowara. It seems pretty clear that it was held off until the end because it was finalized very late in development, and the team may have been mostly or completely done with the main storyline before Misdreavus was finished. Gold and Silver were incredibly rushed, despite having such a long development time (especially at the end of development), so the team may have decided to use the Kanto portion of the game to stick Pokemon that they finished late, just so they didn’t have to rebalance the rest of the game as they put the finishing touches on new Pokemon. Thus, all these Pokemon could have been ones that were considered to be incomplete or on the chopping block, before the team added them at the last second. The explanation fits perfectly for Misdreavus and Larvitar, since they weren’t done until late. It fits Sneasel too, since we know that Sneasel’s design was only changed at the very last month before the game’s release (though your rival, Silver, had Sneasel as his signature Pokemon even before the design was changed). It kind of fits Slugma too. While Magcargo, its evolution, can be found used by many trainers (notably it’s one of Blaine’s signature Pokemon in the Kanto section), Slugma is only wielded by one trainer in the game. We know that Slugma wasn’t properly implemented by the time of the Spaceworld ’99 demo, and so the team may have held back the line because it still needed work. Instead, they used Magcargo, since it was pretty much finalized already, with Slugma only being brought in at the last second to give this Pokémon a pre-evolution. (These Pokemon's appearances just months before the game was released. Slugma and Sneasel's sprites were substantially changed and Larvitar's is only a sketch. Slugma is just named "Pending 10" at this point; Misdreavus' sprites are finalized, but its name is still "Norowara" and still retains the stats for Norowara and Norowara's typing, Ghost/Dark) Houndour and Houndoom are outliers in this pattern: both of them have complete sprites by Spaceworld 1999; they've been in the games since Spaceworld 97, unlike most of these. So if the problem wasn’t that they were unfinished or not ready for use in the game, what was it? It's possible Houndour and Houndoom were almost cut; maybe the team made minimal use of them because they were always expecting Houndour and Houndoom to be replaced for some new design. There were a few other early designs that made it this far but not much further, such as the unused Pinsir and Weepinbell evolutions. Only, because the team was on such a tight schedule, no one ever had the time to cut Houndour or Houndoom and replace them with something else. It might have been dumb luck that we got these hounds and not the monstrosity that was Tsubomitto: Conversely, maybe the team couldn't be bothered to find a habitat in Johto where Houndour fit, and so they dumped it into the end game rather than heavily reworked any of the routes. The only other explanation I can think of is that the team just wasn’t confident in the Dark type and worried it might be overpowered. Misdreavus, until just before release, was Ghost/Dark, and Sneasel was also Dark, so those plus Houndour could have been held until the post-game to avoid them messing up the balance of the game. They may have also have been playing around with the type effectiveness of the Dark type until late in development, and they may have limited Dark types in the main game so that the balance of the games didn’t drastically swing as they changed what they were effective against. Murkrow was never a threat to balance, since it’s a single stage Pokemon with middling stats; Umbreon, the only other choice for a Dark type, couldn’t really be limited to the post-game without limiting Eevee too, and was poison type for a large part of its development anyway. Thus, Houndour and Houndoom may not have been on their way out: they may have been dangerous for the overall balance until the very last polishing of the games. It’s a shame that Houndour and Houndoom as so sidelined, as they’re honestly very clever designs, and fit well into the Pokemon world. Houndoom’s a good idea, even if it seems like the designers didn’t have a lot of confidence in him. ID 390: Urufuman (Wolfman) Urufuman, better translated as “Wolfman,” is the first in a two stage Ice evolutionary line. Both Wolfman and its evolution, “Warwolf,” appeared in Spaceworld ’97, and there’s some evidence that they survived in some form at least through 1998 and maybe into January or February of 1999. After that, they’re gone, though interestingly, their sprites were used as placeholders for newly designed Pokemon. Honestly, these guys are the best. I love them, and while I also like Swinub and Piloswine—which probably didn’t replace so much as filled their niche—I really wish these guys had made it into the final game. Wolfman and Warwolf are obviously based on Werewolves, the mythological creature that transforms into a giant wolf during the full moon. I also think there’s a good chance that these two were inspired by the old saying “a wolf in sheep’s clothing,” meaning someone who appears harmless but is actually dangerous under their exterior. Part of me wonders if Wolfman is the reverse: Under that wolf hood, is there a mareep hiding inside? While these two guys didn’t quite make it to the final, many years later, in Generation VII, Pokemon did finally get it's own werewolf: Lycanroc. Let’s talk about the connection between the Houndour/Houndoom and Urufuman/Warurufuu lines. There’s a lot of small things about both Pokemon families that makes it pretty convincing to me they were supposed to be parallels to each, in the same vein as Caterpie/Weedle or Scyther/Pinsir. The first thing to note is how close they are together in the Korean Index; more interesting, however, is that they're also right next to each other in the Spaceworld '97 Pokedex: Houndour and Houndoom are #235 and #236, while Wolfman and Warwolf are #237 and #238. It’s possible because the two families were dumped into the demo at the same time, suggesting that the designers saw them as parallels. I also think it’s significant and suspicious how late they are in the SW97 Pokedex: everything after them is either a legendary creature, an evolution of a Gen I Pokemon, or miscellaneous Pokemon clearly thrown in at the end of the list at the last minute. While Houndour and Wolfman don’t appear to be added into SW97 as late as, say, Togepi or Snubbull, they do seem to have been the last Pokemon added to the Pokedex as part of the standard list, before the team focused on filling in holes and adding in legendaries. The Wolfman line and the Houndour line don’t have the same sprite creator, but they have a very similar Halloween vibe: Houndour is more or less wearing a skeleton costume and even has a skull mask over its head, while Wolfman is literally a mystery Pokemon hiding in a wolf costume. Both take inspiration from common kid’s Halloween costumes, or monsters viewers of 50s and 60s hammer horrors would be familiar with. Their types are also opposite, Houndour being Fire (or Dark/Fire) and Wolfman being Ice; they also evolve at the same level, the oddly high level 35 (Houndour’s evolutionary level would be lowered in the final game to 24). And while their movesets aren’t exactly mirrors, there are similarities between them: While it’s not uncommon for Pokemon in SW’97 to learn moves every seven levels, it is an interesting parallel that both have the same growth pacing. As well, Wolfman has more physical moves than Houndour, and it has Safeguard to protect itself while Houndour lacks any such defensive moves, but both have a fair share of debuffs, and they both learn a weak elemental move (Ember, Powder Snow) and the most powerful one (Flamethrower, Blizzard) at very similar levels. Again, I wouldn’t say these movesets are mirrors of each other by any means, but I do think the Pokemon would function similarly in a battle. As an aside, its interesting that Wolfman knows Fury Swipes, since up until this point, this was a Meowth specific move. In SW97, Rinrin got it too, but it’s a move closely associated with cats. Odd why a wolf would get a cat move. Because of Wolfman’s flavor—a werewolf that may or may not have eaten another Pokemon and used its skin for a hoodie—and its similarities to Houndour, I can’t help but wonder if the Wolfman line was also meant to be Dark type, but that we don’t have a build in which they were changed to Dark. It would make sense, and strengthen the parallels between the two lines. But unfortunately we just don’t have evidence one way or another, except for the existing parallels and that fact that Houndour was mistakenly single-typed, meaning it could have happened twice. The interesting thing about Wolfman and Houndour, of course, is that while both of them are counterparts like the rest of Period 2b, these are the only case in which both counterparts showed up in Spaceworld ’97. So what gives? Like I’ve said at the beginning of this article, it seems like the team thought of these more as a contrasting mirror to each other, especially because Ice and Fire types fill a different niche in the ecosystem of the games. But I wonder if that’s not initially how they were conceived. Like the other counterparts in Period 2b, Houndour and Wolfman were clearly made by different designers, and so the initial directive could have been simply to design an elemental (or Dark type) canine. Once they had two drastically different designs, the team could have decided they worked better as a contrast to each other, or they could have simply decided they liked both designs equally. Given how rough Houndour’s sprite is, I wouldn’t be surprised if the team had initially chosen Wolfman to fit this spot in the Pokedex, and Houndour was the one that was a last minute addition to SW97. Another reason to think that Houndour was the last minute addition: Wolfman and Warwolf two of the very few Spaceworld '97 exclusive Pokemon that have encounter data in Spaceworld '97! Unlike Houndour, which doesn't appear at all in the encounter data, a player could fight and catch Wolfman in the Blue Forest area, a northern snowy town. They'd mostly be encountered at night, with levels in the mid-30s. So the team had clearly put enough love and attention into Wolfman such that it was already integrated into the game. Which is ironic, because of course Houndour and Houndoom made it to the final, but Wolfman did not. Although Wolfman’s name or stats do not appear in any of the data we have after the Korean Index (in May 1998), unlike many of the other deleted Pokemon, Wolfman and Warwolf’s front sprites are still in use by June 1999, as placeholder graphics for Pokemon that didn’t have sprites; in the case of Wolfman, it’s sprite is being used as a stand-in for Pupitar. Pupitar’s whole family is substantially unfinished at that point, and an early sprite of Cyndaquil and Rinrin are being used as the front and back sprites of Larvitar at the same time. But this stand-in Wolfman sprite is interesting because it’s touched up and uses a new palette compared to the SW97 one: Which means that Wolfman made it at least far enough for the team to redo its sprite for the Gameboy Color, even if it didn’t make it much farther than that. It’s a shame; I wonder what could have been with this adorable ‘mon? UPDATE: As a number of commenters have pointed out to me, we might know what became of Wolfman! It could have become Snorunt, in Generation III! It's easy to see the similarities: they have the same little black stubs for hands and feet, they're both wearing something, and their faces are both hidden, with beady eyes visible. Though they have very different theming--Snorunt is based on a Japanese myth about a mischievous spirit--I'm honestly pretty sold that Snorunt may have been a rework of the idea that led to Wolfman. The redesign also follows the main trajectory for a lot of discarded Generation II designs. Snorunt is a significantly less scary, or gruesome, version of the concept, in that it isn't garbed in the hide of another creature, but just a cute reed coat. If the problem with Wolfman was that it was tonally dissonant from the lighthearted designs of Generation II, Snorunt feels like an obvious way to correct that problem. Obviously, another telling feature is that both are Ice type. But Snorunt's moveset also feels a bit suspicious. Comparing the movesets of Wolfman and Snorunt, it's easy to notice they don't share any moves beyond the basic ice moves that would be expected of any ice Pokemon. But look closer and there are some telling similarities. Wolfman learns Safeguard, for instance, an out-of-character defensive move that protects it from status ailments. But Snorunt learns Protect, at a very similar level (Safeguard at level 21, Protect at level 25). In addition, Snorunt has access to Bite and Crunch, which seem odd to me, given that neither Snorunt nor Glalie seem prone to biting things. But those moves would have made perfect sense on a more developed moveset for Wolfman, especially if my hunch is right that Wolfman was being considered for Dark-type. Of course, this theory isn't perfect. For instance, we know from early data from Ruby and Sapphire that Snorunt's early name was "Nozu," or a portmanteau of "Nose" and "Snow"; that name seems to have almost nothing to do with Wolfman's concept, so either Snorunt's final similarities are a coincidence and it once looked way more "Nosey" than it does now, or that name was after it diverged significantly from Wolfman. The other problem is what to do with Glalie, which, honestly, looks nothing like Wolfman or Warwolf. In fact, it looks almost nothing like Snorunt either, suggesting to me that it was a completely different concept tacked on to Snorunt much later. While its pose does have some superficial similarities to Warwolf, Glalie is either proof that the Snorunt line was developed independently from Wolfman, or just that Glalie itself was originally separate and stapled onto Snorunt closer to development. There's no way to be sure about any of this Snorunt/Wolfman speculation, but I do appreciate thinking about it. It makes me sleep easier knowing that something may have come from the cool concept of Wolfman. ID 391: Waaurufu (Warwolf) Wolfman’s evolution is Warwolf. He’s bigger, meaner, and now has some claws, making him look like a dude no one wants to mess with. Like Wolfman, his sprite was updated as they transferred Gold and Silver to the Gameboy Color in early 1999, and his sprite is even more obviously updated than Wolfman’s. Notice the new shading, and how the tufts of fur on his SW97 design are gone. And of course he’s blue, the color of Ice. Wolfman and Warwolf both disappeared from the games after SW97. There's an obvious question to ask: what happened to them? To answer that question, we need to discuss Piloswine, a Pokemon I’m not going to get to talk about in detail until much much later in this project, after we say goodbye to the Korean Index and we start talking about the Pokemon designed after 1998. Long story short though, Piloswine seems to have been created before April of 1999, possibly as early as January or February of that year, maybe even late 1998. Piloswine is a Ice/Ground Pokemon with a ton of fur on its back, hiding its face; it unarguably shares a lot of features with Warwolf. The team initially wasn’t sure what to do with Piloswine, however. It seems like they wanted it to be a two-stage or even three-stage Pokemon, but they weren’t sure whether the sprite they had was a first or second stage evolution. Swinub did not yet have a front sprite by June 1999, and there was even plans for a third stage of Piloswine, which has data relating to it as late as August 1999, when it was ultimately overwritten with Slugma. Up until that point, that third evolution was using Warwolf’s old sprite as a placeholder, even as they changed its typing to Ice/Ground and gave it stats to match Piloswine. It’s important not to make too much of this: after all, Wolfman’s sprite was at one point being used as a stand-in for Pupitar, and there’s no relationship between those two Pokemon. But there is undoubtedly some connection between the Piloswine family and the Warwolf family. The two possibilities are as follows:
