What Dreams May Come: Prototype Data for Pokémon Red and Green

For in that sleep of death what dreams may come

when we have shuffled off this mortal coil, must give us pause.

-Hamlet, Hamlet

So… you may have watched the Pokéthon broadcast, and read the documentation, and you may be expecting some explanations. We present to you the preliminary rundown and analysis of all the prototype assets for Generation I games that we published right here.

As a response to some of the commentaries and criticisms of how we went about the release, we’ve added some notes and made some changes to this article. Pokéthon did not go exactly as planned; we had no intentions to trick the community, we’re a really small site and we couldn’t care less about views or clickbaity shenanigans. Our main aim is, and always will be, sharing our findings with the community as soon as we prove their legitimacy. No one on the team has ever released something of this magnitude, so we’re learning from this mistake in communication.
As a result, we’d like to apologize in that regard, and we’ve removed the fan names we made for the MissingNos and any references to our ‘restoration’ romhack’s interpretations from this article, in order to prevent any further confusion. There have also been some edits made for factual correction and clarification, naturally.

As there is an incredible amount of information to go through, this is only a summary of the most important parts of the prototype data given to us. We have no other data from this, as this is all that we were given. Additionally, the formats in which we were given the files were non-negotiable (i.e. ROMs were off the table).

Early Title Screens

Do these look familiar to you? We’ve seen both of these logos before: on the map sheet for Capsule Monsters, and the Kaiju Index’s cover. Rhyhorn looks like its early design, with no drill pattern on its horn.

No longer lost to printer ink
If only we had it in higher resolution… oh, wait

The MissingNo. Family Robinson

In general knowledge, MissingNo. is a glitch Pokémon created from garbage data. Through hints like the internal index order and leftover data, like cries, speculation arose that these were Pokémon that were deleted and overwritten. A few of the MissingNo. identities were already known from sources like Satoshi Tajiri’s recently published biographical manga and an NHK broadcast that revealed some early Capumon designs. The information that we received served to further unmask the cut Pokémon. MissingNo. was never Ho-Oh, Lugia, Nidogod, Pikablu, Togepi, Yoshi, Charcolt, or Venustoise…….. it was these guys. And we’re so glad to have them back. For the most part, we only got back sprites, not front sprites, so we had to take some liberties as to what these Pokémon resembled or were based on, and recreated them for the Pokéthon presentation. For the purpose of clarity, there is no fanmade content in this article.

There’s only three unidentified MissingNo. left, not counting revisions that were overwritten or the Fossil/Ghost sprite slots

21 (Omega), who we all know and love, isn’t here because Mew overwrote its spot! However, its name is found in the cry list. Omega, seen in the Capumon pitch documents, is based pretty heavily on robotic kaiju such as Mechagodzilla.

Red and Green’s early sprite style often resembles super-deformed versions of kaiju characters.
Gyaoon’s front sprite from the Tajiri Manga

31 (Gyaoon) – Gyaoon (originally Gyace), the second Capumon sprite, has had a rocky history, and seems to have inspired Tyranitar later on. We also know from the Game Informer cry list that it once had Ivysaur’s original cry. Its backsprite may reflect changes that were made to its design by the time of the Tajiri Manga poll. As we already know from our analysis of the index list changes, Gyaoon was plucked from slot 1 of the list fairly early on and caused a small restructuring of the index order. It’s possible that another Pokémon was originally in slot 31, and then it was deleted to make room for Gyaoon’s newer design.. and then even that was deleted. Woo-hoo.

32 (Nidorino Lookalike) is probably an earlier spot for Nidorino. It has two horns on its back, like Nidorina, and has Nidorino’s general body shape and ears. It also looks like a bigger Nidoran M…. Like Shellos and Gastrodon, maybe Nidorino and Nidorina were once one, but…. How would that even work? Why did it change? Nidorino and Nidorina’s final spots are very late, right after Raticate; perhaps the final Nidorino overwrote a third Rattata-line evolution…?

Barunda’s front sprite from the Tajiri Manga.

50 (Barunda) – Barunda ended up being in a different slot than our hypothesized slot 81 (which is where Mikon is). All of the Pokémon seen in the Tajiri Manga were pretty close together index number-wise, it seems. We are aware of Barunda’s name because it was voted for in the Tajiri Manga’s ballot. Jigglypuff was probably the other potential design for Barunda, as there were (a) and (b) designs at the time of the manga poll, and Jigglypuff ultimately seemed to have been preferred over Barunda.

The name “Buu” from the Tajiri Manga ballot

52 (Buu?) – You can say a lot about this one right away. Its design is sure to bring up some… conversations. The missing ‘Ice Punch Yokai’, and possibly Jynx’s male counterpart, MissingNo. 52 is very unfortunate. We think it was inspired by “Woo”, a similar yeti-like kaiju with a dark face from Ultraman. Woo was actually a female of its species, which may explain the nature of Jynx, and perhaps also the reason for 52’s removal. As a bonus, if its name is really “Buu” (seen in the Tajiri Manga poll as a name on the ballot), it’s an obvious play on “Woo” – and it stands between “Elebuu” and “Buuber” in the index, bridging the names of the trio. If there’s anything that we’ve learned from this, it’s that the Game Freak staff adored Ultraman.

Deer’s name and front sprite from the Tajiri Manga

56 (Deer) – Deer’s here. Its backsprite doesn’t suggest that any changes happened to it between the Tajiri Manga and right before it was canned. It may have eventually been redesigned into Stantler for Generation II.

61’s front sprite from the Tajiri Manga

61 (???ko/???ro) – The elephant from the Tajiri Manga is also here! Now we know for sure that it had four tusks. We still don’t know the beginning of its name because of the way it was cut off in the manga, but if the kana there is actually コ ko, it could be a pun on a famous elephant in Japan named Hanako. Judging by how the next elephant Pokémon was Ground, maybe this one was, too.

Crocky in the Tajiri Manga, with hair

62 (Crocky) – Crocky’s backsprite has spikes on its head rather than hair – a design alteration that was visible in the Tajiri Manga, drawn on top of its original sprite.

122, the supposed evolutionary relative

63 and 122 – A squidlike pair. There is no evidence in the documentation as to their relation, but visually, 63 may have evolved into 122. 63 looks a little like Omanyte; 122, on the other hand, resembles a myriad of squid designs in other video games, like Blooper from Mario, or Squishy from Kirby. 63 could have been a standalone Pokémon with no evolution for a while before 122 was added. It could have been based on an ink pen. Flash was originally called “Squid Ink”, further supporting this theory. Looks like the victory goes to Octillery.

Front sprite from Tajiri Manga

67 (Cactus) – Cactus is back as usual and so is its cry. From our previous research, we still think it could have been a resident of the desert seen in the concept art, if that idea ever even got out of the concept stage.

Jagg’s front sprite

68 (Jagg) – Jagg is Jagg, nothing wrong with that. We first saw it in the Tajiri Manga, and its backsprite shows that it had a dorsal fin.

69 (Zubat Pre-evolution)… nice. The mystery of the identical Zubat cry is solved – either it never got far enough to have a distinct cry, or Zubat had a deeper cry and it took 69’s cry instead when it was deleted. It seems like GF had a hard time making this guy, looking like a Halloween decoration over here. It may have had … only a mouth? Hard to tell.

79 and 80 – These two are a little confusing – they seem to be birdlike at first glance, but then you also notice the fins. They seem to be based on flying fish or betta fish… or the Cheep-Cheeps from Mario. Movesets also exist for them, which we will be talking about soon.

