Era III Overview
The final stretch of the Korean Index is perhaps the most focused, and the most coherent. If Era II ended with a rough draft of a complete Pokedex, Era III was clearly a period of development in which the team was refining that Pokedex into a usable roster. Era III is heavily focused, for instance, on making connections between designs. 35 of the 41 designs in Era III are evolutions of existing designs (either earlier in the Korean Index, or from Generation I). Of those, the last 31 designs in Korean Index, with the sole exception of Togepi, form a string of unbroken evolutionary relatives. In addition to that, 27 of these 41 designs were meant to extend Generation I evolutionary lines, and thus create connections between the first games and this new sequel. Very few Pokemon designed at the end of the Korean Index were completely new ideas; instead, almost all of them purposefully tried to build on existing creations.
In addition to that, only three of the 41 designs in Era III don’t appear in Spaceworld ’97. This shows two things. First, it means that all of these final 41 designs were worked on and developed into usable designs, unlike Periods 1a, 1b, 1e, and 2d, which felt much more like brief sketches of ideas. More importantly, it probably also indicates that these were all made right before Spaceworld ’97. After all, if basically all of these designs appear in Spaceworld ’97, the team probably didn’t have time to replace them with anything: these were probably the last designs meant to finish off that build of the game. Thus, the patterns we find in Era III are some of the best evidence we have that the Korean Index is somewhat chronological. That so many of these designs were used in the 1997 build suggests that they were made very close to the date of the Spaceworld event.
In fact, the evidence we have is that Hitmontop was very likely designed, at the earliest, in April 1997; since Hitmontop is at the beginning of Period 3a, we can likely date the rest of the Korean Index from this date. The Spaceworld '97 build of the game is dated November 15th, 1997, which means that all of Era III was made in just that eight months before the SW'97 build. That's not a lot of time to create a large chunk of the roster!
Era III cleanly divides into three sections. Period 3a, from 408 (Mareep) to 417 (Sunflora), was the last gasp of the designers making wholly new creations, rather than adding on evolutionary relatives to other families. Some of these are evolutionary relatives of other Pokemon, but four of the ten are unique. This is also the last section to have designs that aren’t found in SW ’97; saying that, even these designs are more interesting than those of Period 2d, and it's very likely that all three were cut from Spaceworld ’97 at the last minute.
Period 3b, from 418 (Turban) to 438 (Para), is entirely composed of designs that add an evolutionary relative to a Pokemon from Generation I (Turban is an interesting pseudo-exception and we’ll discuss it when we get to it). This is probably the most disciplined part of the entire Index—with only one possible exception to the pattern in the entire sequence—and my suspicion is that this entire group was designed very quickly as part of an initiative to create as many links to Generation I as possible. Interestingly, babies make a return in this sequence; there’s also a focus on giving underwhelming single-stage Pokemon from Generation I a powerful second form, to make them more relevant.
Period 3c, from 439 (Flaafy) to 448 (Tyrogue) still has some designs that extend the evolutionary lines from Generation I, but the focus has switched to instead fleshing out Generation II lines. These last ten designs were probably the final tweaking of the roster the team did before the end result of Spaceworld ’97: these designs were probably quickly produced to fill last-minute gaps in the Pokedex. On top of that, Period 3c has one strange exception—Togepi—that wasn’t designed as an evolutionary relative. Instead, it was probably snuck in at the very last second of design, so the team would have a cute, lovable, mascot character that they could use to get fans excited for the sequel.
Period 3a: 408 (Mareep) to 417 (Sunflora)
Like I noted above, Period 3a is a bit of a mix. Like the rest of Era III, Period 3a focuses on evolutionary relatives of Generation I Pokemon; unlike the rest of Era III, this section doesn’t only cover evolutionary relatives, but still introduces new designs as well. It also, unlike the other sections of Era III, has a few unused designs: Proto-Dunsparce, Anklosuarus Seal, and Sato the flying fish. Saying that, it’s obviously a divergence from Era II, and Period 2d in particular. Despite featuring new designs, the additions in Period 3a are not introducing new gameplay mechanics, unique type combinations, or Pokemon designed to fill some gameplay niche, like was distinctive of Era II; instead, the additions here seem to expand upon previously introduced concepts. Even the three designs in this Period that are unused in SW ’97 are drastically more complete than the ones in Period 2d: not only are they (mostly) higher quality sprites, but we even have a name for one of them.
Despite the lack of focus in Period 3a, these ten Pokemon are very clearly the beginning of the larger themes of Era III: evolutionary families and expanding connections between Generation II and Generation I. Era II introduced all the new concepts that Generation II would use; Era III, starting here, defines exactly the direction the team would take these concepts.
ID #408: Mareep
It’s Mareep! Everyone’s favorite sheep! Or at least mine.
Mareep is adorable, a slam dunk of a design. The design team clearly agreed with this assessment, because although it got a palette change, and its sprite was touched up, Mareep remained pretty much the same (static, you might say) from the moment of conception to the final game. Still, there’s one key mystery worth being discussed with Mareep.
But first, basics. Mareep’s a sheep. The idea of making it electric type probably comes from the idea that wool can generate a static shock if rubbed together; thus, a sheep with tons of wool could conceivably have its own electric field. It might also come from the famous Phillip K Dick book, “Do Androids Dream of Electric Sheep,” a story set in a dystopian future in which most animals are dead and the only way to tell apart a human from an android duplicate is that humans can feel empathy for animals. As much as I’d like Mareep to have a weirdly-dark origin story, at best the designers were aware of this sci-fi title and took it literally.
Mareep’s original Japanese name was “Pachimee,” which, according to TCRF, is a portmanteau of Pachipachi (Spark) and Mee (Bleat). A very cute name that perfectly describes Mareep. Sometime in early 1999, when development on Gold and Silver restarted, its name turned into Mareep. This name is less clear; it’s definitely a portmanteau of sheep and something else. Bulbapedia suggests Mary from “Mary had a little lamb” but I think “Merry” like it’s happy is more likely. Or maybe Maa is another transliteration of it bleating, like “Baa”. Mareep is also an anagram of Ampere, a unit of measurement for electricity, so maybe they chose the “Ma” part to make the anagram work. Whatever the reason, it’s still a cute name.