My gut leads me to think (2) is more likely to be correct, due in part to how different in concept Piloswine and Wolfman are. Saying that, I think either chain of events is possible, and there’s no doubt that “Ice-type Pokemon with fur covering its face” fits both of these Pokemon equally. In either case, Piloswine probably replaced Wolfman in the Pokedex, and the team probably moved its sprite to serve as a placeholder graphic, while Warwolf’s sprite stayed in place while they worked on Piloswine’s evolution. The replacement of the Warwolf line fits the larger shift in 1999 of dialing down scary or dark designs and replacing them with cuter designs. Whatever you may think of Piloswine, it’s certainly much cuter than Wolfman. As the team tried to make the games fit a specific aesthetic, it makes sense that odder or more outlandish designs, like a mystery Pokemon wearing someone’s pelt, would go by the wayside. It’s a shame, but it’s understandable. Don’t get me wrong; I’m on Team Piloswine, and I’m on record as saying that Piloswine is one of my favorite Pokemon. But I also love Warwolf. I know, logically, that you can’t have both of them in the same game; they cover the same ground too much. But my heart screams out for Warwolf. Counterparts: Bomushikaa versus ID 392 (Bomushikaa art by @Raciebeep) The final two designs in Period 2b are the last two that I’m sure are counterparts, though there is an argument to be made that the next three designs—baby versions of Vulpix, Ponyta, and Goldeen—might have also been initially conceived as counterparts competing for one slot. However, these last two designs are complicated. Without knowing anything about ID 392, its very hard to see what similar design space Bomushikaa, the fire-seal, was sharing with an odd robot bird. Like the rest of Period 2b, the Korean Index here switches back and forth between an unused design and one that made it into Spaceworld ’97, suggesting these two are related, somehow. My hypothesis is that the brief for this slot in the Index was “Unusual typing we didn’t do in Generation I.” That’s clearly Bomushikaa’s selling point, as a Water/Fire Pokemon, and it is one of the most obvious things to do with a sequel like Gold and Silver: give people Pokemon with typings they’d never experienced. I will admit, however, that there’s not a lot of proof of this, and any guesses of ID 392’s typing have to be purely speculative. For these counterparts, I’m going to shake it up a little by discussing Bomushikaa first, even though it comes after ID 392. Because we know more about it, we can establish what’s distinctive about its design. From there, we can draw parallels and see what we can learn about my friend and yours the robot bird. ID 393: Bomushikaa (Colorization of the second sprite was done by @OrangeFrench, and is speculative) Bomushikaa is a seal ahead of its time. Created to be Pokemon’s first Water/Fire type Pokemon, Bomushikaa had a typing that wouldn’t actually appear in the games until Generation VI, years later, with Volcanion. Not only that, but Bomushikaa’s design, as a circus seal, wouldn’t be revisited until Popplio in Generation VII. Everything about Bomushikaa is unique and interesting, and yet it didn’t make it farther than Spaceworld ’97 before it was removed from the games. (I can't help but feel that Bomushikaa is a far better design than the awful, mashed together idea we got for Volcanion, though Popplio is alright) Bomushikaa is obviously based on circus seals, which frequently are depicted with balls that they spin from their nose. It’s a design that doesn’t stray from the animal inspiration very much, but does do something interesting with the ball seals balance. In Bomushikaa’s case, the ball is a fireball! Ready to explode on its enemies! Sure, Bomushikaa has precursors. Seel is, of course, a seal, though it lacks the fireball that defines Bomushikaa. Mr. Mime, likewise, was designed after a circus act, but its design draws from a completely different act, and led to it becoming a more-or-less standard Psychic Pokemon in Generation I. Bomushikaa would fit closely next to both of those, but it has a flare all of its own. Even Bomushikaa’s moveset is interesting. Take a look: Obviously, Bomushikaa is the only Pokemon until Generation VI which naturally learned both Water and Fire moves; in addition, although a lot of other Generation I returnees learn Flame Wheel, the fact that the move creates a “Ring of Fire” makes me think it was originally created as a signature for Bomushikaa, to emphasize the circus aspects of the seal. But look more closely at the movepool. Alongside the fire and water moves, Bomushikaa also learns Smog and Smokescreen. As far as I can tell, Bomushikaa doesn’t have an obvious way of making smoke (maybe throwing its fireball around?). What are those moves doing there? Bomushikaa isn't the only Fire-type Pokemon to learn these two moves, but their inclusion in the moveset suggests an interesting connection. It turns out, Bomushikaa was probably designed specifically as the signature Pokemon for a new trainer class in Generation II: the Firebreather. Firebreathers share Bomushikaa’s circus aesthetic, and in the final game, their signature Pokemon are fire-types and Koffing. Koffing, of course, had Smog as its signature in Generation I, and also knows how to use Smokescreen. So not only does Bomushikaa share a flavor link with Firebreathers, but it also has a moveset extremely reminiscent of what Firebreathers already did with their Pokemon. Bomushikaa might be the first (only?) Pokemon designed around a trainer class? There’s a chance Jugglers in Generation I might have been created to work with Mr. Mime or Voltorbs, and Dragonite might have been designed late enough in Generation I’s development to have been made specifically as Lance’s signature. There’s also the case of our Tanuki friend a few Index slots ago, which I neglected to mention at the time has an uncanny resemblance to the Generation I Burglar trainer class. But outside of those possibilities, creating a Pokemon with a trainer in mind is a unique design inspiration, and one I doubt was repeated often. Bomushikaa is also very unique because it has a post-97 sprite, despite disappearing after Spaceworld ’97. In the Korean Index, Bomushikaa is one of three Pokemon (the others being Elekid and Ho-oh, which we’ve already discussed) to have a unique sprite not found in SW’97. This is odd and not totally explainable, but the most likely possibility for it is that Bomushikaa was still being worked on after Spaceworld ’97, right up until the team’s hiatus in 1998. That means Bomushikaa may have been the very last Pokemon to get tweaks and revisions before the team restarted production in 1999 and scrapped a lot of their earlier work. I will say it’s a shame: the later sprite found in the Korean Index is very good work, and really makes Bomushikaa stand out. There are a million reasons why Bomushikaa may not have made it to 1999. First, it’s very clearly an Animal+ design, and when viewed from that perspective, it seems like an easy place the team could have cut a superfluous Pokemon. Those sorts of designs are often boring and the first on the chopping block. On the other hand, Bomushikaa could have been too weird to really find a home. It’s a single stage Pokemon without an obvious habitat; its unclear which routes you would even encounter Bomushikaa on, and single stage Pokemon are always a hard sell because they can’t grow in the endgame. It could also have been removed because of balance reasons: Water/Fire is a powerful typing combination, and its moveset is extremely good, able to take advantage of both of those types. Or, finally, it could have been abandoned because the team was interested in making new Pokemon when development restarted in 1999 and Bomushikaa just happened to be no one’s favorite. That’s about all we can glean from this lost guy. However, keep this all in mind: what we know about Bomushikaa gives us a few leads to understand what’s going on with ID 392. ID 392: ???Bomushikaa’s counterpart, at a glance, is a mystery. It’s a strange looking bird with a megaphone or a speaker for a mouth; it has weird eyes and very straight wings. There’s a lot going on in its visual design, but nothing that suggests one single inspiration or concept. The most common explanation is that 392 was a very early design for what became Chatot: rather than have a musical note on its head, 392 blares the music out of its speakerphone. Previously, I briefly flirted with the idea that 392 was an early design for Hoothoot; I still don’t really think that’s likely, but there is something about Hoothoot that’s suspicious. In Spaceworld ’97, Hoothoot’s signature move is Megaphone, a move not left in the final that lowers accuracy. Given 392’s visual design, Megaphone seems an obvious shoe-in for its signature; not for Hoothoot, which doesn’t really have anything to do with loud noises present in its design. I do think it’s possible that 392 was included originally instead of Hoothoot, and that Hoothoot’s moveset was leftover from 392. There’s absolutely no evidence for this, but this moveset certainly seems like it’d serve 392 nicely, maybe minus Hypnosis and Moonlight. At first glance, my own personal reading of this sprite was that 392 was a robot bird, inspired by loudspeakers. Partially I think this because of all the straight edges in the design, especially with the way the wings unfurl in the backsprite. The other reason I think this is because of the strange eyes. While those could certainly just be an odd design choice, they look a lot to me like spritework of a loudspeaker grill, made of metal mesh. Thus, 392 might have microphones or speakers in its eyes, and then a megaphone to broadcast what it hears.
That theory, however, doesn’t completely mesh with what we know from Bomushikaa. Obviously, I could be wrong about the counterpart nature of Bomushikaa and 392: the entire theory could be wrong, or these two weren’t very interrelated, or they were a break from the other Pokemon around them, designed around themes. All of that is possible. But take it as a postulate for now that 392 was designed to fill the same niche as Bomushikaa. If we assume that, what could we learn about 392? First, Bomushikaa implies that 392 was also designed with a unique typing, or a typing made up of opposites. That could back up my theory that 392 is a robot speaker, which would imply that it was designed to be Flying/Steel. Except that Skarmory, already designed in the index, was already Flying/Steel. That doesn’t mean it’s not possible, as two different designers could have worked on these two without consulting each other. And Skarmory’s typing could be the reason the team didn’t choose 392 to take up a slot in the Pokedex. But Steel/Flying also doesn’t feel as much like opposites as Water/Fire does. What else could it be? Another very likely candidate for 392’s typing is Flying/Ground. Like Water/Fire, these two are about as opposite types as you can possibly get, and there hadn’t been a Flying/Ground Pokemon yet, though Gligar—designed in 1999, after 392—would eventually fit that typing profile. At first, it’s hard to see how 392 could be ground type, but the more I look at it, the more it makes sense. It’s hard to tell from just a sprite, but I could imagine that 392’s megaphone shook the ground and made earthquakes, giving 392 access to attacks like Magnitude and …Earthquake. Again, there’s no evidence here, but it certainly seems plausible to me. It could also be Flying/Rock, Flying/Bug (those eyes could be compound eyes), or even Flying/Psychic, I suppose, but Flying/Steel and Flying/Ground seem like the most obvious bets. Beyond its typing, Bomushikaa also suggests that 392 could have been designed around a trainer type. Here, though, we come up more or less empty. None of the new trainers introduced in Generation II make an obvious candidate to be 392’s trainer, with the possible exception of the Police Officer. In the final game, Officers only use one Growlithe a piece, and are very boring trainers; if they had a speaker that blares out a Police siren, they’d potentially have a lot more interesting theming. Alas, the Officers don’t exist yet in SW97, so if 392 was designed around this trainer, it hadn’t been implemented, making that whole theory unlikely. That all might sound like a lot of nothing, but we do have at least something to say about 392. First, 392 pretty obviously had a signature move, one that was given to Hoothoot even though it didn’t fit him all that well. Secondly, it could be Normal/Flying and the inspiration for Chatot, but the more likely explanation to me is that 392 was either Flying/Steel or Flying/Ground. And knowing that gives us a hint at its possible moveset. While we don’t have everything, we have a decent outline of what could have been here. It’s also not much of a surprise why it was cut before Spaceworld ’97. As I’ve mentioned before, the Korean Index had way too many bird designs in it; 392 was competing with better bird designs from other places in the index. Secondly, and I think it’s pretty unambiguous: 392 is an inferior, sketchier design to Bomushikaa. Bomushikaa adds a very interesting Pokemon, which a cool design and unquie typing to the new games. A Flying/Ground Pokemon, with an ambiguous sprite, is not nearly as exciting. Anyway, that’s it for Period 2b! Next up, we’re going to talk about the shorter Period 2c, where we delve into the mystery of baby Pokemon!