81 (Mikon) – Oh, GF. You were gonna have Mikon from the start, which explains the Pokedex entries being like that even in the first generation, but you cheekily dropped the (Poke)ball – twice! Mikon doesn’t fit with our hypotheses or the patterns we had made, but we didn’t know about any of those other pre-evolutions existing in Gen 1, either. Its cry is strange and seems unfitting for the Vulpix line. Either way, we’re glad it’s here.

What ARE you?

86, 87, and the art of trying to decipher a backsprite that was virtually only a head: the newest bestselling biopic. The recognizable topknot on 86 gave away that it was something Japanese-based. We originally thought: ninja crickets (based around Substitute originally being called “Cast-off Cicada Shell”), with the orbs on the head being eyes. We also theorized the Kokana/Kasanagi Beedrill prototypes, but that didn’t seem to fit, either (and it’s unlikely that the Beedrill line got moved around when it was redesigned). Our favorite idea was… fighting or sumo-wrestling frogs. The orbs on it could be the cheeks of a frog. It’s also surrounded by Japanese culture-based Pokémon like Ninetales (kitsune) and Pikachu (daifuku). These two really need some clarification.

All in the Family

94 and 95 are next to each other and have their own cries, and 174 was added near the very end, with no cry at all. This lizard- or dragon-like family was a three-stage evolutionary line… And, to be honest, the third evolution looks like it belongs in Bubble Bobble or Kirby. They could be based on Chinese water dragons, which are lizards common to China that spend time in watery areas and are popular as pets. 94 and 95 are right by Sandshrew and Sandslash – the pangolins – so the China/Pacific area as a theme may have been possible.

115 – Nothing. Whatever was here was gone. If we followed index list patterns, it may have been an evolution of an earlier ‘mon from the 60-80 range.

121 – This one is also missing. It was likely an evolution, too.

127 (Psyduck Middle Evolution) – Psyduck had a middle evolution… which was probably cut for redundancy and/or balance reasons. It has evolution data proving its place in the “family tree”, as well as a moveset. Its cry is strange, and may have been some sort of error or typo, as switching around two of the numbers gives it a cry that makes more sense. Here is a theory that we came up with: maybe its Japanese name was “Zuduck”, if the Psyduck line was a mirror to the Zubat line: (“Kobat” -> Zubat -> Golbat / Koduck -> “Zuduck” -> Golduck). This is only a hypothesis, though.

134 in the GCCX Kaiju Zukan video. The image is too unclear

134 (Konya) – Meow! This Meowth pre-evolution was present at Spaceworld, but this seems to be an earlier version. Its backsprite looks more like it was standing up at this point, instead of in Spaceworld, where it was lying down; additionally, the tail is long and straightened instead of short and curled. It’s also noticeably bigger than Meowth’s back sprite, but as we saw in the Tajiri manga poll, Meowth was bigger earlier on, so this may explain the discrepancy. 134 was also spotted in the GCCX video, but we’re not sure how well it matches up with the blurry sprite, which is mostly unrecognizable right now.

This slot was filled… once

135 – Another empty slot. We only have external information for this one – the GCCX Kaiju Zukan footage. Considering some sprites like Venusaur and Cloyster were moved to later slots, 135 may actually have been Weezing, as its sprite does seem to fit the blur.

137– It shares a cry base with Gyaoon, so… it’s probably Baby Gyaoon!! As a pre-evolution, it could have been the original occupant for slot 31 before Gyaoon took it over… Its disgusted expression is so precious. It resembles the Substitute doll a bit, doesn’t it?


140 (Magneton copy?). It’s similar to 032 – its backsprite is like Magneton with some differences, like missing the magnets on its other orbs, and it just doesn’t seem to make sense as a standalone Pokémon. Index-wise, Magneton was very early, this one was somewhere in the middle, and Magnemite was fairly late. They seemed to have been trying out something here, perhaps an early pre-evolution of Magneton or a middle evolution between Magnemite and Magneton, and it was probably cut due to redundancy or maybe balance reasons. We wonder if perhaps it had more to its body than can be seen in the back sprite.

There’s a Justin Bieber song about this

146 (Marowak evolution) – This is the one we were all very shocked by. The common fan theory that Cubone is a baby Kangaskhan came true, in a twisted sort of way; now it seems like they were always meant to be linked together – the Marowak evolution is separate from Kangaskhan, but they share the same theme. In the end, it would have become a loving parent to its baby Cubone, maybe even having adopted it (after all, it still has a skull on its head.) It could be holding the baby, or it may have a pouch like Kangaskhan. We can’t tell for sure what’s going on in those pixels.

156 (Gyopin) – Gyopin was present in Spaceworld 1997, but this seems to be its earliest incarnation. You know, Gyopin makes a bit more sense now that we know it was the original first step in this evolution tree. It got as far as having a moveset and evolution data, too!

159 (Kotora), 160 (Raitora-1), and 161 (Raitora-2) – Like we said. It’s Kotora, as we saw in the GCCX video and in Spaceworld. But….. we weren’t expecting a three-stage line. Raitora seemed to have been originally two different Pokémon, and was combined into a two-stage line for Generation 2. This somewhat mirrors Pikachu’s history, ironically, where it had three stages initially, but it had turned into two by Gen 1’s release.

162 (Puchicorn) – Like Mikon, this one also corroborates the Pokédex entries. It seems like Puchicorn, Kotora, and Gyopin’s evolutionary families may have formed a trio!

172 (Blastoise Pre-evo?) was a real toughie to figure out. We thought it was some sort of cow or pig initially, as the circle on its nose resembled a nose ring. This led us to think it was a pre-evolution of Tauros. Then we realized that the nose ring could be a bubble. And the things on its back would then be bubble CANNONS. Blastoise had seemingly been given a pre-evolution when 181 was still the final evolution for Wartortle.

175 (Gorochu). We thought we saw a walrus / With icy hide and horn / We looked again (at the prototype evolution table), and found it was / Gorochu, long forlorn! Oh, for all the joys of life, we have Gorochu! And, uh, it only has one horn, not two! We also don’t know what the spikes at the back are – they could be a tail, or perhaps a cloud, like some people have theorized since Gorochu’s description seemed to be Raijin-esque.

181 (Wartortle Evolution) is, as shown in the evolution list, Wartortle’s true evolution, right where we hypothesized it to be. Yay! It had only one move in the unfinished moveset list, but Squirtle and Wartortle had full movesets at the time, so we will be analyzing that soon. 181 was probably intended to be the last Pokémon designed, as 182-184 take up slots for Ghost and the two fossil skeletons, which were not intended to be catchable Pokémon – they just needed an extra slot. The Oddish and Bellsprout lines were added afterwards, probably to balance out the game and increase the number of Grass-types.

Omega is the only MissingNo. that we knew only from external information (the NHK broadcast), and its spot was eaten up by Mew. All the other designs, like the Kokana line, old Dragonair and Poliwrath, etc. probably never took up a different slot since they were edited into the existing Pokémon. As for the blank spaces, we still really don’t know – the blurry blobs from GCCX aren’t giving us any more hints right now as to the identities of the remaining missing Pokémon.


Seemingly called “dealers” internally. Some familiar art in here! We’ll go in order of their placement in the comparison table, with the exception of the first one, which is the standalone “Missing Intro Graphic” file:

Yuuichi – He’s you! This could have been the player sprite in earlier stages of development, and also matches a pose seen in the concept art! There are two of these, one that looks more experienced, with a whip. It could have also been that as you played, your sprite would “evolve” like the Rival’s sprite does in the final. The intro graphic sprite resembles the ‘shrinking’ animation frames you see in the intro when Red becomes the overworld sprite.