Mareep’s Spaceworld ’97 backsprite doesn’t match its frontsprite, a minor difference that I find interesting because it implies there was an earlier version of this frontsprite. In the backsprite, Mareep has little pink hooves on each of its feet, while in the frontsprite its legs are simple black stubs. The back sprite is also noticeably lower quality, in that the proportions are odd and it isn’t visually obvious what the tail and the feet are supposed to be at first glance. Clearly the hooves didn’t work, because the designer cleaned up the frontsprite. From there, it basically stayed the same the rest of Mareep’s life.
The big unanswerable question about Mareep is its original relationship to Ampharos. We know Ampharos was one of the first Pokemon created for Gold and Silver, and the team quite liked him. But Ampharos is clearly not a sheep: depending on how you interpret the information we have about Ampharos, it might have been originally conceived of as a space alien or a dragon. So was Mareep designed to be its evolutionary relative, or did that come later, when Flaafy was created? And was the plan always for it to have three stages, or was there a brief time in which Mareep and Ampharos were a two-stage family?
Since Era III is almost entirely made of evolutionary relatives, the evidence seems to point to the idea that Mareep was always made to be related to Ampharos. But it isn’t conclusive. Remember that while Periods 3b and 3c are only evolutions, Period 3a still has its share of new single stage creations, such as Sunflora, Murkrow, and Sato. On top of that, Mareep’s palette, a pink color, looks quite different from Ampharos, and even with both of them lined up it’s hard to see how one evolves into the other. If we change Mareep’s palette to Ampharos’s, it really helps to make them look a little more related; maybe this was Mareep's original palette, and is was only made pink when Flaafy was created? Mareep does look quite good in that yellow palette, and it went back to a similar yellow palette before the final game.
(All speculative recolorings by @OrangeFrench!)
What really convinces me that Mareep was made with Ampharos in mind (and not as an independent Pokemon first) are its ears. While everything else about Mareep and Ampharos look completely different, they share the exact same ears. While on Ampharos they look like some sort of horns, on Mareep they look a little more like a strange exaggeration of a sheep’s floppy ears. Ampharos’ ears are distinctive enough that I can’t really put this down to coincidence. It could conceivably be the case that Mareep had an earlier sprite without these ears, and they were added on later on when Mareep was redesigned to evolve into Ampharos. If that were so, we'd have no evidence, however, because any earlier sprite without those ears would have been overwritten by the SW'97 sprite in the Korean Index.
Wait a second. What about that backsprite we looked at earlier, with the hooves. That sprite was from an earlier iteration of Mareep. What did its ears look like?
It... didn't have ears? Oh that's interesting...
Instead, the back sprite just seems to have a circular protrusion where the conical ears should be. Could this be a remnant of an earlier design of Mareep, before it was made into a relative of Ampharos?
Or is the angle just a little bit strange and we just can't see the ears?
It's also hard to see, but Mareep's tail--which looks just like Ampharos' in the front sprite--instead is a different shape here, like a lightbulb. It's another way this backsprite shows a less Ampharos-like earlier design.
I’m not sure what to think. Draw your own conclusions.
To add to this debate, let's also take into account Atsuko Nishida's personal style in designing evolutionary relatives. Both Mareep and Ampharos' sprites were clearly designed by her, and Nishida has said in interviews that she likes it when the third stage of an evolutionary family is completely different from the first two stages; she’s said she likes to surprise the player with how the Pokemon turns out. The archetypical case of this is Dragonair and Dragonite: Nishida specifically designed Dragonite to look completely different, so it was a surprise for the player to end up with this dopey dragon. So it would be completely within her regular style for Mareep and Ampharos to look so different.
If Mareep was designed to be related to Ampharos, my hunch is that Flaafy was also already in the works, but it just didn't quite have a sprite yet. Flaafy’s not much further in the index, and given the huge gulf between Mareep and Ampharos, it makes sense that even at this stage the team knew a three stage family would work better aesthetically. There isn’t really any evidence either way, just a guess based upon the sprites themselves.
After Spaceworld ’97, when development restarted, Mareep’s sprite got a makeover and a palette change, but its design (and its pose) stayed basically the same. Now its face was blue and its wool was yellow, subtly making it look more similar to Ampharos. It also got an updated moveset after Spaceworld ’97, and its Pokedex entry in Spaceworld ’99 got mildly updated before the final; however, both changes were minor, and didn’t really change anything about Mareep.
Sometimes you create perfection, and there isn’t anything to fix, I guess.
ID #409: Hitmontop
Adding a third brother to the Hitmon family line was a great idea. It doesn't feel like the most obvious move Game Freak could have made--Hitmonchan and Hitmonlee feel like they cover the binary between fighting styles pretty well, and neither are super popular--but the fact that a third Hitmon is so unexpected is part of what makes Hitmontop work. Pokemon loves trios—the Legendary Birds, the Eeveelutions, whatever the Magmar line is—almost as much as it loves version exclusive duos, and turning a duo into a trio is something that they otherwise didn't try in Gold and Silver.
I also like that unlike Slowking, or Crobat, or Politoed, Hitmontop wasn't just an evolution of a Generation one Pokemon. Though the team eventually made it related, evolutionarily, to Hitmonchan and Hitmonlee, the idea of making a connected Pokemon like Hitmontop is an underexplored way of building on an earlier game. I also like how the design of Hitmontop fits its concept: if Hitmonchan was had high attack and Hitmonlee had high defense, then Hitmontop was supposed to be right in the middle of both of them, with balanced stats. And what could be more "balanced" than a top?