14 Comments
5/19/2022 15 Comments 382 to 387: Unknown to UnknownCounterparts: ID 382 & 383 versus Spinarak and Ariados (All the above artwork is thanks to @RacieB) The next set of counterparts are two competing evolutionary pairs. On the one hand, the team developed a bagworm Pokemon that would transform into a moth; on the other, they had the option of a spider and... a bigger spider. It’s pretty clear from these two families that the team wanted to introduce new bug Pokemon for the new games. As I’ve shown, the preceding 82 designs in the Korean Index have a ton of water Pokemon and more bird Pokemon than you’d ever need. However, bug Pokemon are completely absent. If the team wanted to emulate Red and Green’s use of bug Pokemon as weak early-game Pokemon which can showcase evolution, then they’d need to introduce at least a couple more, to make the early game seem distinct from Red and Green's Viridian Forest. It looks like the team came up with two evolutionary lines based on bugs the first game didn’t use. Bagworms have a unique silhouette and a iconic look, while Spiders are probably the best known “bug” that wasn’t used in Generation I. Either one could have been a good candidate to fulfill this gap in the new lineup. The team decided on the spiders. We can’t know why for sure, and there are probably lots of reasons, but I have at a few theories. 383 looks very similar to both Venomoth and Butterfree, and the team might not have wanted to retread that same ground. Likewise, 382 is very similar to the Generation I pupae Pokemon, and again might have been too close to previous designs. As well, I wonder the team had an easier time designing a moveset for the spiders than for the bagworms. While 382 and 383 would probably share a lot of the same moves with previous bug Pokemon—at least Harden seems like a given—spider Pokemon presented the potential for moves based upon web spinning. Which we did in fact get: even though it’s awful, Spinarak and Ariados got a signature move of Spider Web, allowing the team to create a new, flavorful move to make the new bug distinct from the old. As we’ll see, Lebyda comes up soon in Period 2c, so clearly the team felt they had room for two new bug families by the time Lebyda appeared on the scene. This could mean that for a short time the team used both bug families, and that Lebyda and its evolution eventually overwrote 382 and 383 as stronger designs. Even if that was the case, it seems like Lebyda’s existence suggests that the team failed to include 382 not because they didn’t have the space for an extra bug, but because there was enough wrong with 382 to discard the design entirely. ID 382: ??? 382 is a bagworm Pokemon. Bagworm moths are a particular type of moths which make bags of natural materials as a cocoon when they transition from a worm into their adult moth form. While bagworm moths can be found just about everywhere in the world, they are especially noticeable in Japanese, which has some particularly large species native to the islands. They have a very distinctive look and stand out immediately as an interesting inspiration for a bug Pokemon. One interesting trait of bagworms is that they make different types of cocoons based upon the types of materials they're presented with. As you can see, the one of the right, made of sticks, looks a lot more like 382 than the picture on the left. For whatever reason, these little guys are really popular in Japanese popular culture. You’ll find them all over the place in Japanese popular culture and video games: (All colorizations in this article thanks to @OrangeFrench) In addition to being a bagworm, 382 has a couple of other interesting features. First of all, notice his white beady eyes; to my knowledge, no previous Pokemon designs in Gen I or II have this type of eyes, and it suggests to me that 382 was designed by a new designer, perhaps the same person who gave us Period 1e. They almost look a little unfinished: would the eyes have gotten reworked in future drafts? On top of that, 382 has some strange ear-like protrusions coming from the side of its body; we see from 383 (and Morph Moth, above) that they are actually the moth's antenae poking out from inside the cocoon? You can also see a little string on the top of its body, where the Pokemon is presumably attached to the tree branch it’s hanging from. Speaking of this worm's bag, notice how the four top twigs under the eyes look a bit incongruous, like they were drawn on top of the sprite. It seems that this guy got redrawn at least once while the designer was coming up with its look; while there’s no real evidence for it, I wonder if the “antennae” were also added to 382 after its evolution was sprited, to create continuity. There are good reasons for Game Freak to decide on a bagworm as an inspiration for their new bugs in Generation II. As I mentioned, bagworms are common in Japanese popular culture, and have a distinctive look that is well suited to the Pokemon aesthetic. Given that the initial map for Gold and Silver was a map based closely on a map of Japan, the team might have been able to use 382 in places which were known for bagworms hanging from the trees. Likewise, bagworms are known, more than anything, for their transformative properties as they become moths. Like Metapod and Kakuna before it, the metamorphosis that cocoons go through is a perfect inspiration for evolution in Pokemon games. At first, I also thought that 382 might have been designed to show off the new catching mechanic from headbutting (or lathering honey on) trees. Given the flimsy string holding it up, and a bagworm’s penchant for hanging from trees, this seems like a natural connection to make. However, that mechanic wasn’t actually designed yet, and even by Spaceworld ’97 isn’t usable. It could be the case that the team were thinking of using this mechanic but hadn’t yet implemented it, but it’s unlikely. However, it is worth noting that Pineco, which may be a spiritual successor to 382, is mostly found by headbutting trees in the final version of the game. Even though bagworms may be distinctive in real life, in the Pokemon world, I suspect that 382’s similarities to Kakuna and Metapod probably doomed it before it got very far in development. With two cocoon Pokemon already extremely visible in the Pokemon world, adding a third one to the lineup would have been, at best, a very strange choice. Maybe 382 would have worked if the team had found something unique about it that could differentiate its gameplay from Kakuna and Metapod. Generation II introduced the move Protect, so maybe it could have learned that instead of or in addition to Harden. But it isn’t likely 382’s moveset could have been significantly different from the bugs preceding it, making this a dead end as well. Not to mention that bagworms transform into moths in real life, and the first generation of games already had Venomoth. There just doesn’t seem to be a niche for 382, given what had come before. Some people think that 382 is just a very early design for Pineco. Even though 382 didn’t make it into Spaceworld ’97, it could have been heavily redesigned, and brought back once the team had created the headbutting mechanic, as a signature Pokemon that could demonstrate the new mechanic. Pineco does have a number of similarities to 382: it’s a bug Pokemon, its body has the same general shape as 382, and although its primarily based on a pine cone, Pineco is also called “the bagworm Pokemon” in the final. So the idea isn’t farfetched. They do have a strikingly similar shape and "pose" (if you could call it that). However, if we look at the history of Pineco’s development a little, it becomes unlikely that these were ever related. Pineco started development as a single-stage Grass-type Pokemon; at first, it wasn’t a bug at all! While Pineco gained its bug typing later, the timing heavily suggests Pineco started life based off a different concept: it was probably a pine cone before the team realized it looked like a bagworm and also made it a bug. 382 on the other hand, was obviously a bug from the beginning. Maybe they noticed the similarity because the team had already considered a bagworm for inclusion. Whatever happened though, Pineco probably started as a completely separate design. If anything, Pineco's evolution, Forretress, draws more inspiration from this cut Pokémon than Pineco. Namely, here was have the idea of a bagworm which now uses steel chunks to make up its shell, instead of twigs. Moreover, both Forretress and 383 share the design trait of eyes poking out of their shell. But Forretress's sprite is still extremely rough even a few months before the final game, which suggests Forretress was also a new design created from scratch. The much more likely connection is that 382 is a very early idea that would eventually get reworked into Burmy in Generation IV. We know that Pearl and Diamond drew a lot of its ideas from Generation II’s rejected designs: Tangrowth, for instance, obviously has some connection to Monjara, the rejected Tangela evolution from Spaceworld ’97. While Burmy doesn’t look exactly like 382, it has enough similarities to note a passing resemblance, and it is clearly based on a bagworm, just like 382. What really convinces me is that one of Burmy’s evolutions, Mothim, is not only a moth, but closely resembles 383. But what makes Burmy work in Generation IV, if 382 so clearly didn’t have spot in Generation II? The main difference is that while Generation II included Metapod and Kakuna in early routes, by Generation IV the team had made the decision to make some Pokemon from earlier generations unavailable. Instead, by Generation IV, the design philosophy was to create new Pokemon that fit the same niche in-game: so for instance, most generations have a counterpart to Pidgey, an early game bug Pokemon, a Pikachu clone, etc. With the decision made to not use Metapod or Kakuna in Generation IV, the team had a gap that they needed to fill with a very similar bug design. What better design to use than one rejected from Generation II for being too similar to Kakuna and Metapod? The second reason Burmy could find a home in Generation IV but not in Generation II was because the team found a gimmick for it that made Burmy unique enough to be memorable in its own right. Rather than simply being a bagworm Pokemon, Burmys are interesting in that they collect different materials for their cocoons based on the environment: a Burmy looks different based on the last place it battled, which in turn makes it evolve differently based on what its cocoon was like. It’s a really cool mechanic, one that’s very flavorful given bagworms and one that make Burmy into something more than just a Metapod lookalike. It’s also something that probably couldn’t have been done in Gold and Silver, which in turn probably makes it a good thing that 382 never got off the ground until much later. My gut instinct is that 382 was rejected pretty early on in favor of the Spinarak family: both 382 and 383’s sprites look very rough and probably needed multiple drafts to bring them up to standard. The concept was a good one, and it makes me glad that the team was eventually able to reuse it on Burmy: it shows that bagworms were a good idea, they just hadn’t found their place yet. ID 383: ???383 is pretty obviously the evolution of 382. Unlike most other case in the Korean Index, we don’t really have to speculate about its connection to 382, because the two designs share so many common traits. The same materials that covered 382’s cocoon are still visible on 383’s body, though now it has little black arms poking out. 382’s strange ear things are now seen to be the antennae for 383. They both even have oddly white pupils, though 383’s now look more natural and are a part of a wide grin. The connections are obvious enough to see that 383 is simply the butterfly or moth that has emerged from 382’s cocoon. 383 faces all the same problems that 382 did. If 382 was too close to Kakuna or Metapod, 383's design is far too close to Venomoth and Butterfree, two notable winged bugs from Generation I. While 383 is probably a moth, given that bagworms are a type of moth, its wings have a very similar shape to Butterfree’s, and would be nearly identical if not for the added dots on 383’s tips. Change a few things, and they’re the same Pokemon. Looking at them there, with the three different palettes, I can't help but imagine them as the three legendary bugs, Butteruno, Venodos, and Mothitres. In other ways, however, 383 is surprisingly off brand for a bug Pokemon. Venonat, Butterfree and Beedrill all have compound eyes, along with mandibles in the place of their mouths. Meanwhile, 383 has a cartoony grin. Again, this is probably evidence that 382 and 383 weren’t made by the designers who made the bulk of Pokemon designs, and it’s also probably an indication that this guy would probably have gone through a couple more drafts if he wasn’t abandoned pretty early in the process. And while it’s certainly pretty innocent in this case, I’m sure after the Jynx incident the team would probably make sure 383 didn’t have a completely black face by the time of the final. More in general for the moment, I wonder if the team was having trouble coming up with more bug Pokemon. As I’ve mentioned, there are tons of water and bird Pokemon in the Korean Index, but barely any bugs: just the used Spinarak family, the Ledyba family, and the 382 family. There’s also almost no new bug moves in Generation II: just Fury Cutter, a Scyther and Scizor signature move, and Spider Web, which was made with Spinarak in mind. Including the Generation I bug moves, this brings us to a grand total of five bug moves. So what gives? Why so little love for bug Pokemon? In part it could be because it’s difficult to come up with new bug archetypes. All the obvious ones—bees, moths, caterpillars, butterflies—were already used in Red and Green. Worms, Slugs, Snails and Spiders are probably also obvious, but worms are functionally very similar to Weedle and Caterpie, while the other three (which are technically not bugs anyway) did find their way into the final version of Gold and Silver (as Spinarak, Slugma, and Magcargo). In addition, three other bugs—Shuckle, Pineco, and Foretress—made it into the final game but look almost nothing like a normal bug archetype, demonstrating how hard it may have been to come up with new original bugs. Combined with Heracross and Scizor, the final game did eventually end up adding nine new bug types, so the final games are hardly wanting for bugs. But at least this early in development, Game Freak seems to have been struggling to find new ways to iterate on the bug Pokemon