Red Backsprite – Looked rather cartoony early on, but it matches with the early protagonist sprite (except for the sleeves and black hat brim) and the concept art with the whip-wielding person in a similar pose.

“Student” – This one looks very rough and early design-wise. Her pose may have been borrowed and reworked into the Cooltrainer F. Like Rhydon, it seems like she may have occupied a 0-slot once.

Shinjuku Jack – Probably one of the most bizarre cut Trainers. Shinjuku is a popular commerce district within Tokyo, which doesn’t really connect with this guy as a nickname. It seems obvious that Jack is a robot Trainer! His name was replaced by Juggler’s name, but his trainer sprite was simply blanked out. Jack’s name also survives into Gold and Silver in the unused trainer list, by the way. Jack and Omega would have made a pretty good team, wouldn’t they?

Firefighter – This was a cool design, but was replaced with the Psychic. They could have had Water and Ground types. Swimmers already existed, so perhaps it was cut to diversify the ‘types’ that Trainers represented.

Silph Chief – We tried to warn you! Here he is taking his original spot next to the Scientist. While we still don’t know his purpose, he may have been the President or even someone trying to take over the company, like Faba in Sun and Moon. None of his teams survived, but perhaps he had Bug-types, because Giovanni’s Silph teams in both earlier lists feature Scyther and Pinsir. Who knows?


Yujirou – This “C” for “Capumon” shirt kid went through a lot. The first design for him resembles a Youngster. In his second incarnation, he became the proto-“first” Gym Leader that we saw in Sugimori’s concept art on his Twitter. He may have been the original Gym Leader for Viridian City. His design doesn’t seem to evoke a type – perhaps he was a Normal-type trainer.

Ichitarou – Brock used to look like Buu saga Krillin, and his name was different. “Ichitarou” sounds like a pun on the number one in Japanese, “ichi”, but considering Yujirou’s status as the first Gym Leader, who knows what it meant for poor Ichitarou.

Blaine – The army man seen in Sugimori’s twitter post! The gym he would be in does have a lot of holes in the floor, for someone who allegedly had bombs in his gym, according to Sugimori.

Junior/Umezou – We think this guy is Blue/Gary’s little brother. Here’s why: In Japan, there is a three-tier system used occasionally for ranking. Ume (plum) is the lowest of the three. It’s a pretty obvious pun name. Blue had an earlier name of “Take” (bamboo), the middle rank. Does that make Oak “Matsu” at some point? It’s also possible he was YOUR little brother, like Ken was in Spaceworld. Who knows!?

Overworld NPCs

You may have noticed that many of the NPCs have similar designs but are rougher, while others were cut entirely or reworked. Seeing them makes one really appreciate how much work Game Freak put into making the overworld sprites look like believable people. Here are a handful of important personalities that we can go over in more detail at a later point:

  • Early Protagonist – You’re already familiar with this guy from the NHK broadcast and the New Game Design cover sprites.
  • Danpei – The original Gym coach. He’s based on the coach named Danpei from Ashita no Joe, the famous boxing manga. That includes the eyepatch. Gee, I wonder why they changed him? Also, Danpei’s name survives internally, even in X and Y! Thanks to Twitter user pokesupe29 for pointing this out.
  • Dad/Otousan – It’s your dad… probably! His sprite’s removal shifted a copy of Oak’s research assistants/the Scientist into his slot, so……… he certainly is a mystery, isn’t he?
  • What we thought was a jet-ski turned out to be a Virtual Boy! Can only imagine why this extremely relevant gamer was removed.
  • Disappointingly, no sign of the female protagonist seen in the concept art – she may have been removed by then.
The NPC sprite in question and an image of Danpei from Ashita no Joe’s anime

Trainer Parties

-Agatha had the three Eeveelutions and Tauros – fitting for Oak’s longtime rival.
-Most trainers underwent changes. For example, Youngsters had Bulbasaurs and Geodudes everywhere.
-Giovanni was a Flying-type trainer. Look at the Earth Badge – no kidding.
-Both lists were created after the Pokémon list was finalized – none of the Missingno. are there, and the Bellsprout/Oddish lines at the end of the index are present.

The Other Cry List

Probably one of the most important pieces of evidence given to us was the cry list, but it was actually not the one which we had partially uncovered in the Game Informer interview video. The names are out of order from the in-game base cry list, and feature some strange differences; this could have actually been made in the order that the cries were designed or put into the game, as opposed to the cry list in the video and in the final game, which was more based around the Pokémon index order.

Many of the cries match the Pokémon that they were named for, but some… don’t.
The order and names in this cry list seem to imply a very early stage of game development, as it seems like the game was originally based entirely around kaiju-esque creatures – the first eight cries here are all kaiju Pokémon. We can see some evidence of this in Capumon, where most of the creatures – even test creatures like Gorillante – resemble kaiju. Additionally, some of the known Capumon names have different romanized spellings that don’t entirely match their Japanese kana, which may or may not indicate that the names were different at one point.

On a similar note, Game Freak intended the cry system to be like how each kaiju in series like Godzilla and Ultraman have a distinctive sound that they make. Recall that Capsule Monsters had a significant base in the “capsule kaiju” of Ultraman, as well. After the first five cries (excepting that of Nidoking), they probably started creating cries based on how the Capumon looked like they may sound (and thus named them after the actual Capumon). In fact, if you go on the Ultraman Wiki and look at the cries with the same names as Ultraman kaiju, the base cries may sound similar. Try it out!