I’ll just start with it right from the gate: I think that Hitmontop wasn’t initially created to serve as an alternative evolution for the Hitmon family. Given that Tyrogue is the very last Pokemon in the Korean Index, my suspicion is that Tyrogue was a last minute idea to unite all three into one family, rather than let them be a trio like the Legendary Birds or the Jynxmarbuzzes. One reason to think this is because the evolutionary method by which Tyrogue (Gong in Spaceworld ’97) evolves into each of the three Hitmons is clearly incomplete in Spaceworld '97. In the final game, Tyrogue evolves based on which of its stats are higher: attack gives it Hitmonchan, defense Hitmonlee, and if they're the same, it becomes Hitmontop. But in Spaceworld '97, this evolutionary method was not implemented. There, it evolves into all of them at level twenty, at the same time: first it’ll turn into Hitmonlee, and then into Hitmonchan, and then into Hitmontop, but only if you cancel each evolution. Obviously, this is unintuitive and lacks any clear flavor; it was likely just a placeholder before the team figured out exactly how they wanted Gong to evolve.
Now, this isn’t conclusive by any means: the team could have meant for a Pokemon like Tyrogue to exist as early as the creation of Hitmontop, but just didn’t manage to get to designing it before the final push right before the SW ’97 build. After all, Era III almost entirely made of evolutionary relatives. But given the way Hitmontop appears earlier than Tyrogue, and how it’s found in Period 3a and not later (where we do still have a sizable amount of designs which weren’t evolutions), it seems plausible that it was originally designed just as a new connection to Gen I, rather than explicitly as an evolution.
While it was a good idea to add a third Hitmon and make a trilogy, Hitmontop seems like it was a bit difficult to design. First of all, the team didn’t fall back on the same naming scheme of Hitmontop’s brothers. In English, Hitmonchan and Hitmonlee are named after famous martial artists—Jackie Chan and Bruce Lee, respectively—and in Japanese Sawamular is named after the kickboxer Tadashi Sawamura, while Ebiwara was named after Hiroyuki Ebihara, a boxer. As far as I know, there’s no famous martial artist named “top,” and Hitmontop’s Japanese name—Kapoera—is named after a style of martial arts from Brazil named Capoeira. It’s a fitting name, since practitioners of Capoeira tend to fight by doing acrobatic stunts, like kicking from a handstand position. Eddie, from Tekken, fights with a Capoeira style; why didn't we get Hitmoneddie?
Then there’s Hitmontop's sprite. The Spaceworld ’97 sprite for Hitmontop is a really fascinating concept. Shaped like a top, it has three legs on its head that seem to be spinning around and a needle of a top below its body. But it also has a devious look on its face and is holding its hands like it’s meditating while its legs spin (Notice that in its SW ’97 moveset, Hitmontop even learns the move Meditate, but doesn’t learn the move by the final). It’s also, strangely, got another eye around where its belly-button would be, giving it the appearance that it might have a second, upsidedown, head that it’s spinning on. Or maybe the eye is a mystical third eye, to go along with the meditation theme? Also notice that on its frontsprite, Hitmontop’s legs are equidistant from each other, while on the backsprite they seem to be clustered around one side. Maybe that’s to give the sprite a sense of movement? Or maybe this was an artifact from an earlier design?
It’s a really complicated concept for a fighting Pokemon: it has two faces, it doesn’t walk on its legs, it seems to be meditating, etc. Why does Hitmontop have feet on its head, and if that’s how its supposed to be oriented, how does it move around? It also looks nothing like the other Hitmons, except for its palette and some similarities on its feet-talons and its upside down head. It all feels too high-concept for me.
It looks a little more like the other hitmons when you turn it upside down, but even then it's a big of a stretch:
It’s no surprise to me that Hitmontop was heavily reworked before the final version of Gold and Silver, given its strange design. You can see the team were thinking this through by June 1999, the next time we have a sprite for Hitmontop. Its palette is weird, and I’m not sure exactly why that is: it seems to be using the palette that Tyrogue uses in the final game. Which is strange because Tyrogue is using this palette already in June 1999, but with its old design, and it looks wrong on that sprite too. So I’m not sure where that palette initially came from, and it's strange Hitmonstop still uses that palette in the final, even though it doesn't match the palette for the other two Hitmons at all.
But that’s all an aside, beside the point. The important thing is to notice how the team have changed Hitmontop’s face in this sprite:
It’s subtle, but now its face looks more like the final version, and it’s clearly spinning upside down! You can see that Hitmontop now has an open mouth and that its extra eye has been moved up on its body. Hitmontop is no longer meditating, but now it also doesn’t seem to have legs sprouting from its head for some reason. By the final version of the sprite, Hitmontop got a complete revamp—probably by a new artist—but the design was still moving in the same direction started earlier that year. By the final, Hitmontop no longer looks like a bizarre top, and instead was redesigned to be a Hitmon which can spin upside down to kick its opponents. Its legs and body now look humanoid, and its head looks like the rest of the family. It isn’t unique because of how its body is composed anymore, but because of how it fights.
There’s another strange oddity regarding Hitmontop’s visual design, but to understand it, we have to go back before Spaceworld ’97. About seven months before the Spaceworld ’97 demo was shown to the public, a Japanese video game magazine, Microgroup Game Review, featured a cover with new Sugimori artwork, plus an interview with Ken Sugimori about the upcoming Pokemon 2. The artwork looked like it featured new Pokemon designs next to Pokemon trainers, but Sugimori denied that was the case at all. In the interview, Sugimori explained that the three Pokemon-like creatures featured were not Pokemon designs, but something similar to Pokemon. The cover showed “another world, like Pokemon, but not Pokemon.”
But that’s clearly a lie, because if you look at the cover, Tyranitar is right there, clear as daylight:
(Credit to Dr. Lava for the high resolution scan of this artwork)
This magazine cover has become infamous among Pokemon fans, and I’ll talk about it a lot more when we eventually get to Tyranitar. But the reason I bring it up now is because of the Pokemon to the bottom-left of Tyranitar. It’s a bit of a weird design, but it looks a little like a merger between Clefairy and Hitmontop. When you consider the original Spaceworld ’97 design of Hitmontop, the similarities are even more unarguable. So what’s going on here?