formula. However popular the design of 383 was, the invention of Ledyba just a little bit later in the Korean Index may have put the nails in the coffin of this design. If the team already had the Spinaraks, Ledyba easily filled any other bug-shaped gaps the early roster might have needed. Given how derivative 382 and 383 are, it’s no wonder they weren’t worked on much more, and didn’t appear in Spaceworld ’97. ID 384: SpinarakLet’s talk about Spinarak. The most obvious choice for a new bug Pokemon was a spider, which was conspicuous in its absence from the first generation of games. There’s nothing too major about Spinarak’s development, but by looking at Spinarak we can see why this design worked, as opposed to the proto-Burmy line above. Spinarak’s cute; his original name, Kokumo even means “baby spider,” while his final name, Itomaru, probably means something like “Thread kid” or maybe “Thready,” since the ending is a diminutive for a young boy, like “Johnny” would be for “John.” The sprite is tiny, and yet it’s elegant: using just a few pixels, we have a very clear idea of what this Pokemon is, what it does, and what we can expect from its evolution. This is mostly signaled by the face on its back. Based on how real life tarantulas often having markings on their bodies to confuse other animals, or make them look scary to predators, Spinarak has a spots that form into the shape of a face. In particular, Spinarak has a sad face, while the early versions of Ariados transformed it into a frightening monster. Notice the faces? The one on the left is even called a "Hawaiian Happy Face Spider." Spinarak’s design didn’t change very much from Spaceworld to the final game. Its face went from neutral to sad, and it changed its pose, to look less like it’s hanging from a web and more like it’s crawling. One consequence of this change is that its face is deemphasized in its final design, making Spinarak look a little more like a generic spider. This would make sense as the team reworked Ariados to lose the two-headed theme, but in my opinion it is a shame, since it makes Spinarak lose a bit of its overall distinctiveness. In Crystal, Spinarak went through its most drastic change: it became green! There’s actually an interesting story to this. Helixchamber noticed something interesting: that on the original Gameboy Color hardware, Spinarak’s specific purple would actually appear to be a greenish yellow very similar to its Crystal palette. On an emulator we lose this effect so it’s been easily missed. The Crystal palette might have been changed to match the color that Spinarak actually shows up as, or, Helixchamber suggested, Spinarak’s official art could have been colored incorrectly because it was based on an artist viewing the sprites on the hardware. Thus, Crystal’s sprites were then changed to match the official art. It’s hard to say if this is for sure, but it’s an interesting, overlooked mystery. On the other hand, it's possible that Spinarak's purple color was simply a holdover from its earliest design when it was purple. The team probably never questioned that palette under their tight deadlines; only once they had a chance to breath did they decide to change Spinarak. Just on the basis of looks, Spinarak is already a far superior design to 382 and 383. It looks distinct from any other bug Pokemon, it has an instantly recognizable design as a spider, and it has an easily graspable quirk: the Halloween face on its back. The other thing that makes Spinarak work is that its (and Ariados’s) designs already hint at what exactly their movesets are going to look like. Here it is, for reference: First of all, because they’re spiders, with fangs, Spinarak are Ariados are pretty obviously signaling to players that they're part Poison-type, explaining Poison Sting. In addition, as a spider we’d expect to see a “web” move, which, predictably, was created to be a signature move in "Spider Web." Not only that, but Scary Face seems a shoe-in given Spinarak’s design. It doesn’t have the move in Spaceworld ’97, but of course it learns it by the final. In addition, you’d expect a spider Pokemon to be able to tie up the opponent—in SW97, it has Bind, and in the final, Constrict—and you’d expect a spider to either bite with its fangs or suck blood, which it does. Also, given its menacing theme, Night Shade feels like an interesting and logical addition to its moveset: it was previously only the signature of the scary ghost line, and many people find spiders scary. While I would be hard pressed to come up with a moveset for 382 and 383 that was distinct from Metapod and Butterfree, Spinarak’s design more or less builds a moveset all on its own. No wonder it was chosen while 382 and 383 were dropped. Spinarak’s might look a little like Weedle, but Spinarak has the benefit of learning its whole moveset without first going into a pupae stage. It’s most similar to Venonat, in that they both share Psychic and Leech Life, but Venonat’s heavily balanced towards using status ailments, whereas Spinarak is more designed around more solid offense, with Nightshade giving it a nice way to deal with type resistance. Spinarak’s movepool, importantly, carved out a niche unique in bug-pokemon-dom. As an aside, I’ve always been a little confused about exactly why Spider Web was designed to be so bad. It’s a move that prevents the defending Pokemon from switching out, an ability that does almost nothing in the regular game and is only moderately useful in multiplayer. And that’s all it does. It doesn’t lower speed or other stats, it doesn’t do any damage, it does nada. Now, sure, some moves have to be bad, but why make a signature move, a move that’s supposed to define the flavor and playstyle of a Pokemon, so crappy? Two ideas present themselves to me. First, the team had never tried out the trapping mechanic before. Since they didn’t know if it might turn out to be powerful, the developers may have been cautious in how competitive it was. As it’s a new ability, they may have also wanted the move to be as simple as possible, to show off the effect without anything else. Or, secondly, maybe the move was developed purely around the flavor of a spiderweb: what does it do, except trap someone? The team may not have been thinking at all about balance when they designed its effect, just what a spider web would logically do. Either way, it kind of sucks; I can’t imagine why anyone would keep Spider Web on their Spinarak, despite it being the defining move of the family. Another thing to notice about Spider Web is that it’s only the fourth new move in the move index for Gold and Silver, after Triple Kick, Sketch, and Thief, and right before Mindreader. I’ve previously discussed whether or not the move index might correlate to creation order when we discussed Smeargle, and my conclusions are inconclusive. Though some of these moves look like they correlate to the Korean Index, some—like Triple Kick, which is the signature move of the last Pokemon in theIndex—don’t correlate at all. So this could be an indication that Spider Web was created early, in close coordination with the development of Spinarak, or it could not. Saying that, Spider Web is surrounded by Thief and Mindreader, two other moves that have unique effects not seen in Generation I. It could have been the case that these three moves were designed to try out new ideas before the team had a clear roster, and then Pokemon were designed around the moves later. If this is the case, then 382 may never have been in strong competition with Spinarak, as Spinarak was already designed to fit an existing move. One more aside about Spinarak’s movepool. What’s the deal with Spinarak learning Psychic? Nightshade makes sense given its association with creepiness or fear, but why is a spider all of a sudden able to attack the opponent’s mind? The only other bug to get this move is Venonat/Venomoth, but they also learn Psybeam, and there’s reason to think they were partly psychic in an earlier build of the games. Moths also have an association with Rorshach tests, with hypnotism (“moths to the flame”), and other ideas loosely connected to psychiatry, so this association almost works. But spiders have none of that. Maybe the idea is that the creepy face on their abdomen plays mind games on the opponent? I can only speculate. In the final game, Spinarak was made a version exclusive counterpart to Ledyba, and was also made exclusive to nighttime, to show off the day/night cycle and to differentiate it more from Ledyba, a daytime only 'mon. This hints that the design purpose of both the Spinarak line and 382/383: they were originally designed with the plan of being early game introductory bug Pokemon. And Spinarak certainly hits that design goal. It’s a simple design that also has a lot of depth. Spinarak shows what happens when the team hits on a good design: almost everything else flows directly out of that initial concept. ID 385: Ariados In contrast to Spinarak, Ariados had a much harder time finding its niche. Ariados went through many redesigns throughout development as the team tried to get its look just right. While I’m convinced that they dropped the ball with its final look, it’s clear that the team were trying to make a menacing spider befitting Spinarak, and were struggling to figure out exactly how to make it the right balance of menacing and lovable. In Spaceworld ’97, Ariados—then named “Tsuheddo” which translates to the extremely on the nose “Two-Headed”—was the perfect evolution of Spinarak. The small neutral face had been replaced by a demonic smile, showing the Pokemon’s true colors as a predator, not just a cute lil boi. While Ariados’s early head isn’t all that exciting to look at, the face on the back exudes character, making Ariados instantly iconic. Saying that, something is off about its backsprite. Don’t they look like completely different faces? On the front sprite, the face is smiling, joyful and ready to strike! On the back, it’s frowning, with two teeth protruding from its mouth. It's a different expression entirely! It's also interesting to me that when you flip the backsprite, it looks almost exactly like an unrefined and cropped version of the front sprite. Maybe the backsprite was simply a cropped version of an earlier sprite with a different face? Either that, or initially in conception Ariados’s abdomen face wasn’t just a series of spots, but could move and show emotion. That would be a really cool take, and give him a lot of character: the spider itself might not be able to emote, but its other face could show it reacting to things. In fact, that's apparently exactly what the anime did with Spinarak, so it's not all that farfetched! So it's backsprite face is either a reflection of this, or like I said, it could be an earlier version of the same sprite. Or it's just a simple inconsistency; who knows? It’s probably nothing, but I can’t help but notice we’ve had a lot of menacing, smiling faces featuring on Pokemon in the last few entries. I’d say it’s a sign that maybe the last five entries in the Index were by the same creator, but the other parts of 382 and 383’s spritework are too different from Mantine and Ariados to match. Still, I’m still curious if whether the face on the back on Mantine in some way inspired a creator to base an entire Pokemon design on the face, leading to the Spinarak line? It seems possible given the similarities of the faces. Anyway, after Spaceworld '97, the team decided to heavily rework Ariados' design. By June 1999, its head stayed the same, and its basic pose is the same, even though it was redrawn to have a little more perspective in the pose. But the abdomen is completely redrawn: instead of looking like a menacing grin, its abdomen looks a bit more realistic, has lines down it almost like a Jack O’ Lantern, and a sad, gaping face. I’m not sure what the point of this change was, because the old face looks much better and the new abdomen doesn’t really add anything. Maybe Sugimori asked the artist to make it look more realistic? Or maybe the artist was explicitly trying to evoke a Jack O’Lantern and they thought the new face looked more fitting for a carved face? We had a Christmas themed Pokemon in Delibird, so maybe this was a deliberate attempt at a Halloween Pokemon? There are two more intermediate sprites on the Scratchpads that give us more of a picture of what Game Freak was trying to do here. At first, I thought they were both permutations of a sprite after Spaceworld '99, right before the final's rework, but when OrangeFrench colorized them for me, he pointed out that one of these sprites was clearly made for the Super Game Boy palettes that Spaceworld '97 used. Notice how the left-most sprite looks way better in the first palette, and the second one looks more natural in the second palette: As you can see, the dithering on the left sprite looks too white and out of place with the Gameboy Color Palette of the final, but it fits the style of the earlier sprites. Which means that Ariados's sprites were probably designed in this sort of order: So what it looks like is that the team briefly tried out a new pose for Ariados, before deciding it didn't work and going back to the old pose. Even so, as they went back to that pose, they gave its face two teeth on the bottom row (potentially to look more like a Jack O' Lantern?). Then, in the fourth and fifth sprites, the team considered getting rid of its fangs almost completely, before moving back to the old pose they tried and then completely overhauling it. The reason they spent all this time reworking its sprite is obvious: the new Ariados with the sad, Jack O' Lantern face doesn’t look at all like it deserves to be a final evolution. Evolved Pokemon are often designed to be imposing or dangerous looking; a clear upgrade to what came before as a reward to the player who put time into raising them. The SW’97 version of Ariados certainly succeeded at this, mostly due to the way the evil grin was a natural upgrade from the tiny spider before it. But once the team replaced that grin with this dopey, sad, Jack O’ Lantern face, the incentive to evolve Spinarak just no longer seemed to be there. If Ariados is sad to be alive, why should its trainer want to use it? In part, I bet that the team was having difficult with Ariados’ design due to the unique problem spiders have in game development. The “spider dilemma” is borne out of the fact that a lot of people have arachnophobia, and a fear of spiders can be acute enough that some players might not like to view them in games. The dilemma is this: do you make your spiders cartoony, or realistic and scary? If you make it too cartoony, it loses all sense of presence, or no longer becomes