  • 0. Earthtron – The unused base cry which survives in the final game, it actually uses the name of a kaiju from Ultraman, アーストロン/Arstron, which can also be romanized as “Earthtron”. Both the ingame cry and Arstron’s cry start with a high note and then have a short lower-pitched note at the end.
  • 1. Mycor – This is a romanization of “Maiko マイコー”, the early name for Nidoking. Nidoking iself is based on the kaiju Baragon, but no kaiju from any series seem to have the name Mycor – it may have been original.
  • 2. Kingthaurus – This is another Ultraman kaiju, Kingsaurus III. The cry is used by Pokémon like Aerodactyl. No Pokémon has the ‘base’ version of this cry, but both the base cry and Kingsaurus’ roar are a high-pitched screech.
  • 3. Sutegon – Another Ultraman kaiju, Stegon. The ingame cry (used for Graveler and others) is a deep, rolling roar, similar to how Stegon sounded.
  • 4. Eleking – No, it’s not Electabuzz’s secret scrapped evolution. Eleking is an Ultraman kaiju. Its roar from the Redman series, which is actually the roar of Bemular, seems the closest to the cry base’s melody (used for Pokémon like Rapidash.)
  • 5. Gagarth – Rhydon’s old name (also romanized as Gagars). This is where the cry list starts to follow the index order… sort of.
  • 6. Garura – Kangaskhan
  • 7. Gyace – Gyaoon / Gyaasu
  • 8. Omom – Spearow’s old name, Omuomu.
  • 9. Hariusagi – This has to be Nidoran, judging by how ハリウサギ literally means ‘needle rabbit’… and it points to Nidoran’s cry base (1). Nidoran M’s cry is the first base cry in the final game’s list, and was also the first in the Game Informer video’s list, but it’s number 9 here.
  • 10. Pip – Romanization of Pippi, Clefairy. The early romanizations for the names are really interesting.
  • 11. Biriridama – Voltorb
  • 12. Gengar
  • 13. Bettobeter – Grimer
  • 14. Tamatama – Exeggcute
  • 15. Beloringa – Lickitung
  • 16. Yaddon – Slowbro, who (as evidenced before) seemed to have switched names with Slowpoke. The double D would seem to imply that it was called ヤッドン instead of ヤドン, but we haven’t seen that anywhere in kana form yet.
  • 17. Nasshii – Exeggutor
  • 18. Hariusagi 2 – Nidoran F
  • 19. Mycor 2 – Nidoqueen
  • 20. Fushigiso – Ivysaur. Ivysaur’s final Japanese name is “Fushigisou”, with an extended “o” sound at the end, but whether this was just a shorthand name or its original name is unknown.
  • 21. Cyhorne – Rhyhorn
  • 22. Karakara – Cubone
  • 23. Strike – Scyther
  • 24. Rapruth – Lapras
  • 25. Monjara – Tangela
  • 26. Omega
  • 27. Caiross – Pinsir
  • 28. Mimi – Staryu
  • 29. Wing – Arcanine’s early name, as seen in the NHK broadcast photo.
  • 30. Gyaradoss – Gyarados. Not sure why it has two S at the end, because its name has never been seen as ギャラドッス, at least not by the time of the Capumon pitch document.
  • 31. Gouth – Romanization of ゴース, Gastly
  • 32. Ambler – Tentacool’s Capumon-era name. You might have noticed our social media manager slip up and spell it exactly this way a few months ago instead of something like Umblar. Oops.
  • 33. Karabatheo – Caravaggio, Blastoise’s early name.
  • 34. Shellder
  • 35. Yaddohogera – It’s Slowpoke, according to the index order. “Yaddo” comes from “Yaddon”, using the double D again. “Hogera” is a nonsense word used as a placeholder in Japanese programming, like “foo” and “bar” in English. As a result, this name can basically be translated to “Slow-whatever.”
  • 36. This cry is called 34, but refers to cry base 33. Onix probably belongs here as it is index 34 and the cry sees plausible for it, even if this was eventually used for Krabby instead. The use of numbers may indicate that this is where they stopped naming cries after individual Pokémon and started making placeholders again.
  • 37. Called 36, but refers to cry base 34. There was something in between there, wasn’t there. Hmm. Maybe in this pseudo-‘early index list’, there was a “35” base cry that got deleted. Pidgey would belong here if 36 was referenced index-wise, which seems somewhat fitting, since it’s a birdlike ‘honk’ and is used for Psyduck.
  • 38. Dog – This one could be Growlithe, which is nearby in the index list. It has the same naming convention as Arcanine’s early name, Wing. The cry has a name again; Growlithe comes before Pidgey and Fearow in the index, so perhaps something was moved around.


We were given four ‘versions’ of the move list – two move name tables, a move effect table, and an evolution/moveset list. All of these have differences from the final version and hint at different setups of the move list.
Things of note from “Move Names V1” and “V2”:

  • Punch – This was a generic Fighting-type or Normal-type move that got cut and replaced by other, more exciting punching moves.
  • Upper Cut – Concept would later be reused as Sky Uppercut.
  • BaiBai Punch – No idea what this means! BaiBai sounds like “bye-bye”, so perhaps it referred to a knockout?
  • Shibire Hasami – “Stun Cut” – Sounds like Razor Wind. Possibly a reference to Final Fantasy’s “Shibire Basami” (Numblade).
  • Shippo – Probably Tail Whip. There are two copies of this, as well as for Itohaki (String Shot).
  • Hinotama – Fireball.
  • Mega Fire – This seems to be a progression of Fire-type moves by power. There is a second copy farther down in the list. Possibly a reference to Final Fantasy’s Megaflare.
  • Konayuki – Name is identical to Gen 2’s Powder Snow.
  • Hydro Jet – Early Hydro Pump name.
  • Elekitel – This one has a bit of interesting history. The Elekiter is a Dutch static electricity device that became popular in 1700s Japan. This could have been the original name for Thunder Wave.
  • 50Man Volt – An upgrade from Thunderbolt, which is “10-man volt” (aka 100,000 volts) in Japanese.
  • Ieki – A different move than “Youkaieki” below it, which is Acid. Ieki was used in later games as the name for Gastro Acid (which is exactly what Ieki means, anyway).
  • Honoo – Seemingly separate from Fireball, this is just “Fire”. A second copy of Mega Fire exists with it, suggesting that this set of Fire-type moves got moved around a bit while being edited. Exists in the TCG on Growlithe as “Flare” in English.
  • Star Freeze – This took up Rock Slide’s slot when Rock Throw was called Rock Slide instead. Also seen in the TCG on a Starmie card. The sound effect for Rock Slide seems out of place, so perhaps it is a leftover from this time.

Things of note from “Evolutions and Attacks”:

  • This list is set up differently from the others, because the moves aren’t labeled with their names, and instead use numbers. When comparing these to the final games’ move index, they don’t match up, indicating that the move list was going through a lot during the time that this list had been made. Gorochu, Wartortle’s evolution, Gyopin, Puchicorn, Zubat’s pre-evolution, and Psyduck’s middle evolution populate the list, as well, meaning that it was made when they were still being considered for the final game. The table isn’t complete, either. When we attempted to compare these movesets with the two move lists and the final move list, nothing fit enough to draw solid conclusions about the identity of each move.
  • The number of moves that existed at this point was very high – the list mentions move numbers up to 237, hitting close to that 256 limit. For comparison, the final game only has a total of 165 moves. There could be more, but 237 is the highest number listed in the file. The final attack list is the abridged version of this huge list, as the structure of the list seems to have stayed mostly the same, while redundant or less notable moves were cut out. Despite the challenge that this presented, we were able to create a reference point: 237 is definitely Slash. It’s the last move in the final list, and is learned by Pokémon such as Parasect. Other later moves correspond well to the final moves of the final list.
  • Only about 90 Pokémon are present in this list, predominantly early-game Pokémon like the Nidorans and starters. Only some of them have full or even partial movesets. This may imply that the list was filled according to some in-game Pokémon order, like an early Pokédex. When this list is sorted by attack numbers and who learns them, we can start to deduce other comparisons, like what moves Nidoran and Rattata would have in common.
  • Evolution seemed to be more level-based at this point. Clefairy is the only Pokémon to evolve by an item, the Moon Stone – every other Pokémon on this list evolves via level-up, including the Pikachu line.
  • Pokémon whose learnsets are outlined in this file often have more moves than in the final. Sometimes the moves will repeat themselves, too. They don’t learn any new moves after the level at which they evolve; this implies that evolutions were mandatory and the player couldn’t cancel them, as this would render the Pokémon unable to actually grow stronger outside of stats and TMs.

We’ll have further move analysis that also covers the other move lists later, because the amount of stuff already covered in this article is immense by itself!


There’s a LOT to process here. We were amazed at the way that the world of Kanto developed over time. This is also going to take up its own subpage, but we’ll provide some highlights.

WOW!! Some of these tiles can be seen on the New Game Design book, but there’s a lot more. The art style really resembles the concept art. This might need its own article.
We can see how the world developed from start to finish. Route numbers were literally written into the routes they belonged on in grass as a template. Cities didn’t have a lot of defining features. It has a very strong traditional RPG feeling to it.