The best explanation is that when Sugimori drew these designs, he meant what he said: these were just concepts that could have been Pokemon, but at the time weren’t planned to be Pokemon. However, as development continued, the team would have needed more designs, and Sugimori may have reached back to this magazine cover for new ideas. Sugimori says later in the same interview that he gets the ideas for his designs by looking back at his earlier concepts and constantly revising them; it seems absolutely possible that this design inspired the team to come up with Hitmontop, or that they came up with the idea of a new, "balanced" Hitmon, and Sugimori's mind went back to this top concept. If that were the case, we know that Era III can be dated past April of 1997. This, interestingly, means that all the rest of the designs in the Korean Index were done in a hurry, probably all within eight months or even less. No wonder Hitmontop looks odd: the team probably had time to get a basic concept sprited, and then had to move on to other designs.
So this strange design on the cover, then, was possibly an inspiration for Hitmontop, but reworked drastically when the team wanted to create a new Hitmon. Despite the connection, it's unlikely that the cover design was ever programmed into the game.
The only other thing worth noting is Hitmontop’s movepool. Honestly, it’s not very focused in either Spaceworld ’97 or in the final. In both, Hitmontop has a smattering of moves that the other Hitmons also learn: Pursuit, Rolling Kick, Detect, and Focus Energy. On top of those, it has Rapid Spin—quite flavorful even if it’s a bit of a bland move—and Triple Kick, which was a signature move created just for Hitmontop, to match the way its three head-legs would spin and hit the opponent in succession. In modern games, Hitmontop has a bit of a priority-move theme going on: it has a lot of moves that go first, like Quick Attack, Sucker Punch, Helping Hand, Quick Guard, and Counter, and some that go last, like Revenge. But this theme doesn’t really exist yet. Unlike Hitmonchan, which gets a flexible assortment of punching moves, and Hitmonlee, which gets four kick moves of increasing power, Hitmontop lacks a lot of identity, being a mostly middle-of-the-road Pokemon with the signature Triple Kick. Which makes sense given that its supposed to be the balanced of the three brothers, but it feels a little bit lame to play. Its clear that Hitmontop’s strange concept was hard to execute as a theme, even though the team had a lot of time to make it work.
Overall, I like Hitmontop. I used to like its Spaceworld ’97 design a lot more, since the final looked kind of bland to me, but on reflection, I think they made the right choice: a weird meditating creature with legs for hair was too strange a theme for this Pokemon. Hitmontop is stronger for the redesign, even if it became a bit blander.
ID 410: Betobebii
We’ve seen baby Pokemon before this point; at this stage in the Korean Index, we’ve seen Elebebii, Pichu, Cleffa, Puchikoon, Gyopin, and Mikon. But Betobebii’s something new. It demonstrates how the team, in Era III, had decided to move in a new direction with baby Pokemon. While almost all the baby Pokemon that follow from Betobebii didn’t make the final cut, Betobebii shows that baby Pokemon don’t have to be cute or iconic Pokemon; the team was willing to try to make almost any two-stage evolutionary family into a three stage one.
Previous to Era III, you can basically divide the baby Pokemon designs into three groups. Pichu and Cleffa are smaller, cuter versions of already cute mascot Pokemon; their purpose is presumably to build upon the popularity of Pikachu and Clefairy by creating new variations in the sequel. The second group was the most interesting sequence in Period 2c: Puchikoon, Gyopin and Mikon were all discarded designs from Generation I repurposed into baby Pokemon for Generation II. Thirdly, Elebebii falls into a strange category all on its own: probably not designed to be a baby Pokemon originally, Elebebii was later retconned into one once the team decided to create Magby and Smoochum.
But Betobebii doesn’t fit any of these categories. Given its cuteness and where it appears in the Index, it was clearly designed to be a baby Pokemon, unlike Elebebii. It’s not a design discarded from Generation I, at least as far as we know, and given the elements in its design that make it cuter than Grimer, it seems unlikely it was made for the previous games. And thirdly, Betobebii was not designed to take advantage of the popularity of an existing ‘mon: Grimer is rarely listed as anyone’s favorite Pokemon and it hardly had any sizable appearance in the anime.
Instead, Betobebii is the first in a line of baby Pokemon which seem to have been created just to create a more sizable and diverse lineup of baby Pokemon in the new generation. Era III has eleven separate designs for baby Pokemon in it, and of those, only Igglybuff and Koonya are related to particularly popular Pokemon. Instead, the team seems to have found as many one-stage and two-stage families from Generation I that could potentially have another relative, and created whatever design they could think of.
(ID 400, if I’m right that it was designed as a Drowzee pre-evolution, might actually be the first pioneer of this new type of baby Pokemon; if so, Betobebii is an acknowledgement that this was a good idea, even if the team didn’t like ID 400’s design.)
Why would the team start designing babies out of seemingly random Pokemon? First, baby Pokemon could have proven popular with the team working on the games, and thus they decided to mine this design space to see what they could come up with. Secondly, maybe it was an attempt to revitalize some of the less popular 'mons from Generation I: if they were unpopular, maybe a cute baby version would give them a fanbase? Or third, the team was getting close to end of development of the Pokemon roster (remember, it was probably only seven or six months until Spaceworld ’97), and the team may have been making a lot of baby Pokemon because it was easy to design smaller versions of existing creatures. While the Pokedex was theoretically full after the end of Era II, the team could have already grown unsatisfied with a lot of those earlier designs; baby Pokemon might have been quickly iterated to fill those slots, at least until a better idea came around.