an interesting enemy to face. At the same time, if you make it too realistic, you flirt with the possibility that some players might just decide not to play your game at all. My suspicion is that the team couldn’t quite find the right balance with Ariados. The team may have toned own the first version to make it less frightening, but the version they ended up with became dopey. Eventually, however, one of the team members started experimenting with the design. As we can see from extra sprites found on the scratch pad, the artist tried changing the pose and experimented with losing the fangs. It still looked dopey. The problem was that Ariados’ abdomen face is pretty silly looking, and the face just wasn’t doing its job of making Ariados look interesting or imposing. They could have gone back to the original face, but for some reason the team hadn't liked it, so they moved ahead in a different direction. Using the last pose experimented with, the artist decided to simply buff it up. The eyes are now mean instead of dots, the pincers are larger, its got a spike on its head, and the legs are thinner to look less chubby and more agile. In doing so, modern Ariados was born. Except…it completely lost the face. The entire concept that the Spinarak line was designed around was discarded into order to make Ariados more extreme. And yet, it still has “dos” in its name. Its original name was “Two-Headed,” so the “dos” must refer to a second head still. But where is it? Maybe the Pokedex entry can shine some light? “It spins string not only from its rear but also from its mouth. It is hard to tell which end is which.” Wait, so its rear is hard to tell from its face? Let’s check those sprites again: …Oh. The team gave Ariados a butt-face. That’s certainly…a choice. But it seems the concept did survive all the way to the final game. I definitely prefer the SW’97 design more than the final, but I can see how a series of decisions led the team to the…strange choices they made by the end. Despite losing the most of its theming, the final Ariados certainly looks a lot more like a final evolution, and something I wouldn’t want to mess with. Before we leave this, I want to just mention one thing I’ve noticed while staring at these sprites. Has anyone else ever wondered why the Spinarak family doesn’t have eight legs? If there’s one thing that defines spiders, isn’t it their eight legs? Maybe eight legs just didn’t look good on the sprites, so just like cartoon people often have only four fingers, the artist deleted one set of legs? My deeper suspicion is that Game Freak deleted one set of legs because, technically, bugs only have six legs, and so if you made a Spinarak with eight legs, it technically couldn’t be bug type. But if that’s the case, why does the final Ariados lose one more set of legs? If I'm not mistaken, doesn't it have four now? I get that maybe the last pair of legs is whatever’s on Ariados’s back, but what the heck is on Ariados’s back? Why would it have two half-legs aimed in the air like that? What are they for, or supposed to be? OrangeFrench has suggested to me that they could be the weird legs of a Peacock Spider. Who knows? Counterparts: 386 versus 387 (Thanks again to Racie B for the above artwork) The next two Pokemon in Period 2b end up both being unused, and so the information we have about them and what design niche they were supposed to fill is very scant. One of them is a bizarre raccoon who seems to be permanently on fire; the other is a seemingly normal looking Koala with rings instead of ears. Who drew these, and what exactly they were supposed to do, is a mystery: even their typing is up for grabs, though I suspect 387 was at least part-fire. My best guess is that the team wanted more woodland creatures to fill the early forests of Gold and Silver, and so they brainstormed these two based on the environment they were going to be located in. 386 is pretty exotic for Japan, so whoever drew it may have tried to find a woodland creature that Japanese kids would find interesting and different. 387, on the other hand, is based on a folktale about a tanuki, so it would fit right into the Japanese environment that the early versions of Gold and Silver more emphasized. There are two reasons I can suggest as to why neither of these were chosen. First, the next set of counterparts—the Houndour and Warwolf lines—were both chosen for inclusion into Spaceworld ’97. If the team liked both counterparts so much that they wanted to use them both, they might have cut something earlier from this part of the index, and 386/387 are the obvious choice. Secondly, 386 and 387 have all the tell signs of being designed by the same mystery designer that created Period 1e and parts of 1b. Almost all of those designs are unused by Spaceworld ’97, and my suspicion is because whoever designed these groups of Pokemon was relatively disconnected from the main group of Pokemon designers. Thus, their designs were different from the house style of Sugimori, Nishida, and Morimoto, and thus were often not Pokemon-like enough for inclusion. As well, whoever this designer was probably didn’t work full time with the monster designers, and thus probably wasn’t around to be an advocate for their own designs as much. For whatever reason, the Pokemon designed by this mystery person were almost never chosen, and 386/387 are no exception. ID 386: ??? 386 is a Koala hugging a tree branch. At first glance, it’s a pretty standard Koala: the proportions aren’t much different from the real life creature, it’s hanging on what looks like eucalyptus, just like a real-life Koala, and its pose is pretty much what you'd expect from the real life animal. I’ve talked before about the simplest Pokemon designs, in which the designer took a real world animal and put a slight twist on it. I call those designs “Animal+” but that’s almost too charitable here. 386 isn’t “Animal+”, it’s more or less “Animal.” There are two subtle distinguishing features of 386. First, its paws are darker colored and look a little like they’re gloves. More significantly are its ears, which are not what they seem. While at first glance they look like regular ears, if you glance at the backsprite, you can see part of its branch through the middle! That means that the ears are actually rings. It’s easier to see them once the sprite has been colored. (OrangeFrench colorized this sprite, like all the others in this article, but I changed the background to a light blue, to help show you what's significant about the ears.) What does that say about 386, and what it may have done? Not much. Rings kind of remind me of a satellite dish, so maybe 386 would receive messages from somewhere and act upon them. Like a radio-Koala, or a robot Koala. Rings might also be flotation devices, so that 386 could swim. Or they could mean nothing and were put there by the sprite artist on a whim. 386 fits in with a lot of the other simplistic animal designs, especially from Periods 1a, 1b, and 1e. Like I said above, I get the feeling that whoever designed the Pokemon that populate these periods was only probably not one of the main three, and so their designs (whoever they were) were usually not picked up for inclusion in the games. But whoever this designer was, they really liked designing Pokemon around exotic animals. We've seen a lot of this in the unused designs: There is one other Koala Pokemon in the series, designed for Generation VII, years and years later than when the Korean Index was populated. That koala Pokemon is Komala. Komala is a cute, normal type Pokemon, which is described as permanently asleep (despite not learning snore or rest). There are a few similarities between Komala and 386. Most notably, they’re both Koalas, but they also share a pose and are both holding onto a tree. Some people theorize that 386 is an early proto-version of Komala, but this seems very unlikely to me. First, of course, is that Komala was designed 18 years after 386. There’s no way they kept the design around that long and then decided to reuse it; far more likely is that a later designer had the idea for a Koala Pokemon, and they happen to look similar. Sure, both of them are grasping onto a stick of some sort, but that’s hardly conclusive, as that’s more or less the default pose for Koalas. Literally if I google “Koala” the first two images I find are of Koalas in that exact pose: Secondly, 386 and Komala, while they are both the same animal, don’t share each other’s defining features. Komala is defined by its perpetual sleepiness, but you’ll note that 386 has its eyes wide open. Meanwhile, 386’s defining feature is its ring-ears but Komala’s ears are conspicuously not even circular. Essentially, though both of these are based on the same animal, they really have nothing else in common. But even if Komala is unrelated to 386, it can still give us a sense of why 386 never made it past the concept stage. Look, for instance, at Komala’s moveset in generation VII: …It’s really generic, isn’t it? Smogon says that Komala’s main use is its ability, which makes it immune to statuses, and the move Rapid Spin, but otherwise its generally outclassed by other Normal-type Pokemon. Defense Curl, Rollout, Slam, Flail, Thrash: pretty much every Normal-type Pokemon learns these moves. What there reminds you of a Koala? Komala was not brought back in Sword and Shield, the very next generation, which suggests to me it was not very popular and Game Freak didn’t make it a priority. And I think the moveset nicely demonstrates why. It’s hard to make a Koala Pokemon that isn’t incredibly generic. Koalas don’t fight in real life, and their main gimmicks are eating weird leaves, a very stupid sounding scream (listen to it if you don’t believe me) and their ability to carry chlamydia. None of these things makes a good Pokemon hook. Adding the sleeping angle to Komala was a nice try, but it clearly didn’t help the team figure out an interesting moveset. Given the much smaller range of choices Game Freak must have had in Generation II to develop an interesting Koala Pokemon, I don’t blame them for failing to here. Koalas are cute. Koalas are cuddly. I’m not convinced that Koalas make a good Pokemon concept. ID 387: ???Let’s talk about Tanukis. The tanuki is a Japanese animal that is normally translated into English as a “racoon.” It looks superficially similar to a racoon, but is otherwise almost completely unrelated to that animal. The main significant difference is that while racoons are most closely related to cats and bears, Tanuki’s are much closer to foxes and to dogs. Getting the two mixed up is literally confusing cats for dogs! In Japan, tanukis are a common staple of folklore. They’re most often the trickster character that is commonly found in folklore, like Raven in Pacific Northwest mythology, Coyote in Southwestern American mythology, or Brer Rabbit in African American folklore. Tricksters are usually portrayed as heroes, even though they are often selfish, self-centered, and troublemakers: their stories are most often fun because even though they are quite human in their flaws, they mostly make fools of those in power or take advantage of people otherwise seen as oppressors. They’re weak creatures, but clever and sly, which allows them to become champions of people without power in a society. For the Japanese, Tanukis often trick farmers or landlords who have been cruel of selfish. They appear literally all over the place in Japanese video games: Mario dresses up as one in Super Mario Bros 3, the titular Rocky of the Pocky and Rocky games is a Tanuki, the greedy venture capitalist Tom Nook is a tanuki. They also show up a lot in Japanese anime, most famously in Pom Poko, an early Studio Ghibli creation. Tanuki have a range of magical abilities. They can use a special leaf to transform their form into other animals or into humans, or to transform objects into valuables. They also have, strangely, extremely large testicles that they carry around in a sack, often thrown over their backs. Their sack is often said to be the source of their powers, and losing it spells doom for the tanuki. They’re also often depicted as alcoholics and beings that love to trick people just for the fun of it. Most of the time, tanuki are the good guys, but in one particular story, that of Kachi-Kachi Yama, the tanuki is the villain. The story tells of a old farmer who constantly had to deal with a tanuki ruining his fields. The man decided to set a trap, and captured the tanuki, who pleaded to be let free. Eventually, while the man was gone, the tanuki convinced his wife to set him free, but as soon as she unleashed him, the tanuki murdered her, chopped up her body, and made her into a soup. Then, to trick the farmer, the tanuki changed into his wife, fed him the soup, and then taunted him for eating his own loved ones. It’s…kind of dark. Enter a wily rabbit, not unlike Brer Rabbit. The rabbit finds out what happened and promises the farmer to help him get revenge. From that point on, the rabbit pretends to be the friend of the tanuki but instead plays tricks on him to torture him. The most prominent trap the rabbit plays happens when the tanuki is gathering firewood to bring back to his home. While he’s walking back home with the wood the rabbit lights it on fire, but he doesn’t notice because the firewood is packed to high for him to see. He asks the rabbit what the crackling sound he hears is, and the rabbit just tells him it’s the sound of forest creatures. Because he doesn’t suspect a thing, the tanuki is eventually badly burnt by the fire, and the rabbit helps the farmer get revenge. What I’m trying to tell you here is that 387 is the tanuki from the story of Kachi Kachi Yama. Tanukis are a great subject for a Pokemon: they have all sorts of cool magical powers, they’re a staple of Japanese video games, and as a cute mischievous racoon, they have obvious mascot potential. And yet, for the most part, tanukis have never been used as the inspiration for a Pokemon. Zigzagoon is probably based on the real life tanuki, but doesn’t share any of the folklore of the creature. Instead, this is the only time the team tried to make a more traditional, mythological tanuki work. However, 387 clearly doesn’t work. The sprite we have is more or less an action pose: it shows 387 being burnt by the fire on its back, carrying kindling. It’s already hard to naturally explain Pokemon carrying objects: do Farfetch’d always have their green onion? Where does Alakazam get its spoons (are they a part of its hands?)? Is every single Cubone ever wearing the skull of its mother (this implies a lot of childbirth deaths. How does it get the skull? Does it personally remove it from its mother?)?