Later, more updated maps with the final’s tilesets had placeholders for certain buildings like Silph Co. (an entire island of pegs) and the Celadon Department Store (a stack of wide one-story houses). At this point you can tell that they had plans for each city and probably the story as well.
As a side note, something to look for more coverage of in the future: Silph Co., according to our donor, was the original host of the Pokémon League, which was also quite a bit larger than in the final game. This concept was moved to Indigo Plateau and then cut very late into development. They had apparently started to work on adding a replacement area to Silph’s old League which had every Gym Leader and trainer type and would have culminated with a dungeon that hosted Mewtwo. It can be assumed that this second attempt surrounded the Trade Center and Colosseum in the index list, as one of those maps is the default warp for Silph’s elevator.


There is still so much to dig through, document, and analyze. More articles will come soon about this, but now it’s out there for everyone to see. Thanks for your patience!


  1. I’d like to question why we were told in the stream that it isn’t a romhack, yet now there’s written proof stating otherwise. It’s an edited rom with custom sprites following as closely as possible to what was theorized on their appearances, and that turned out rather disappointing to find this out. An infodump without a misleading rom hack would’ve been far more interesting than watching a romhack, being told it isn’t a romhack, and finding out shortly after that it is a romhack. It kind of feels like those who donated just to find out the sprites were tricked.

  2. Hi,
    I’m a big fan of yours and have loved all of your beta coverage thus far.
    However, I feel that this particular incident was handled really poorly. The way that it was announced and the ROM implied to be “real” really has let a lot of people down, and left them feeling that their time was wasted.
    If anything like this comes up in the future, I believe it would go down a lot better to have someone who knows their beta content explain and go through the ACTUAL leaked information, as opposed to playing out a vague ROM that is largely false and unofficial.
    I feel that this is going to reflect really poorly on yourselves and other groups that were involved. Fans want you to be forthright, not righteous!
    I appreciate if this has been read, and really urge you to consider what I have said. You will see many, many people echoing my sentiment.

    1. Sorry, of course, it wasn’t our intention to trick you, we just though the playthrough was going to be a fun and quick addition/introduction to the topic, we didn’t subtract nor withheld a thing from the assets and took care in giving you the usual thorough analysis. Thanks for your comments 😀 we’ll definitely take notes for the future.

      1. That still doesn’t explain why you withheld the fact that it was a rom until after the stream ended. If you respected us as audience, you’d would have informed us beforehand, so we could make the decision to watch or not. Instead you played coy probably so you’d get more views.

        1. Aw, sorry we’re not that cunning, the thing was handled in a really spontaneous way, thinking it was fun and might help a good cause, we don’t really mind about views, the actual assets would surely make more buzz. Sorry again it wasn’t our intention.

          1. I’m sorry but why should we be willing to believe anything in this article when, despite your “aw shucks” attitude, a romhack was streamed under the pretense of being a genuine beta, while asking for donations. That’s actual fraud, my dude.

            I want to believe this is real, I love me some betamons, but releasing and streaming a romhack without disclosure, at face value, makes everything about this look fake. How do I know you aren’t going to show me other interpretations as opposed to raw data that won’t be disclosed as such until a later time?

            If you genuinely believed that you were being “fun” and “spontaneous” rather than “lying to their faces” and “misrepresenting your assets for publicity and donations”, then you need to make a real public apology. Otherwise, HC’s reputation is gone.

  3. Any chance to get the 3 remaining sprites? Is possible for them to be based on Growlithe, Tangela, or Grimer preevolutions? After all, we got Ponyta, Vulpix, Goldeen, and Meowth preevolutions there.

    1. they’re probably related to Farfetch’d, Crocky or Cactus but anything is possible! Also no, there’s nothing more at the moment :/

  4. This is fascinating, but as someone who didn’t see the Pokéthon broadcast because it was in the middle of a weekday night my time, it doesn’t wholly hang together as a concrete piece – you talk a bunch about what you did with the designs of the Pokémon in the ROM hack, but don’t actually show those designs, for instance. It’s good and reasonable to not want to mix your fanmade sprites in with the actual prototype info to avoid spreading misinformation, but then why are you talking about your fanmade sprites in this post to begin with? Why not make two separate posts, one that just explains the prototype info and one that shows and discusses the fanmade ROM hack properly (with appropriately huge disclaimers attached and a link to the post with the actual info)?

    Similarly, you make a vague reference at the top to “you may have watched the Pokéthon broadcast, and read the documentation, and you may be expecting some explanations”, and then just dive right in. Unfortunately, someone who didn’t watch the Pokéthon broadcast or read “the documentation” (I guess you’re referring to the Twitter thread?) would simply have no idea what any of this post is on about. In other spots you reference names, etc. that one would have to have read some of the previous posts on the blog to understand, without linking back to a post with the relevant info (I can’t for the life of me remember what Gyopin is). And when you start talking about the cry list, I’m similarly confused – what is this cry list? Where do the names on it come from? As it is, I don’t feel like I can share this post and all the cool information in it, unless I wrote up my own multi-paragraph explanation to accompany the link (and I could only do that with an incomplete understanding of what I’m talking about).

    I really wish you’d make a bigger effort to make these posts accessible to people who haven’t been following everything you do with a microscope! I want to spread the gospel and promote all this cool beta stuff, but there’s no proper accessible stanadalone writeup that I can link to, as far as I can see.

    1. Agreed, we’re working on updating the whole site, it might take a while to rewire all the links and make a more accessible documentation. As far as the present article goes, I feel the core info is all laid out but we’ll work on the details. Thanks for the advice, we’re still learning.

    2. Okay, but how is the ‘crylist’ not very obviously self explanatory? It’s the list of cries in the actual game.

  5. I feel it was a bit disingenuous during the stream to say this wasn’t a Rom hack. Just dropping the assets would have been better. Plus, if you take into consideration the front sprites were fan made, there’s not even enough real content that this hack felt worth while. Most of the “meat and bones” was actually fan content. I would have been far more interested in the stream if this was explained while playing. I realized early on it was a Rom hack and didn’t appreciate people bring it up getting muted.

    1. Honestly, we thought the “meat and bones” were actually more than the fan content, might’ve been a miscalculation of course. Sorry again, we’ll learn from this. Again we just thought it was a fun addition/teaser for the reveal, no malice at all. Perhaps it dragged a bit longer than expected and I understand this might get to your nerves.

      1. Lol no-one cares about fan content when it comes to this sort of thing. If we had a thing similar to the Spaceworld Gold we had then it’ll be fine. Give us the information and the fans can make art

  6. I love when beta stuff comes out! Through reading some comments, I’m assuming a livestream showing a romhack using the beta stuff was used as an introduction for all this information?

    It seems the livestream is impossible to watch now, and somehow some people are playing the hack itself. Is there any way to do that? If not, is there any place to hear the cries? I’m really interesting in that, but the file just has a txt document listing the cries. Thank you for compiling all this information.

  7. This is undeniably very cool, but wow. You guys really dropped the ball presenting this. It was extremely disingenuous of you guys to say multiple times during the stream “This isn’t a romhack” only for it to literally end up being a romhack. I honestly feel bad for the people that worked so hard to make the romhack because this whole stunt was executed so poorly. Showing off front sprites and pretending that they’re real during the stream only for it to be announced that they’re actually /not/ real and you were only given backsprites is a huge disappointment. If you were upfront about this, the romhack could have been a fun and cool way to show off the content you guys were given as well as the front sprite interpretations that you made. But instead, you hyped us all up for what has turned out to be 50% fanmade content. The cagey act was not funny nor was it fun to watch. It got old as soon as the runner started doing it. Don’t act like you aren’t cunning enough to pass this all off as real for your stream in order to get more views and donations because that really doesn’t require much cunning and that’s clearly what you did. You guys really shot yourself in the foot doing this. You’ve certainly left a bad taste in my mouth and the same can be said about many others. Obviously you don’t owe me and everyone else anything and you didn’t have to show any of this stuff off at all, so don’t get me wrong, I really appreciate that you did. This is an incredible find. But from one person passionate about gaming history to another, you messed up pretty badly. I hope you read this. Thank you for your time.