That could partly explain why so few of these baby Pokemon survived to the final game. Of the eleven baby Pokemon which appear in Era III, only four of them survived past Spaceworld ’97 (none of the three that appeared in Period 2c did). This could suggest that these were all concepts that the team was only lightly devoted to, and that most of them were seen as easily cut when a better idea came along. Another explanation is that the team discovered that, despite how cute and adorable baby Pokemon are, they’re don’t really contribute anything to the gameplay of the games. They aren’t found in the wild (mostly, although Mikon is found in some encounter tables in Spaceworld ’97), they are incredibly weak, they share the same movesets with the adults (making them redundant), and they evolve so quickly that most players wouldn’t have babies in their team for long.
The biggest problem is that these flaws are all baked into the idea conceptually: babies, by design, have to be bad in battles and somewhat redundant. After the team restarted development in early 1999, they may have simply decided that the slots devoted to baby Pokemon could be better devoted to new ideas that would flesh out the world they were building.
Betobebii survived slightly longer than the other deleted baby-mon’s. This guy’s sprite can still be found hanging around by June 1999, with ever-so-slightly tweaked eyes, and a completely gray palette. At that point, it doesn’t have a name, or a moveset, and that gray palette makes it feel like they didn't bother coloring the guy! It’s in the Pokedex slot that was eventually filled by Magcargo and then Slugma, and so my guess is that the sprite is just there as a placeholder; they had scrubbed the data clean of Betobebii but didn’t yet have sprites to overwrite it with. None of the other deleted baby-mon appear in this June 1999 sprite collection; only Betobebii. Maybe there was something special about this guy, or maybe all the other ones were deleted only after the team had come up with a new Pokemon.
Betobebii’s name, by the way, is actually pretty cute. “Betobeto” is a Japanese word that can mean “sticky;” it’s also an onomatopoeia, for the sound a sticky thing makes (maybe when you walk on it and it sticks to your shoe?). The word is used in the Japanese names of Grimer and Muk as well—Betobeta and Betobeton—so it makes sense to use it for the baby version. And, of course, the second half of Betobebii’s name is just a transliteration of the English word “baby,” making this Pokemon literally “Sticky baby.” If it had gotten an English name, it probably would have been some like “Grimy.”
(As a quick aside, I just noticed that Betobebii doesn't have quite the same palette as Grimer and Muk. Maybe that's a sign they were sprited by different people?)
I find its sprite a little bit adorable; its amazing how the team were somehow able to make Grimer of all things precious. However, there is a little bit of ambiguity about its sprite: namely, what is the dark circle at the center of Betobebii's face? Some people think it’s a pacifier in the baby’s mouth, but I feel like if it was a pacifier, the sprite would have looked more explicitly like one. It could also be a nose, possibly like Diglett’s nose, or the nose of a clown. That’s personally what I see when I look at this guy. Finally, the more outlandish among us see it as a really tiny mouth, with maybe a single tiny tooth. While I like this interpretation a lot, and it certainly seems fitting for the Grimer family, I’m also skeptical; it feels like the shading is implying the “mouth” actually protrudes outward, rather than inwards. We’ll never know, unfortunately, since I doubt we’ll never hear more about Betobebii from Game Freak.
Poor mystery baby.
ID 411: Murkrow
Murkrow’s a simple guy. His English name is a mix of “murky” and “crow,”; his Japanese name is a mix of “yami” (darkness), and karasu (crow). Can't get more straightforward than that. The feathers on his head look like a witch’s hat, and his tail is supposed to resemble a witch’s broom. He was made to be a Dark/Flying Pokemon, and he shows off exactly what a Dark/Flying Pokemon should do. Everything about Murkrow is right there on the tin: you get exactly what you’re asking for with Murkrow.
Murkrow looks like a Sugimori design to me, in that its proportions are mostly well-drawn, its concept is simple, and it didn’t go through much changes over the course of development. The main difference between the earliest Murkrow we have and the final version is that his theme was a lot more obvious in Spaceworld ’97. There, Murkrow is clearly wearing an actual witch’s cap, rather than its feathers just looking a lot like a cap. It’s also got a slightly more malevolent grin, which was toned down as well by the final.
(Credit to @Raciebeep for the great beta Murkrow art!)
Sugimori has gone on record as not liking Pokemon that look too obviously like they have human tools, or if they don't look like they'd fit in the wild. Many of his other designs from Generation II that were based on real objects-- Remoraid, Octillery, Mantine, and Delibrid, for instance—all had their themes made more subtle by the final game, so the change to Murkrow’s hat fits with the same changes that were being made elsewhere. The change to his grin fits larger patterns as well: even though Dark-type was characterized as “evil” early on in development, Sugimori has made a point to say that he didn’t think any Pokemon should be evil, since any of them have the potential to be the player’s friend. It’s possible that maybe making Murkrow a little less mean looking helped turn it into a more appealing companion.
If you look at Murkrow’s early movepool, it suggests that the team hadn’t yet fleshed out how Dark-type Pokemon should act. Murkrow’s initial moveset is all over the place, filled with “sneaky” or “dirty trick” type moves: it gets Sand Attack, Spikes, Detect, Foresight, Faint Attack, Stalker (ie, Mean Look), and Perish Song. All of these moves, to some extent, could be themed around sneaky behavior: Detect dodges an attack, Sand Attack throws sand in their eyes, Spikes punishes them for switching out, Perish Song punishes the opponent for not switching, and Mean Look is, well, mean. Detect’s an especially weird move that the team also gave to Sneasel, one of the few other resident Dark-types in Spaceworld ’97 (Twinz, Girafarig, Rinrin, and Berurun being the others). In the final, the legendary birds and some fighting Pokemon learn Detect, but not much else.
Instead of coming up with a single theme for Murkrow, it seems like this early moveset just decided to try everything. If a move seemed sneaky or underhanded, just hand it to Murkrow. Why not? It's dark type. It doesn't seem like the team got much past "Dark-type bird" in concept for early Murkrow.