But 387 is on another level. Does 387 always carry around tinder? Is it always on fire? Is the Pokemon in a constant state of pain, running around avoiding the flames on its back? How does it sleep, or eat? Potentially, this could be reworked as they refined the design: for instance, maybe they could get rid of the tinder and the fire could just erupt from its back, Cyndaquil like, when it gets excited. But as it is, 387’s sprite is showing us a slice from its life, but nothing that actually shows us what its really like. Above all that, there’s a very notable push in the development of Gold and Silver to move away from references only Japanese players would understand. With the international success of the first games, the team decided after Spaceworld ’97 to make the world map less obviously resemble Japan, and deleted or edited Pokemon designs that leaned into Japanese mythology too much. Kyonpan, for instance, was probably removed for being too closely associated with Chinese mythology. 387 didn’t make it even into Spaceworld ’97, before this was a concern, but even if it had made it there, it’s almost certain that it wouldn’t have survived to the final, at least without being heavily remodeled to remove most references to Japanese folklore. Which, in the case of 387, is more or less its entire identity. I doubt this guy ever got farther than a sprite, but it’s fun to speculate about what it could have been like as a more fully fleshed out Pokemon. Would 387 have been a Fire-type, and would it have used the flames on its back to attack other Pokemon? Because it’s a tanuki, would it learn transform? If this guy did get all the way to a moveset (at least a partial one), then I’ve always suspected the move “Thief” was created as a signature move for 387. While, like I said before, the new move index is probably not entirely chronological, Thief does appear third on the list, right before Spider Web (the signature move of Ariados, two slots above this), and two moves after Sketch (the signature move of Smeargle, 38 slots above this one). If these moves were created even somewhat chronologically, then Thief would belong to something around this part of the index, and a mischievous, fast moving tanuki seems exactly the sort of creature who might be stealing other Pokemon’s held items. In addition, in the final game, Thief isn’t learned by any Pokemon, and is strictly a TM only move. That’s weird, and suggests to me that it may have been designed for something that got cut. While that cut Pokemon could have been anything on this list, 387 is my prime suspect. We’ll never know though, because neither 386 nor 387 were chosen, despite being made to fill the same slot. Instead, the next group of counterparts, an Ice wolf and a fire dog, were both so well designed that the team decided to choose them both, taking slots that were probably meant in part for one of these guys. Not a big loss, in my opinion. Period 2a was, overall, a pretty straightforward section of the Index. There, we saw the Pokemon team create all three starters (though of course two of those starter lines would eventually be replaced), along with Pokemon like Rinrin and Skarmory that showed off the new typings the team had come up with. Alongside those were the first baby Pokemon, the first Pokemon made the breeding mechanic/Daycare system, and the legendary trio of these games. Period 2b, upon first glance, looks much stranger, and a pattern is initially much less apparent. While almost everything in 2a was used in Spaceworld ’97 (at least), about half of 2b didn't even make it that far. Much of 2b feels like a throwback to sketchier ideas, back into the Animal+ designs that characterized the first era of the Index, and it doesn’t seem ordered in a way that makes a lot of sense. After the logical ordering of the previous section, Period 2b is a 180 degree turn. However, once you examine this section more, a pattern does emerge. Notice first how almost exactly 50% of Period 2b is discarded, and the discarded designs almost always trade off with used designs. The pattern we see emerging is of two Pokemon in the Index right next to each other, one used and the other one discarded. What seems to be going on is that the Pokemon were in competition with each other. Both were made to fill the same concept or niche, and only one was chosen to appear in Spaceworld '97. Why did they switch to this pattern now? It could be related to how many slots in the Index the team had already used. By Period 2b we’re up to 78 new Pokemon designs, and even the final game only added 100 new Pokemon, so the team must have begun to know that the open slots for new Pokemon were getting filled. Rather than just throwing out more designs, it seems at this point the team started creating competing designs to fit the last niche's they needed for the new roster. It seems as though the designers were designing to a checklist, trying to create a Pokemon that fits in a needed role in the final game. Thus, Period 2b is defined by a series of “either/or” competing designs. The first two in this section are a Stork or Pichu, both of which fit the niche of designing something to do with baby Pokemon, and only one of which was chosen to make cut off for Spaceworld ’97. After that, we continue to see competing designs like this: Pokemon designed to fit odd type pairings, a choice of two bug Pokemon evolutionary families, woodland Pokemon who could live in forests, Ice or Fire dogs, and a pattern that’s more difficult to pin down. It looks like each slot in the next sixteen was paired up with the one around it, and the team was supposed to decide on one or the other for inclusion. It’s notable that exactly half of Period 2b found its way into Spaceworld ’97, and the rest was scrapped. It isn’t clear if Octillery and the Kudagitsune were part of this pattern; my gut tells me they are and I misplaced them in the previous section. However, they don’t have an obvious counterpart nature, so maybe not. On the other hand, it might be a mistake to imagine there was a direct contrast between every pair in Period 2b; the designers may not have thought about it that strictly. Maybe Octillery and Kudagitsune were a pair, but they weren't made to directly contrast each other. Anyway, as we go through this next section of the Index, I’ll try to focus on how each design contrasts with its rival, and speculate on exactly why one was chosen over the other. Counterparts: ID 378 and Pichu (On the left was created by Racie B; the right image is official artwork) To start off Period 2b, we have an odd pair. One's a Stork that didn't even make it through design long enough to appear in Spaceworld '97. The other is Pichu, one of the most memorable and beloved Pokemon of Generation II. The reason these two are competing in the Index is because they both represent alternate directions the team could've taken in regards to baby Pokemon and the breeding mechanic. Period 2a saw the first Pokemon designed with breeding in mind, and also showed us Cleffa, the first time a baby Pokemon was designed explicitly as a baby Pokemon (Elebebii was probably just thought of, originally, as the first evolution of Electabuzz). It’s likely that the team came to a crossroads at this point in development. Should they focus more on adult Pokemon (like Miltank) that interacted with the breeding/daycare system? Or should the team create more baby Pokemon, like Cleffa (and Elebebii) to capitalize on the system in a different way? From our vantage point, it’s easy to see what the right decision was. Babies added a natural incentive to play around with the breeding system, and were natural mascots for the series: take the most popular Pokemon, and make them chibi and even cuter! However, I’m sure there was at least some uncertainty about whether that was a good option: Cleffa is cute, but the team assuredly knew it was useless in battle. Making more baby designs would fill up the precious last slots of the game with Pokemon generally useless except for collecting. To choose Pichu over the Stork would be to swerve from the designs goals of Generation I, in which the Pokemon were, above all else, there to fight. Instead, the inclusion of so much babies would send Pokemon down a new genre path: that of a monster collector. We all know which direction the team went in. ID 378: ???To start off Period 2b, we have an adorable stork guy. Look at it's happy mug, it's so excited to be here! The overall look of the design reminds me a lot of Nishida’s handiwork, though really, this could have been made by Oota or any of the newer designers for whom we don’t have much information. As cute as it is, the Stork clearly didn’t fit the design needs of Gold and Silver, and although there is some evidence here that it went through at least a couple of drafts, it’s nowhere to be found in Spaceworld ’97, leaving it somewhat of a mystery. In a lot of ways, ID 378 looks a lot more like a Generation I design than the style that the team has settled on since the beginning of Era II. It’s a lot cuter and less monstrous in design than what we got out of Generation I, for sure, but it is very clearly an Animal+ design, with very little interesting features besides its identity as a stork. However, the mythology surrounding storks and babies is apparently more uncommon in Japan, making this feel more unique than it does to western eyes. Saying that, 378 is not the first stork to have shown up in a Japanese game. 378 reminds me most of Pokemon like Kangaskhan, or the unused Marowak evolution. All three emphasize a motherly role in taking care of a baby Pokemon. There's also some Dodrio in this design, in its pose and its tailfeathers, though admittedly Dodrio is a much angrier dude. It’s likely to me that Kangaskhan was the inspiration for 378. By Period 2b, the team was deep into trying to figure out how the breeding mechanic would work, and creating another motherly ‘mon seems like a natural experiment to undertake. (378 colorized by @FrenchOrange) However, parent Pokemon are clearly a bad idea once we think about them more, and the whole idea of 378 kind of falls apart. First of all, if you bred 378, would it hatch already holding a baby in its mouth? And the baby looks significantly different than 378 itself, a coincidental callback to Kangaskhan's blue baby. The obvious thing to do would be to make the baby stork a pre-evolution. It could eventually evolve into 378, and get a baby in its own mouth! However, nothing like the baby is anywhere in the Korean Index, so the team likely never got that far before they decided against 378. But even if the team did design a pre-evolution to the Stork…then they just created another baby Pokemon! And if the team already committed itself to making more baby Pokemon (remember, babies are useless in actual battles, so presumably going the Stork route in development would have the primary perk of making more usable Pokemon), than why not make Pichu, rather than a generic bird baby connected to a Stork no one’s seen before? It’s pretty clear, then, that this concept was doomed from the start. On top of the other problems, 378 is yet another example of the huge amounts of uninspired bird Pokemon in the Korean Index. Given the heavy competition between new birds in the game—and the fact that Hoothoot, Natu, and the upcoming Murkrow are far more inspired designs than 378—it also seems as though the Stork already was competing for a crowded niche, even if you ignore the problems with its concept vis a vis breeding. There are just...so many bird Pokemon in the Korean Index. Another possibility is that 378 was an even older design taken from Generation I's design and reintroduced as the team was brainstorming more ways to use breeding. This isn’t the only time old designs would return in a new context during Era II: two slots after this, ID 380 is a Squid that is suspiciously similar to unused Generation I backsprites. As well, wait until Period 2c to see how the team repurposed a lot of older Generation I designs into baby Pokemon. I don’t have a lot of evidence for this, except for the design similarities between 378 and Kangaskhan or even Dodrio. If it was repurposed for Gen II, it certainly got redrawn, since the sprite in the Korean Index doesn’t look like one initially drawn for Generation I. One thing that could be significant, or could be nothing, comes from a comparison of the stork's two sprites. The backsprite has a few discrepancies from the front sprite, such as it's lack of tailfeathers; this could be an indication that the front sprite was redrawn and there was an earlier design, or just artistic license. Anyway, for some or all of the reasons given above, 378 never saw the light of day, and because of that we know very little about how it would have functioned. Instead, the team quickly decided that its slot would instead be filled by ID 379: Pichu. ID 379: PichuPichu might be the Pokemon from Generation II with the most name recognition; really only Ho-oh is in competition. It’s the only Generation II Pokemon to feature in Smash Brothers, and Pichu is a regular presence in all sorts of spin-off media. While it certainly isn’t the crowd pleaser that Pikachu turned out to be, Pichu has a legacy that few other Generation II designs have. Pichu’s important because his place in the Korean Index denotes the point at which Game Freak decided to go all-in on baby Pokemon. Elebebii was probably designed around a different idea, and Cleffa was an experiment, but Pichu shows that the team had finally signed off on making this subgroup of Pokemon a feature of the new games. In addition, Pichu also seems to be where the team made another important change. While Cleffa was the first baby Pokemon designed, the team wasn’t sure what the design philosophy was going to be for a “pre-Clefairy,” and the sprite looks kind of odd. But with Pichu, the design team realized what was obvious: that babies should take what was successful about the parents’ designs, and exaggerate them, making them even cuter. Pichu’s initial design dialed the cuteness up to eleven. It did so by simplifying Pikachu as much as possible, into a round circular puffball with pikachu’s signature cheeks and some exaggerated ears. The Cutting Room Floor suggests that this initial design was meant to mimic a daifuku, or Japanese dessert. This was the same dessert that the original version of Pikachu was originally designed to resemble, which makes this theory about Pichu make sense. If Pichu really was based off a Daifuku, then it is possible that Pichu