  8. A note about Jack: his design seems to be based on Jacky Bryant from Virtua Fighter (particularly his very polygonal model from the first game). Ken Sugamori is a known SEGA fan, so it’s not too surprising that he might try to sneak in a reference.

  9. (Tried to comment before but didn’t go through. Will try to break it up into parts – hopefully this works):

    I’ve been following this blog for a while now, and have always appreciated the information that you’ve shared. It’s fantastic that these early designs and understanding of Gen 1’s development are now preserved on the internet; especially considering the far-reaching influence that Pokémon has had over the past few decades. However, I feel like that the way that this new info was revealed was executed poorly and (given your reputation) irresponsibly.

    The decision to reveal this new information by incorporating it into a hack that includes fanmade content just seems senseless to me. I can understand the excitement that must have come with receiving this information and the desire to reveal it to some spectacle, but I think the long-lasting repercussion of this is that legitimate discussion of pre-release content is now going to be muddied with what is essentially fanfiction. Just to name a Pokémon-related example that we’ve already seen before: drawings of Chikorita and Qwilfish, made by a Japanese fan who had played the original Gold Spaceworld demo, were widely circulated and became accepted as being depictions of beta designs for years (these pictures were up on Bulbapedia for ages). When the demo was finally leaked, we obviously learned that this wasn’t the case and the ‘beta designs’ were just the result of faulty memory, but the damage was done.

    Some people were aware that those drawings were fanmade reproductions, and that they were probably inaccurate to some extent; but, like anything on the internet, people just circulate this kind of stuff without the original context – I’ve seen it claimed often enough that those images were early design sketches from Game Freak themselves. My fear is that the images from your hack are going to make the rounds on social media, and your fan designs are going to be confused with the legitimate pre-release content. As these get shared from site-to-site, it’s very unlikely that the original context with which they were released will be circulated alongside them – people aren’t going to know about the stream, that half of the content in the hack (and the hack in-and-of itself) was ‘fake’, and that only certain overworld sprites, move data and back sprites are ‘real’. It’s all going to be jumbled and confused.

    1. (cont.)

      I don’t see how you can’t have predicted this, and while I hope that you wouldn’t intentionally reveal content in a misleading way to leave some sort of ‘mark’ on the fanbase, I can see why it’s been suggested. If you wanted to make a hack based on this unused content, I think that’s a great idea (I personally love fan-attempts to recreate beta versions of games) – but that’s something that should be done afterwards. The actual beta content should have been revealed separately, so that there was no confusion between the two. If you were just some tactless click-bait channel then this wouldn’t be an issue, but Helix Chamber has earned a name for itself among the fanbase as being a legitimate source for pre-release Pokémon information; given that reputation, it seems extremely irresponsible to reveal new information in a way like this. I know you’re stating that this was all done to make it fun and that you were upfront about it being a hack, but the feedback has demonstrated that this wasn’t the case – people weren’t sure what they were seeing, people were told that it wasn’t a hack, people feel they were tricked into donating for ‘fake’ content, and things were just generally communicated poorly all-around. I’ve already seen designs from your stream cropping up in different forums with the claim that they’re actual beta sprites – it’s only a matter before some YouTuber picks it up and reacts to these “early designs” and they spread like wildfire.

      Even this article is a mix of actual beta content and a description of how you interpreted said content in your hack, which just makes it a confusing read. For example, my understanding is that all we know about 94, 95 and 174 is the design of their back sprites and move data. You describe them as “lizard- or dragon-like”, which is a fair-enough assessment of the images that we have, but then go on to say that they “ended up being based on Chinese water dragons” – are you referring specifically to how you they were depicted in your hack, or is this the case for the original Pokémon designs? Again, if you want to do an article on your hack, then that’s fine – but the information specific to it should be clearly separated from the actual beta content. You’ve also given names to some of these early designs, some of which are based on your original interpretations of their back sprites – this is all essentially fanfiction, and I don’t think it should be mixed in what is presenting itself as an analysis of actual pre-release content.

      Honestly, I found the way that you chose to reveal this disappointing. Pokémon is the largest media franchise in the world, and the Spaceworld demo already drummed up tons of interest in beta Pokémon content. There was no need to embellish something like this, since it was always going to get people’s attention. I don’t think it was your intention to be misleading, or that you had some sort of nefarious machinations behind it all; I just hope that the excitement of all this just led to a momentary lapse in judgement, and that you’ll be clear about new information in the future as you’ve been in the past. Still looking forward to any further updates and articles by you all, and hope that everything is back-on-track.

      1. Thanks for the understanding. Of course we didn’t mean to trick the community, we’re a really small site we couldn’t care less about views and clickbaty shenanigans, it’s clear that our main aim is and always will be sharing our findings with the community as soon as we prove their legitimacy. we’re learning from this mistake in communication. Sorry again.

  10. Echoing the sentiment of others here, this just wasn’t the right way to go about it.

    It’s not worth pointing fingers and I absolutely understand the reasoning for the romhack as an exciting reveal but ‘artists interpretation’ just within or after the article would have served a similar purpose and probably been more welcome. Given the fact people have doubted the validity of the Spaceworld content, this seemed very short-sighted to reveal with.

    Not discrediting the artists or people who worked on it. It was beautifully done and absolutely necessary in helping piece this all together.

    The findings are fantastic and for someone who grew up with this franchise from day one, peeling back the mysteries is fascinating and I hope you’re able to share more discoveries with us in future.

    Massive thanks to the donor, and thanks for your work!

  11. (Apologies if this ends up getting posted twice – comments section was messing up for me.)

    What a find! It’s not cool that the guy with the beta refused to release everything (What’re they gonna do, wait decades to reveal it? Just sell it now while the iron is hot if you want to make a quick buck so bad, Guy Who Has The Beta), but at least they gave us something. A very substantial something.

    Unfortunately I do have to echo the people who are saying the release was handled poorly. The official stuff and the fanmade stuff are now forever going to be entwined because they were released together instead of the real stuff coming out first and then fanwork filling the gaps. Making a hack was a great idea – an amazing idea, even! Every time beta Pokemon show up, people talk about wanting to catch them in new Pokemon games, fangames, or hacks. So making the hack was a no-brainer. But it shouldn’t have been the introduction to the information, it should have been the follow-up, like what happened with “Pokemon 2” last year. And though I did not see the stream (I was asleep), apparently the streamer was pretending the hack wasn’t a hack? That is not at all the sort of behavior I expect from this site and its’ associates. You guys have built up a reputation for dealing strictly with facts and clearly marking any theories or educated guesses about anything. Enough falsehoods surround Pokemon’s history as it is, especially every time there’s a new gen announced and hordes of “leaks” flood the Internet from fakers. And it’s been a problem with Pokemon since day one – who else remembers the schoolyard rumors about all sorts of nonsense that wasn’t in the games? The Mist Stone, Bill’s secret garden, Pokegods, “it’s true my uncle works at Nintendo”, etc?