In the final, Murkrow’s moveset underwent a huge revamp. It lost most of its moves except for Peck, Faint Attack, and Mean Look, but it gained Haze and Night Shade. The final Murkrow was a lot less tricky, and now its main form of attack was a Ghost-type move. The team must have trying to make Night Shade much more of a staple for “mysterious” style Pokemon: they also gave it to the Natu line and the Spinarak line in the final games. Final Murkrow seems, to me, to be much worse in combat. Which, I guess, makes sense, given that it’s unusable by the player for most of the game.
I’ve mentioned this before, but it blows my mind how Dark-type Pokemon were basically unusable by the player prior to the post-game in Gold and Silver, with the only exception being Umbreon. Like Houndour, Murkrow is only found on three Kanto Routes, after the Elite Four and way past when it would be useful to have this guy available; unlike Houndoom, which at least has the stats to be vaguely useful in the late game, Murkrow’s designed much more like something you’d get in the early-to-mid game, compounding its uselessness by the time you get it. At least Murkrow has a little more representation than Houndour: four trainers use a Murkrow in Gold and Silver, while only two bother to use and Houndour or a Houndoom. But it’s still not a lot.
Again, it feels bizarre to introduce a whole new type into the game, and then make it this scarce. Maybe they were trying to make the new type feel special if you finally found one, or like I’ve suggested before, maybe the team wasn’t confident about the balance for Steel and Dark types, and so they hid them in the end game, where they could do minimum damage if they were overtuned. A commenter, Paul, has also suggested that maybe the Pokemon team saw Pokemon Stadium 2 as the true endgame of Gold and Silver, and left a lot of rarer or more unique Pokemon in the Kanto portion of Gold and Silver so you could collect them specifically to beat that game. While absolutely possible, I don’t know of anything interviews or other evidence that could confirm or deny this theory.
The only real oddity concerning Murkrow is its placement here in Era III. It sort of fits the larger theme of Era III, in that Murkrow was clearly designed as a way to have one more example of a Dark-type in the Pokedex; I mean, looking at its sprites, Murkrow feels almost as if it was the first idea that anyone had when asked “what would a Dark/Flying type look like?” Era III seems to have been a period to expand upon existing ideas in the Pokedex, and Murkrow is doing exactly that.
On the other hand, Murkrow is one of the few Pokemon in Era III not to evolve or be designed as an evolution for something else. That’s not entirely out of keeping for Period 3a: after all, next up is Dunsparce, and four and five designs further down the Korean Index we find Sato and Sunflora, both of which didn’t evolve. Still, given how Murkrow has pretty low stats—compared to Gen I single-stage evolutions, its stats are higher than Farfetch’d but lower than Tangela and much lower than Tauros—and how it looks unimpressive as it is, I wouldn’t be surprised if the team had at least floated the idea of giving Murkrow an evolution later in Era III and just never got around to it. Murkrow, like Sneasel, Gligar, and Misdreavus, was a Gen II single-stage Pokemon that got an evolutionary relative in Generation IV, and we know that some of the Pokemon which had an evolution in Spaceworld ’97 but not in the final—Tangela and Lickitung—got a revised version of those evolutions in Generation IV. There’s no evidence anywhere for a Murkrow evolution in Generation II, but all of this leads me to speculate that the team may have briefly considered the idea.
(Credit to enumise for the art!)
That’s all I’ve got for Murkrow. Next up, everyone’s favorite snake…thing. It’s Dunsparce!
ID 412: Dunsparce?
(Note: I can’t confirm that this sprite was the sprite it had before Spaceworld; this sprite could have been updated between November 1997 and May 1998. However, given that it didn’t appear in the Spaceworld ’97 demo nor the spriteset from June 1999, I doubt the team spent much effort updating it in this period).
ID 412 is an odd one. It’s probably an early version of Dunsparce, but it didn’t appear in Spaceworld ’97, so we can’t be sure. Furthermore, if it was an early Dunsparce, then Dunsparce went through quite a journey: Dunsparce didn’t appear even in June 1999 and was nothing more than a thrown together concept by Spaceworld 1999, only a few months before the final game. So what happened here?
First of all, lets talk a bit about what this early sprite is, and what Dunsparce is based on. ID 412 is definitely based off of the Japanese mythological creature called a Tsuchinoko, which translates to “child of gravel,” “child of the earth,” or “child of the hammer.” It’s a strange creature with a ton of random mythology surrounding it. First of all, the Tsuchinoko looks like a fat snake; it’s not very long compared to other snakes, and has a fat lump in the middle of its body, almost like its digesting something. Some legends say that it will bit its tail and turn into a wheel to travel more quickly; others say that its venomous and jumps at its prey; others say that the Tsuchinoko loves getting drunk on alcohol; others say that this little snake guy can even talk, but spends most of its words lying and being generally untrustworthy. Like the Tanuki, it’s a creature that could plausibly exists (though Tanuki definitely exist, and a real Tsuchinoko has never been spotted), but both of them have so much magical backstories attached to them that a real-life version would be almost completely different from the imagined one.
One touch I like about the final Dunsparce is that in mythology, Tsuchinoko’s are only found in secluded mountains of Japan; In the final game, Dunsparce’s are an extremely rare encounter only in the Dark Cave, which happens to be a secluded mountain cave in Johto. It’s small, but its nice that you find a Tsuchinoko Pokemon exactly where you’d expect it to be.
On top of its obvious Tsuchinoko-like characteristics, ID 412 also has an eye-like pattern on its forehead, which suggests to me that the initial concept could have involved 412 having psychic or mysterious powers of some sort. It’s also a pretty well-done sprite that looks pretty finished, making it even stranger that the sprite doesn’t show up in Spaceworld ’97. It’s possible that the sprite could date to after Spaceworld ’97, but like I said above, I’m not sure why they’d bother updating it very drastically if 412 didn’t even appear in the Pokedex as late as June 1999.
ID 412 isn’t the only design from Era III that doesn’t appear in Spaceworld ’97, but it’s close: of the next 37 designs in the Korean Index, only 415 and 416 also didn’t make it into Spaceworld ’97. And 416 was probably cut from SW’97 only at the last minute: we found a revision of the sprite in a trash folder in the internal files, named Sato, indicating that it was worked on a significant amount before it was dropped.