was, oddly enough, inspired by Pikachu’s lost original design, rather than just by chibi-fying the existing Pikachu. (Left is by @RacieB, right is a daifuku) Note that by the time of Spaceworld ’97, Pikachu’s design hadn’t quite evolved into the sleek, anime inspired design that Pikachu sports in Yellow, and that may have influenced the original design for Pichu. In fact, when you put the original Pichu up against the “fat pikachu” Generation I sprite, or the sprite from Spaceworld ’97, there a much more natural resemblance than if you compare Pichu to the final design in Gold. While Pikachu hadn’t quite evolved into the modern design we all know by the final Gold sprites, you can tell that the amendments to Pichu’s design, which made it sleeker and gave it a more identifiable neck, were made with this design change in mind, and that the final Pichu sprites much more match the thinner design for Pikachu. It’s also worth putting Pichu up against the early sprites for ID 306 and 307 as well, given their possible connection to the original design for Pikachu. Way back when, when we discussed ID 307 in Period 2b, I speculated that 306 was the original Pikachu design, brought back from the grave for another chance in Generation II. If 306 does have a connection to the original Pikachu, then maybe Pichu’s sprite was developed out of 306’s; they’re obviously different designers, given the differences in how the sprites are drawn (look in particular at the shading around the eyes), but maybe there are enough similarities to buy that 306 is related here? You be the judge. I’m also including 305 here, because it’s round ball design also has its own similarities with Pichu. I really love this round Daifuku design for Pichu, much more than I like the final design. It’s elegant in its simplicity. However, the final designs for Gold and Silver often moved away from these sorts of designs, instead adding a lot of details that made the final pokemon resemble cartoons a lot less and real creatures a lot more (see, for instance, how Quagsire became less exaggerated, or how the Hoppip line grew arms and legs). The final Pichu design also better matches the final Pikachu design. However, we actually have a third design of Pichu as well! Unlike most of the other Pokemon in the Spaceworld ’97 build, there are some sprites of Pichu's awkward transition that the team experimented with between the original and final sprites. Behold: awkward derp Pichu! What’s weird about this is that the sprite isn’t used in the June 1999 sprites we have, only as the silver sprite in the Spaceworld ’99 files. Regardless, it’s definitely an earlier design, since June 1999 uses a back sprite that exactly corresponds with this guy. He’s clearly a rough design: the expression on his face looks weird, his ears are still being refined, and the pose is strange and unnatural. It’s clear Game Freak wanted to move away from the round puffball, but weren’t quite sure how to yet. Elements of this design were reused when the team created the final Silver sprite for pichu: if you pay attention, you can see that it has the same face and mostly the same tail, but that the rest of it was reproportioned and pichu was given a neck. It makes sense why the team were so slow to give Pichu a neck, as Pikachu didn’t really have one at this point in Pokemon history, and it would have been weird for him to lose his neck upon evolution. Funny enough, even if the no-necked design didn’t work very well in Pichu’s case, some later “Pikachu Clones” have this same body type: Gameplay-wise, only one thing stands out about Pichu's evolution to the final. For the most part, Pichu follows the same pattern through development as Cleffa and Igglybuff. It initially evolved into Pikachu at level twelve (which doesn’t make a lot of sense lore-wise, as you can encounter Pikachus in Viridian forest at a much lower level than this). Then, like the rest of the babies (except for Elekid, Magby, and Smoochum) its evolution level was standardized at ten, then changed to a condition of high friendship. Otherwise, its base stats stayed the same, and it’s moveset followed what we saw with Cleffa. Initially, Pichu had the same moveset as Pikachu, just learned at a lower level, but by the final, Pichu had lost all of its moves except initial level one moves and the addition of Sweet Kiss. Pretty standard stuff. However, while the three babies (Pichu, Cleffa, and Igglybuff) were standardized to evolve with high friendship, Pichu was changed to a friendship evolution one month earlier than the other two. It wasn’t the first friendship evolution: Chansey and Espeon were the first to evolve with high friendship, while Umbreon and Crobat were changed to evolve with low friendship at the same time as these two (and then changed to high friendship when it became apparent a low friendship evolution was a terrible idea). Thus high friendship evolutions were an idea that preceded Pichu. Saying that…the friendship mechanic was initially developed in Pokemon Yellow, as a way of tracking how much your Pikachu in that game loved you, and what mood it would show off if you talked to it. Which means that the idea for a high friendship evolution might have been inspired by the bond you form with your Pikachu in that game. While Blissey and Espeon pioneered the idea, high friendship was a natural choice for a Pichu evolutionary method, given where the idea originated. The team may originally have decided on this evolutionary method as being unique to Pichu out of all the babies, and then only expanded it to Cleffa and Igglybuff when it showed promise. One last thing worth noting about Pichu is that the creation of Pichu formally closed off the possibility of the unused Raichu evolution, Gorochu, ever being brought back in a later generation. No Pokemon has ever had four direct evolutions, and so once Pichu was chosen to be in Generation II, Gorochu no longer had any space in the evolutionary line. To my knowledge, this is the first point in Pokemon’s development that one of those old unused designs from Generation I were formally trashed, with no chance to be reused. With the existence of Pichu, the designers had already changed how they conceived of the Pikachu family line. Most significantly, the creation of Pichu is an acknowledgement that Pikachu was the mascot for the franchise. Had the team known Pikachu would be the massive hit that it was, they probably would never have created Raichu in the first place: why encourage players to trade away the iconic fan favorite for a less beloved design? They even try to stop players from getting a Raichu in Yellow, by preventing the player’s Pikachu from ever evolving: that’s not behavior we’d expect from a team who believed in Raichu’s design. Had Gorochu existed, this would have just compounded the problem, by making Pikachu the insignificant starting member of a line that most players would evolve into Gorochu as soon as they could. However, by replacing Pichu with Gorochu, Pikachu’s predominant place in Pokemon is established: Pichu only exists to be a proto-Pikachu. While Raichu is still a problem, the evolutionary family is now much more Pikachu focused, promising that players won’t forget or discard the mascot of the entire series. Anyway, Pichu’s inclusion in the games marks the point at which the team decided to go all in on baby designs. Rather than experimenting with parenthood depicted through Pokemon, now the idea of chibi mascots is established, and from this point on in the Korean Index, the team would design nearly a dozen other baby Pokemon, so many so that most were cut from the final game. Between Pichu and the Stork, this was an easy decision for the team to make: other counterparts in Period 2b will prove to be much harder. Counterparts: ID 380 and Mantine It’s a little less certain exactly what the trade off between the next two Pokemon in Period 2b was. My best guess is that the team was trying to design a water Pokemon with an unusual secondary type. Thus, the next two Pokemon pair Water with two types that you normally wouldn't see alongside it: 380 probably paired water with Ground, and Mantine paired it with flying. Yes, Gyarados had previously shown players what a Water/Flying Pokemon would look like, but Gyarados’s flying typing never connected much with its design or gameplay: it learns no flying moves, and its resemblance to a koinobori (and windsock shaped like a dragon) doesn't really stick out in a significant way, and is something completely lost to western fans. On the other hand, Mantine, in it’s original design of a fish with feathers, soaring through the air, is much more obviously designed after the brief: what would a water pokemon that could fly look like? Mantine can answer the question of how that typing would actually work, though admittedly even in Spaceworld ’97 it also didn’t learn flying-type moves. Mantine’s counterpart is ID 380, a squid that may have originated in Generation I. At first, 380 looks like a simply animal design, except for the fact that it has drill arms and a drill on its head. These augmentations clearly suggest that 380 was supposed to be Ground type, or at the very least, Rock type. There were no Water/Ground Pokemon up until this point, and so, like Mantine, 380 would be charting new territory. There’s also a nice symmetry in that while Mantine was a water Pokemon who could fly, 380 was one that burrowed in the ground. It isn’t clear why one was chosen over the other, though it is worth noting that Mantine was almost completely revised by the final games, suggesting the team wasn’t happy with how on-the-nose the designs of Mantine or 380 were. In the end, the decision between the two could have come down to game balance, or the team may have just liked Mantine more. We don't have a lot to work off of in speculating why Mantine was chosen over 380, so your guess is as good as mine. ID 380: ???The unknown Pokemon occupying slot 380 in the Korean Index is intriguing for many reasons. First of all, it’s very likely a holdover from Generation I, where it was also unused. Secondly, 380 seems to be sporting an otherwise rare typing combination of Water and Ground. As I said above, this Pokemon was probably meant to compete with ID 381—Mantine—for a slot in Spaceworld ’97; in the end, Mantine seems to have been the more favored design. The Pokemon 380 reminds me the most of is Inkay, though obviously the design similarities are a coincidence. Both are designed after the same type of squid and that the same sort of triangle design on their heads. There's no way 380 was redesigned into Inkay, given how far apart in creation they. It's still kind of interesting to see a similar idea much later in Pokemon's lifetime. So far in the Korean Index, we have, admittedly, had a lot of water Pokemon show up: Gurotesu, pre-Gurotesu, Ikari, Marill, Quagsire, the Manboo family, Politoed, Quilfish, its aborted evolutionary relative, and the Viking boat Lapras-like beast. Obviously, the team wasn’t lacking any inspiration for water Pokemon. However, they were lacking diversity, and it could have been the case the team was having difficulty making these new water types appreciably different from what they already had in Generation I. Generation I was full of water types, and they had covered most of the archetypes available to the team, given the narrow range of water moves in these first two generations. The Tentacool family combined water with Poison and had the annoying bind effect; Gyarados and Magikarp gave us a weak Pokemon turning into a terrible dragon; Seel and Shellder were both water Pokemon that gained the ice type upon evolution, and one was offensively statted while the other one was a tank. Finally, Staryu and Starmie explored the design space for Special attackers, and alongside Slowpoke, they tried out the concept of water/psychic types. Compared to the diversity of typing and design space taken up by Generation I water types, what do we have in the Korean Index? Looking through them reveals a startling lack of diversity. Ikari, Gurotesu, and Manboo were also Water/Steel by the time of Spaceworld ’97, but they probably weren’t originally designed that way, especially since Steel wasn’t formally introduced as a typing until Skarmory. Ikari might have been the first idea for a Steel Pokemon, but given that Manboo and Gurotesu have their own evolutionary lines in the Korean Index, its very unlikely they were originally anything but pure water. Meanwhile, Qwilfish and Quagsire, while being dual-typed in the final games, were pure water Pokemon by Spaceworld ’97, which probably means the proto-Qwilfish that appears earlier in the Index was also pure water type. Slowking was Water/Psychic, but it didn’t really expand the diversity of water Pokemon you could catch, since it was attached to Slowpoke. Remoraid was pure water (though making it the “beam” Pokemon it did find a unique niche), as was Octillery. Politoed was pure water, Marill was pure water, and while we can’t say for sure, I can’t imagine what typing the Viking Ship Pokemon would have been except for pure water as well. So even if the team had designed a lot of water Pokemon, by this point in development all of them basically fit the exact same niche. Thus, as the team got to the last few slots they had available in Period 2b, they might have decided they needed to fix the lack of diversity in their new water types. Thus, ID 380 and ID 381 were designed specifically to address that problem. While we’ll talk about Mantine in just a second, let’s look at exactly what’s going on with ID 380. The first thing that jumps out about 380 is the drills on its arms. While we can’t know for sure the purpose of these drills, they seem designed specifically to signal that this guy is more than just a water type. While the drills could be related to the rock type, it seems more likely that they indicate that 380 can burrow into the ground, and thus that it was dual typed as a water/ground type. Up until this point, Pokemon hasn’t had a water/ground type, though Omanyte and Kabuto were Water/Rock. If the team was trying to explore new design space, then it seems reasonable for them to find an unused typing combination and design a new Pokemon to that typing. That’s what we also see in the case of Mantine, and as I’ll talk about much later (at the very end of the Index), there’s reason to believe the team was considering making Quilfish Water/Electric, another unused type. So it isn’t a stretch to think that this is the origin story of ID 380. However, even if this was