    But all that aside, your work is fascinating and extremely appreciated, regardless of any snafus related to fan content. Never forget that.

  12. Was it really that hard to say you had a recreation rom of beta Pokémon? Seriously? Don’t tell me you thought it was fun. It was not.

    First we got promised SpaceWorld in a certain hour. Ok, the guys were delayed, but did that justify that jyggy dude playing with everyone and being a total ass? He didn’t even try to go fast, in fact he purposefully was slow at times and it was noticeable!

    Ok, then, after some of us had waited two extra hours on A SUNDAY TO MONDAY NIGHT, it was surely time for the Spaceworld exciting stuff that was promised just but two days beforehand? Oh no, silly me, nah, you get this Green rom, if you were just for Spaceworld and didn’t care about it, hah get rekt you wasted two hours its not like a tweet saying Spaceworld didn’t make it on time and you had another exciting stuff could be written. In fact, you tweeted ABOUT SPACEWORLD right before the stream. Do you think we’re stupid? Is this being excited and funny, too?

    Then, when it looked kinda too ‘complete’ for a beta at the beggining (apart from that Red sprite, Pallet Town was actually the same) you just wanted us to waste 3 extra hours to say it was a recreation. Was that fun too? Fun for you I suppose, fun because you tricked all of us.

    How am I even supposed to believe this when you lied throughout the span of the damn night, not disclosed information, and then you come and brush it off with “sry guys maybe next time we’ll try”.

    1. Nobody is brushing it off that abruptly, first of all, we never said “sry guys maybe next time we’ll try”, we clearly said this was a mistake and won’t happen again FOR SURE. I know this is not a consolation for you, but right now we’ve nothing to offer apart from our apologies.

  13. Mixing in your interpretation/fanmade stuff in what you claim to be factual data makes it really confusing. Like the C shirt kid being the first gym leader and used a Lickitung. Did you just give him a Lickitung because you thought it was a good fit or was there actual data that showed that’s what he would use? Was he the first gym leader in the data you got or do you only have concept art to base that on?

    1. The Lickitung thing was ambiguous, the article is currently undergoing a revision because of the very issue you pointed out, sorry for that u.u
      Sugimori himself stated he was the first gym leader on his concept art, anyways he seemingly occupies the slot before Brock (now Bruno’s slot) as predicted by the trainer ID list analysis.

  14. You’ve obviously strained your own credibility here, but the stream and fake sprite nonsense aside, how can one reconcile the fact this content seems to span the complete gamut of various builds you’ve talked about in entries on this site, in a manner that doesn’t seem to make much sense?

    We know GF completely lost their hard drive data of the game as they’d been working on and had to go to great lengths to get the hard drive operational again – this apparent build seems to circumvent the issue entirely.

    Why is a last minute add-on, Mew, apparently in the game data? (n.b. I may be mistaken here but 4chan /vp/ has been crying foul on this point and I’m taking their word for it.)

    Let me put this another way: how am I, an “end user” of this data reveal, supposed to take the mysterious collector as credible? If I give you the benefit of the doubt, it looks like someone who has been following the Spaceworld leak and disassembly decided to cobble together a Red/Green romhack based on things this site has speculated on, that GF has incidentally hinted about, and appeared on Game Center/Tajiri manga, and leaked the virtual equivalent of catnip to you.

    The fact there are no front sprites is a dead giveaway to me. Some character and logo design can be derived from the aesthetics of earlier GF games, like Pulseman, but going full on with entire prototype Pokemon design in the style of early designs would’ve been a bit extra, so they conveniently decided to not provide it to you.

    I think you should be releasing this data, but putting very strong qualifiers on it that, unlike Spaceworld, there is nothing you can currently point to that makes it credible.

    1. As we previously said, it’s a collection of assets from various builds, a really confusing one, but nothing that contradicts previous knowledge, including the fact that GF is often misleading in their statements.
      At least two piece of info are real, Kotora and Raitora’s backsprites and the crylist mentioning Omega, Ambler and Wing.
      Also, I repeat, we also get fake leakers and fake proto stuff, and we’ve never approved any of it before, this one seemed different to the eye of many experienced people in the team. I know we’re not in the position of convincing anyone right now, but hopefully we’ll be able to clarify things in the near future.

  15. Wait… Isn’t Mew a late programmed easter Egg into the original Pokémon Red and Green?? Why it is on the beta for Capsule Monsters….

    Something is wrong here…

    1. Hi! This is not the Capumon beta! It’s a collection of assets from various builds, actually – besides the logo and some text – we don’t think the oldest graphics provided come from the earlier Capumon, the one pitched in the documents of course.

  16. I like how you can see a point in development where they decided to stop fanboyeing over stuff)

    Also, lol, my ancient dark-type Raichu evo was closer to Gorochu than I thought.

  17. BaiBai punch: probably bye-bye punch, presumably a damaging move that force-switches out the opponents pokemon with an effect similar to roar.

  18. (Yes, this is my usual username. I also go around by ax6 or asdf14396. If it looks like a spambot, you’re welcome to email me and verify it.)

    I’ve been making quite a few comments about the events that unfolded today, and I don’t like talking behind people’s backs, so here’s what I think: it’s going to take an awful lot of effort to convince people that the ZIP is real. This is going to taint just about every other release as well, primarily SW97. And the effects are already spreading.

    First of all, let’s focus on the actual release. The ZIP file seems interesting, and I’ll take your word for its legitimacy; its discovery doesn’t seem too different from what SW97 was, and I was there to witness the latter. If this is exactly what you got, though (and I have no reason to doubt your word on this matter), then it wouldn’t be too far-fetched to assume that you were scammed.

    The text files are the most revealing part of the release when it comes to determining its legitimacy. I’ve been told that the text files are the exact copies that the leaker had, completely unmodified, in Shift-JIS encoding as they would have been written back in the ’90s. Only that they aren’t. The files are encoded in little-endian UTF-16, with or without a byte order mark depending on the contents of the file. This points towards the files being generated by Windows Notepad, on a version of Windows no earlier than 2000 — thus necessarily being written several years after the games were released. Even if we were to believe that they used some fringe editor that defaulted to UTF-16, this encoding’s standard was published on 20 March 1996, which is still three weeks after the games’ official release.

    Their content doesn’t rank much higher than this. Consider the “Encounters V2.txt” file. It mentions 97 different species of wild Pokémon (65% of those in the final game), yet all of them are the species we all know and love — not a single one of the scrapped Pokémon appears in the encounter lists. Despite two thirds of the released Pokémon are named there, none of the unreleased ones are, which certainly points to those encounters having been built based on the final, released Pokémon list. This is further reinforced by the “Trainer Parties V1.txt” and “Trainer Parties V2.txt” files, which respectively list 138 and 135 distinct released Pokémon (90% of the whole set), but not a single unreleased one.
    A similar argument can be constructed for “Move Names V1.txt”, which lists 165 move names — the exact number that was released — in spite of references to as many as 237 moves in the “Evolutions and Attacks.txt” file.
    The “Move Stats.txt” file also contains 165 moves, in the same exact order and with the same exact names as the released games, with very similar stats — and written in English, in the same way your average researcher would have written about the game back in the ’90s when little was known about it, and certainly not the way a Japanese person would have written about it (the complete absence of Japanese text being the obvious indicator of this). The accuracy values are particularly revealing, because despite the actual accuracy values in-game are out of 256 (which is what gives rise to those silly percentages like 89% due to rounding error), these values are written as percentages, out of 100. (Note that the values are stored as a fraction out of 256 due to pure convenience, so there’s no logical reason to believe they were once stored as a value out of 100 and then converted to a value out of 256.)