It’s my belief that all three of these designs—Proto-Dunsparce, Xylophone Seal, and Sato—were in the Pokedex until the very last second before Spaceworld ’97. We have evidence that five designs were cut right before SW'97: there are five Pokemon in the Spaceworld ’97 Pokedex that weren't in a logical spot in the Pokedex, but were haphazardly dropped into the Pokedex in the last few slots, after the legendary Pokemon. The way these Pokemon just seem appended to the end of the list suggests that were dumped into the game at last second, as a late change to make sure some of the designers' favorites were included. These five out-of-order Pokemon were Sneasel, Togepi, Snubbull, Leafeon, and Aipom. it’s probably the case that those Pokemon overwrote the three unused Pokemon from Era III, plus two others that got cut (I suspect Stantler and ID 334) in the final revision before the demo. If not for the developers dumping those five into the game at the last second, we might know more about early Dunsparce and its colleagues.
As an aside, some people theorize that ID 415, the Anklyosaurus Armored Seal, was actually designed as an evolution for 412. Looking at them above, I can see the reasons you might think that: they were definitely sprited by the same person, they have very similar shading, and their poses look almost the same. Not to mention how close they are in the Index. Let's save this line of thinking for the entry on ID #415, but also keep it in the back of our mind.
Anyway, if Dunsparce was so close to appearing in Spaceworld ’97, then what happened to it? It’s likely that by the time production on Gold and Silver restarted in early 1999, the team had decided to create new Pokemon rather than draw from the Korean Index; the only sprites from the Korean Index that show up for the first time in June 1999 are the early Cyndaquil sprite, the Elekid sprite (and we’ve discussed what probably was going on there), and Stantler. Meanwhile, June 1999 is filled with completely new designs to replace the ones they’d already discarded: Gligar, Granbull, Ursaring, Piloswine, Tyranitar, Sentret, Furret, etc. It’s very possible that the team designing new Pokemon had been shuffled up since Spaceworld ’97 and the new designers were more interested in inserting their own ideas than pulling old discarded designs. It certainly seems like the new designs in 1999 have a different style than the unused earlier designs.
It's also possible that ID 412 was added into the game after Spaceworld ’97, like Stantler was, but I don’t think that’s likely. If that were the case, we’d have likely seen that sprite used somewhere in June 1999, even as a placeholder, but it’s nowhere to be found.
We do, on the other hand, find something odd going on in the sprite that Dunsparce would eventually fill, but it doesn’t seem to have anything to do with Dunsparce yet. Instead, Pokedex #206 is a strange, in-between amalgamation. It’s named “Mitei 06,” (Mitei means "Pending") which meant that the team hadn’t yet come up with a name. It uses the sprites of the discarded evolution to Rinrin, Berurun (again, if 412 was ever inserted into the game, I’d expect its sprites to be used as placeholders here, not Berurun’s). 206's typing is Rock/Ground, and it evolves from Pokedex #205 at level 22. Which is strange, because #205 is Shuckle, and at this time in development, Shuckle was Rock/Ground.
(Shuckle and...Super Shuckle?)
So there’s almost no chance the team were thinking about Dunsparce by June 1999; instead, they were clearly thinking about giving Shuckle an evolution! We’ll talk more about this once we get to those two Pokemon, but the long and the short of this is that the evolution idea didn’t get very far; Shuckle’s evolution never got a sprite or a name (as far as we know). I have no idea why they decided not to give Shuckle an evolution; maybe the team couldn’t come up with a good visual design, or maybe the team decided to make Shuckle into a gimmick-mon and an evolution was no longer required. But when the Shuckle evolution fell through, the team still had an empty slot in the Pokedex that needed filling.
As a slight tangent, I do want to mention that 206 did have one thing in common with Dunsparce: it shared the same palette that Dunsparce would use in Spaceworld ’99 and in the final games. That palette looks natural for Berurun, so it was probably chosen originally for Berurun and then repurposed for Dunsparce. Which is weird, if you think about it: was Dunsparce designed to fit an existing palette because they didn’t want to bother with changing it? Or did they come up with some early sprites, decide the old palette still worked, and gave Dunsparce its color scheme completely by coincidence?
Dunsparce finally makes its glorious debut into Gold and Silver in Spaceworld ’99, but it’s…odd. Dunsparce is very incomplete at this point, and its sprites are no more than jokey sketches of a Tsuchinoko with its tongue sticking out. On the one hand, this tells us that Dunsparce was still very unfinished just a couple months before the game was finalized, making it one of the very last Pokemon to be completed (alongside Sneasel and Wobbuffet). But secondly, it’s always struck me as strange that the team was using these sketchy sprites, because they already had well polished ones available in the Korean Index! Why not just go back to those sprites, and use them as placeholders while they redesigned Dunsparce?
Here's what happened: Hironobu Yoshida joined the team. Yoshida’s really important to the Pokemon franchise: he’s been working on the series since 1997, when he first joined Game Freak, all the way to the present. Outside of Ken Sugimori and Atsuko Nishida, Yoshida’s probably the most influential artist for the series, and many of the more modern designs, especially those in Generations V and VI, owe much of their sensibilities to Yoshida’s ideas. He’s absolutely important to the Pokemon franchise, but he only becomes important at the tail-end of Gold and Silver’s development.
Sources online say that Yoshida joined the team in 1997, but if he did, I doubt he was working on the Pokemon designs; instead, he was probably working on another aspect of the game. Because in interviews, Yoshida has said that his first assignment was to design Celebi, Wobbuffet, and Dunsparce. Those are three very interesting Pokemon for him to mention, because two of those were only sketches in Spaceworld ’99, and the last, Celebi, was far from its finalized design. There’s only a few other Pokemon in this state of unfinishedness: Larvitar, Wooper, Forretress, and Sunkern are rough sketches in Spaceworld ’99, and Slugma, Azumarill, and Lanturn are still very different from their final design. That Yoshida was working on three Pokemon that were at their least finished in Spaceworld ’99 hints that they were unfinished because he was only put in charge of designing them recently by that point.