the inspiration, it’s got to be a bit more complicated. Notice how the back sprite of 380 conspicuously lack any drills, 380's most defining feature. While we’ve seen other instances of back sprites in the Korean Index that don’t match the front sprites, this is by far the most obvious case of that discrepancy in the entire index. Presumably, the lack of drills on the back sprite means that there was an earlier design without drills that got overwritten before 380 was scrapped. This could mean a few things. First, it could mean I’m wrong, and that 380 was initially designed not to a mandate to create a diverse water type, but simply brainstormed as another generic water type before being repurposed. Secondly, it could mean that 380 was always meant to be Water/Ground, but the first pass at its design didn’t make that obvious enough: it might have looked just like a normal squid and so the designer went back to it and added the drills. The third explanation, and the one that seems most likely to me, is that ID 380 wasn’t designed whole cloth to fit the need of a Water/Ground Pokemon: instead, it was an old design from Generation I, copied into the Korean Index and then repurposed to be a water/ground type. Under this theory, the team was asked to find other designs of water Pokemon they could use and make sure those new designs had the potential to be dual typed in some way. One member of the Pokemon team—likely Nishida—went in one direction to design Mantine, while another looked back through their old designs and came across two squids originally dropped from Red and Green. We have their backsprite designs, which you can see here: While we don't have any idea what these squids looked like from the front, RacieB has made fan art that approximates what they could've been: In Generation I, these were probably evolutionary relatives of each other, though its unclear which evolved into which. They were probably dropped for redundancy: Tentacool already filled the “sea creature with tentacles” theme and Omanyte already fit the idea of a sea creature with tentacles living in a shell. But while flipping through these old designs, someone on the Pokemon team may have realized that while they were redundant in Generation I, a squid Pokemon would be completely unlike what they had designed so far for the sequel games. Furthermore, it might have occurred to the team that the Squids could easily be repurposed into a dual type: notice how the first backsprite’s spiral shell resembles a drill, and this might have inspired the team to make that a core component of their updated squid. Thus, a designer probably copied a drill-less version matching the original design of the right squid first and then edited it to reflect the drills. The design probably didn’t last long enough for them to ever get to redoing the back sprite. If you look closely at 380’s front sprite, you can tell the drills were added hastily and somewhat sloppily to the original sprite. First of all, look at the way the drills are just straight lines drawn over the tentacle/hands. The darker shading of the left drill doesn’t match the shading of the tentacle at all, and the lines of that drill don’t match it’s companion at all. They don’t even connect to the tentacles at straight angles, but are ever so slightly off. Beyond its arms, look at the drill on its head: not only does it bend at the tip, much more like a natural head of a squid than like a real drill, but the two fins on its sides don’t really connect logically to the side of the drill: how would that even work when the drill was spinning? As far as 380s head is concerned, it looks like someone drew in the drill on an existing sprite, but didn’t actually draw a new outline for its head to accommodate the drill design. It’s pretty clear overall that the drills were added quickly to an existing sprite, possibly as a proof of concept of the idea to make it ground themed. Now, there are reasons to think 380 isn’t related to these back sprites. First, you might ask which of the unused Squids 380 was based on? It doesn’t seem to be either, as its head and body are proportioned differently from either. If it was an updated version of the squids, 380 was most likely an amalgamation of both designs, rather than just one or the other. The second objection is that 380’s tentacles don’t look anything like those of either squid: it’s got short stubby tentacles, while they have long, tentacool-like ones. However, maybe that’s the reason 380 got smaller tentacles: to differentiate it from tentacool. Furthermore, we know that designs brought back from Red and Green underwent heavy transformations before they were transferred to the Korean Index: just look at how different the unused elephant’s initial design was from ID 309, or how much Politoed changed from it’s origins as Missingno. 84. ID 380 certainly doesn’t look any more drastically different from the squids than those Pokemon do. So why did the team decide to use Mantine in the end, rather than 380? We can only speculate. The choice could have been about balancing reasons: maybe in the initial Spaceworld ’97 version of the game, there just wasn’t a good environment to find a water/ground Pokemon, while Water/Flying fit into the world more easily. It could also be that the team liked the new design of Mantine more than they liked this hastily pulled together reuse of a previous design: Mantine would have looked a lot more fresh and original to the team than 380. Maybe it was luck: whoever designed the drill squid may have begun to work on other projects, and thus when it came time to choose between it and Mantine, Mantine had the more developed and cleaner sprite. We’ll unfortunately never know the truth. In the final game, the team solved the lack of water Pokemon diversity by adding secondary types to other water Pokemon. By Spaceworld ’99, Quagsire and Qwilfish both got a secondary type. Quagsire’s typing could be because the team liked the idea of introducing the heretofore unknown Water/Ground typing, even if they ultimately rejected 380. Certainly, Quagsire doesn’t look obviously like a Ground type, so the change must have been for more reasons than its flavor. In the end, the extra ground moves and typing in Quagsire might be the last legacy of 380: having proved that a Water/Ground Pokemon would be a neat addition to cover new Ground, the team found another Pokemon better able to fill that design space. ID 381: Mantine (Haneei) Mantine started life looking very different, but the core concept remained more or less the same in the final. Probably designed around either the brief “create a water/flying Pokemon” or the broader mandate, “create a water Pokemon with an unusual second typing,” Mantine’s original design seems closely designed to answer that exact question. By the final game, they’ve toned down Mantine’s associations with the flying type , but it gained an interesting relationship with Remoraid which gave it a unique twist. Like I discussed in the opener, ID 380 and ID 381 were probably designed to compete for the same slot in the Index. While ID 380 was probably derived from pulling an old design from the archives and improving it, Mantine was a fresh idea. In its earliest form, Mantine had fins styled after feathers, an obvious start if a designer was trying to imagine “what would a water type look like if it could fly?” The decision to use a manta ray as the base inspiration also makes sense from this foundation. A Manta Ray has large fins that seem akin to wings, and overall looks like a living kite, further giving Mantine a thematic link to flying Pokemon. It’s original name, Haneei, which was used for almost the entire time in development, even means “Winged Manta Ray,” or “Jumping Ray,” showing that the inspiration for this guy wasn’t exactly nuanced. (Credit to @RacieB) The original design also drew a little more obviously from Manta Rays in that it had a sinister face on its back. It’s meant to mimic the elaborate designs on the backs of real manta rays, which often look like faces, or at the least a Rorshach test. This is a striking design element: this evil-looking pattern visibly contrasts against the graceful, more angelic bottom half. Sadly, this trait was ultimately watered down, not unlike a lot of ther Gen II designs, not sticking around past Spaceworld ’97. The final version of Mantine still has two dots on its back, possibly a leftover or an evolution of its original design's evil face motif. This face also reminds me of Spinarak and Ariados, which were also designed in Period 2b and originally both had a face on their back as a central design element. Maybe the same developer made Mantine and the Spinarak family; maybe they liked the design on Mantine and it inspired them to try it on a creature who fit the idea better? ...I mean, the face looks very similar to Ariados's Spaceworld '97 design. When the team redesigned Mantine, the team completely changed its look. Like most of the redesigns the team undertook in the transition from Spaceworld ’97 to the final, this one made Mantine’s core properties subtler and more nuanced; unlike most of those redesigns, this one actually made Mantine resemble a real life animal more, rather than abstracting away from it. The new Mantine is less sleek and now looks like a grey Manta Ray; except for a vague sense of it soaring through the air in its sprite, any indication that it’s flying--like the defining "feathers" of the SW'97 design--are now gone. Mantine’s also now grown some horns/antennae on its head, and interestingly enough, torpedoes under its wings. Maybe taking a page out of Octillery’s book, the team decided to base the new Mantine vaguely off of a submarine, further divorcing it from it’s original origins as a showcase for a water/flying type. It could also be a fighter jet, since it still kept its Flying typing and even gained Wing Attack, a flying-type move, by the time of the final.
Speaking of its moveset, almost all of Mantine's moves were changed in the redesign. None of them really changed it's flavor but instead just fit its body type more. The new Mantine doesn’t pound or thrash at its enemies, but instead tackles and slams into them with its bulk. The new redesign isn't a drastic rethinking of Mantine, but even though each piece is small, it adds up to an almost completely different Pokemon. My favorite bit about Mantine is how the designers weren’t really sure what it’s face should look like. In succession, the team tried out a Mantine with a giant grin (my personal favorite), then one that looks like it's screaming, and then finally a smaller, cartoonish, closed mouth with puffed cheeks. The grin is… a lot, design-wise, and while it may have been deemed to be too human-like, the "scream" face is a bit too bland and realistic for its own right. Personally, I think it was a bad choice to get rid of these unique facial designs. The final Mantine seems to lack the ''je ne sais quoi'' that the earlier iterations had. The last change to Mantine was to its "torpedoes." In the initial redesign of Mantine, he had gained some grey torpedoes/missiles just under his fins, to go with the submarine/jet aesthetic. It looks like the very earliest designs for these were simple torpedoes; however, by June 1999, each of them has gained two tiny black dots, resembling tiny eyes. Who knows what the design team’s intention for the torpedoes was at this point; the two dots are barely visible and might not even have been intended as eyes, or perhaps Remoraid was supposed to "evolve" in a way similar to Slowbro's Shellder. The torpedoes stayed like this until right before the release of the game: even by Spaceworld ’99, three months before release, they were still torpedoes. Only in the final game were they replaced with a single Remoraid hanging from it’s left wing. While it still somewhat resembles a torpedo, the connection is much less obvious. What happened? My guess is that the torpedoes disappeared for the same reasons that Octillery stopped looking like a tank and Remoraid became less obviously a gun. Fears of parents objecting to weapon Pokemon, especially in the wake of the Columbine shooting, could have contributed to the team deciding to back off of any designs that resembled real world war materiel. All three lost their weapons designs in the same span of time (three months before release), further making me think the redesigns were all linked. However, while this hurt Remoraid and Octillery’s overall design, it ended up helping Mantine. For the first time since Slowbro, the team designed a Pokemon that had a relationship with another species depicted in its sprite. It’s an underused idea, and one that makes Mantine memorable; in Generation IV, they even made the proximity of a Remoraid integral to evolving its baby form, Mantyke. Interesting enough, the first time this relationship between Mantine and Remoraid is referred to isn’t in Mantine’s sprite. Instead, it shows up in Remoraid’s Spaceworld ’99 Pokedex entry: “It uses its modified, suction cup-like dorsal fin to cling onto a Haneei, with which it shares leftovers.” The final Mantine Pokedex entry makes reference to this same idea, but it was put into the game three months after this Remoraid entry. At the time the team had written this Remoraid entry, Mantine still had the old torpedoes present in its sprite. This either means that the team hadn’t yet updated the sprites (but were planning to), or more intriguingly, they hadn’t yet come up with the idea to depict Remoraid on Mantine’s sprite. It’s possible that whoever wrote this Pokedex entry just had a fun idea for relationship between Mantine and Remoraid, and the team liked it so much they updated the sprite to match. Either is possible, but the latter possibility is interesting to me because it reflects a more freeform, improvised creative process, like the process we know was the case for the development of Red and Green. Like I said in the above entry, ID 380 was never reworked and was dropped from the games before it could live, while Mantine was iterated over and over until the team got it just how they liked. Again, this could be luck, or it could be because Mantine had a champion in the design team that the squid lacked. Either way, Mantine was probably the right choice: it’s relationship to Remoraid is very memorable and it’s a likable Pokemon that I think many people have fond memories of. |
AuthorMy name's Aaron George, and I'm both a historian and a fan of Pokemon, especially of development. Reach me at @Asmoranomardic ArchivesCategories |