Given that none of the three exist in June 1999, I think that Yoshida was brought onto the Pokemon design team between June and August of 1999, and he was given these three Pokemon to start out with. In interviews, he mentions asking what the Pokemon at the very end of the Pokedex was, and being told it was supposed to be a legendary Pokemon like Mew. That would make perfect sense, given what we know about the sprites done by June 1999: there is a slot for Celebi, but it uses a placeholder graphic and has no other information attached to it. If Yoshida joined the team about this time, that’d be the first thing he’s logically ask about.
So imagine that, in June or July 1999, Yoshida was introduced to the rest of the Pokemon designers and told to remake Twinz as a single-stage Pokemon, figure out to do with their Mew analogue, and told to come up with an idea for Pokedex #206, which no longer had a Shuckle evolution in it. Given the short timeframe he had, I suspect Yoshida would have gone back to the old designs the team had discarded prior to Spaceworld ’97. And what did he find there? The very first sprite was a nature spirit, Kokopelli, which may have matched his brief for Celebi, which was already conceived as a nature spirit. He also found a Tsuchinoko, probably made by Sugimori. And, given the time constraints, he decided to use those for inspirations.
Those old sprites for Celebi and Dunsparce weren’t used because they weren’t Yoshida’s initial sprites in the first place, and because he probably just drew inspiration from them when he needed ideas for his very first Pokemon designs. By Spaceworld ’99, he had worked enough on Celebi that he had a rough first draft that looked a bit like the old Korean Index design. More importantly for our purposes right now, Yoshida had also sketched out, in his own style, what he thought a Tsuchinoko Pokemon would look like. It looks pretty obvious to me that he based it on the old sprite. The giveaway here is the pose: although the SW’99 sprite is very different than ID 412, it’s clear Yoshida used the first as a reference for the camera angle and the pose.
I don’t know where Yoshida got the idea for Wobbuffet, but based on how sketchy its design was in SW99, it was the Pokemon of the three that Yoshida had worked the least on. My guess is that he didn't draw an inspiration from the Korean Index for it, but came up with his own idea after he had finished the other two.
I like this theory as it provides a plausible explanation why those sprites for early Celebi and early Dunsparce appear in the Korean Index. While Sugimori, Nishida, or Morimoto probably had no interest in returning to old designs by 1999, a new designer with a very short deadline, looking for ideas? That I can buy.
The relative lateness of Dunsparce’s design explains a lot about its final appearance in the games. Dunsparce is beloved and kind of infamous in the Pokemon community for being an awkward and more-or-less useless addition to Gold and Silver. It’s an incredible rare encounter, but it has pretty low stats and a mediocre movepool. It has tiny wings, which suggest that it might evolve into a majestic beast, but then it…doesn’t. On the ratio of “difficult to catch” to “usefulness in game,” Dunsparce is maybe the most disparate example we have. But if Dunsparce was an extremely late addition, I’m not surprised that the team didn’t have time to make it into something unique or helpful. With the amount of time they had, Dunsparce was probably thrown together, put on a single encounter table in the Dark Cave, and then the game was shipped out.
You can see just how late Dunsparce’s concept was worked out by checking out its early Pokedex entries. In Spaceworld ’99, Dunsparce was an almost entirely different creature!
“When it attacks its prey, it strikes them with its head so hard that it knocks them senseless.”
“It has a vindictive personality. Once it has set its sights on its prey, it will hunt them unceasingly.”
“If anyone sees it, it digs into the ground with its tail and burrows away backwards.”
“If someone spots it, it escapes by digging into the ground with its tail. It can uses its wings to float for a bit.”
Quite the difference! In the original version, Dunsparce sounds like a horror movie villain, stalking its prey without sleep nor rest as they run desperately for safety. When it finds its victim, terrified and exhausted, it beats them senseless with its head! In the final…if you spot a Dunsparce, it borrow into the ground and runs away from you. Overall, a much less threatening foe.
(Fanart by Uluri)
The reason for this change seems obvious to me. When these Pokedex entries were written, Yoshida hadn’t finished drawing the sprites; his first concept was much more in line with the Tsuchinoko, which is a hunter according to legends. But when he finished the sprites, the team realized that the final Dunsparce was far too dopey and goofy to be taken seriously as a predator, and so they changed the entries to match. Interestingly though, Dunsparce’s movepool still better reflects this early lore: the very first move Dunsparce learns is Rage, and it also has Pursuit (to chase after its prey) and Spite (for its vicious nature). Its most powerful attack is even Take Down, which matches the description of how Dunsparce strikes its prey with its head. I think that’s so interesting! What was a strange movepool in the final game turns out to have been based on an entirely different concept. Its sprites were probably finished so late that the team either didn’t have time to brainstorm a new movepool, or they simply overlooked it in the rush to put the finishing touches elsewhere.
Because of its last-minute nature, Dunsparce has always been a strange oddity in the Pokemon world. People admire its dopeyness, fangames design bizarre dragon-type evolutions for it.
(Credits, in clockwise order from top left: From Pokemon Uranium fangame, Wooden Plank Studios, Runesparce by AgentKirin, @thelastshaymin)
Generally though, poor Dunsparce has been ignored, a forgotten weirdo from an generation with lots of forgotten weirdos. Except, finally, just last year, there was justice for Dunsparce! After twenty years, Dunsparce got an evolution that could finally make it useful. Did it turn into a majestic dragon? Did its wings grow into falcon’s wings? Did he finally become the predator that he had always been supposed to be?
Nope. He became a slightly longer Dunsparce. After being a mystery for twenty years, Dunsparce became the longest shaggy dog story in Pokemon history. And, honestly, I love it. It’s the perfect design for a perfect weirdo.