    The “Evolutions and Attacks.txt” file is in a much better standing. It does list 190 Pokémon for starters, and without Mew (which we all know was a late addition to the game); it also shows a work-in-progress state, since many Pokémon learn no moves at all (although they learn plenty in the released games), and evolutionary lines are different — notably, there are no trade evolutions. There is no pretense regarding whether this file was ever any form of documentation — it is very obviously a dump of binary data, since moves are referenced by number and Pokémon are only named by their release names, with the scrapped 39 going unnamed.

    Finally, let’s focus on the image files. The back sprites seem legitimate, but that’s as much as we can determine — the front sprites shown on today’s stream also seemed legitimate, yet we know they are not. However, since the one bit of proof claimed over the whole set refers to the back sprites, they are the part I’m most willing to call legitimate.
    There isn’t much more to say about sprites. However, maps are a completely different story. Tools to dump complete maps as images out of a ROM’s map data are of recent development, and they only work on the public disassemblies; tools that dumped data out of whole ROMs made a lot of assumptions about addresses, and thus would only work for a specific version of the ROM (which was acceptable to binary hackers, since they only needed the tools to work on the release ROMs). Image versions of maps are completely useless to actually build maps; even today, with computers a thousand (or more) times more powerful than in the ’90s, building such a tool would be extremely complicated and failure-prone, and the resulting tool would be rather cumbersome to use — thus, nobody has tried. Therefore, we’re left to assume that the images are data dumps, but made without modern tools that allow creating such dumps. Either the distributor wrote their own dumping tools (very doubtful) or the source code contains some universal tool that worked across vastly different stages of development (unlikely). It is interesting to note that many of the maps are obviously broken (and by that I don’t mean unfinished), have incorrect tilesets, or they simply don’t make sense.

    Overall, we’re left with three possible scenarios:
    – the ZIP file contains a collection of data files taken from multiple development images at multiple points in time, plus text files written by dumpers as they came across said development images (the optimistic scenario);
    – the ZIP file contains a mixture of legitimate data, such as the back sprites, mixed in with some “artistic recreation” of development data (the middle-ground scenario); or
    – the ZIP file is an elaborate fake (the pessimistic scenario).

    The pessimistic scenario is most likely not true. But sadly, the optimistic one doesn’t look much better. I’m forced to admit, after analyzing the contents of this ZIP, that the middle-ground scenario is almost certainly the correct one.

    Now, it’s time to talk about the delivery of this information. Disclaimer: I couldn’t be there to watch the events unfold myself, but by this point we’ve all seen enough to make a judgement based on the hundreds of comments (of all kinds) that have been posted about it.

    Creating a ROM hack in order to be able to show off all of these assets in a game context instead of dropping a ZIP file was actually a nice idea, and today’s charity event was a very fitting time to show it. This was not the mistake. Everyone wanted to make an artistic recreation of SW97 when it was released, and you went ahead and actually did it with this data. That’s some wonderful work, and you have nothing but my praise for it.

    But then you tripped and started falling downhill at terminal velocity. Instead of explaining what was going on up front (which would have still gathered the interest of virtually everyone online), you made the worst possible choice: you lied to everyone about what was being shown and pretended that the ROM itself was legitimate (when you undisputably knew it was not, since you had made it). When people started questioning the obvious, the bans started to come; it is well-known by now that many people were banned for attempting to expose the ROM hack as such. And finally, you started to ask for donations in order to show more of the beta that wasn’t, showing Pokédex pages that were completely made up (since there are no front sprites, names, or any other Pokédex data in the ZIP) in exchange for further donations while pretending they were created by Game Freak.

    For this last part, for asking for money in exchange for a fake, many people online have accused you of actually committing fraud. I sincerely hope, from the deepest of my heart, that you don’t get into legal trouble because of this. But the fact that people left your stream wondering if you had committed a crime should show you the depth of the pit you dug.

    Even now, after the events have unfolded, your attitude remains off-putting. I’ve repeatedly described your responses as an elaborate form of “oopsie!” — no attempt at all at apologizing (sincerely), barely any acknowledgement of what actually happened, and no actions towards improving the outcome. I don’t believe you’re malicious, and I believe your intentions were nothing but the best. But if this is true, then you should really consider making a serious apology, perhaps in a separate post, written carefully and thoughtfully, that shows that at least you can see what went wrong and you regret the results. Because otherwise, your credibility will remain as low as it dropped this morning… and with it, the credibility of your releases.

    All in all, this was a release of some good beta content, mixed in with some questionable beta content, in the form of a very well-made ROM hack, that was delivered as nicely as a brick through the window. Overall, I’d rate it at 3 and a half stars for the ZIP, 5 stars for the ROM hack, and negative one star for the on-stream delivery.

    Which results in an average of 2½ stars. And while you might think that two and a half out of five isn’t half bad… that’s exactly what it is.

    1. I’m really impressed by your thorough analysis of the zip, I’ll show it to the other researchers. I’m glad your critic is grounded in a deeper understanding of the matter and not just “the frontsprites were fake, it’s all fake” ahah anyways sorry but at the moment “I’ve repeatedly described your responses as an elaborate form of “oopsie!” Is something that I can’t take right now, I’m sincerely apologizing to everyone both privately and on this site, I’m not addressing the stream itself because I don’t want to create drama, further misunderstandings or sparkle discussions, and I can’t speak for the people who were actually there, the mods, etc.
      I reckon their behavior is our responsibility, should’ve they dropped the stunt “it’s not a hack” at some point, of course. And we’re – again – apologizing about that. Why did it happen? Who knows, it WAS a big OOPSIE. I can’t apologize for plotting a scam because that’s just not true. Was it a scam? It’s also something I can’t address en passant on a blog, because it’s serious and needs serious scrutiny.

      About the assets themselves, it’s obviously a recent re-compilation of assets coming from various builds, but the contents seem to follow a logic and I can’t ignore the fact that 1: the leaker promptly answered to all our early tricky tests 2: the contents of some of his lists have been proven real in a matter of months 3: members of the team more experienced than me in investigating this kind of stuff think it’s real.

      Can we convince you? Apparently not, especially now that we’ve disappointed most of the community and nobody is listening to us, when the storm will pass we’re definitely going to address the matter.

      1. Between the time I posted that comment and now, I see you’ve approved several more comments to this website, and that the tone of your responses has changed towards a more serious and apologetic one. I commend you for that, and I appreciate your apologies over the matter.

        Now, as for the content itself. The issue with it is that it is an uncorrelated mess of files. If the files are real, they belong to at least four different versions of the game; while it would make sense for data to not completely add up during development, the differences are too large.

        Due to the files’ encoding, we know for a fact that they were created years after the games were officially released, in all territories. The natural assumption, therefore, is that someone dumped the data from some prototype and wrote it up. There are many details about the data (moves referenced by ID, move accuracy given as a percentage, all names (Pokémon, moves, maps) referring to the release English versions…) that support this hypothesis.

        And thus, we wind up with a collection of text files written by dumpers who came across prototypes long after the official releases. Do they belong to the same prototype? We know they don’t. Were they written by the same person? We don’t know, although the writing style differences suggest that they weren’t. Are they all equally valid? We have no way of knowing, and this is where the issue lies. Since the data files are pretty much uncorrelated, even if part of the data is proven to be real, that says nothing about the legitimacy of the rest. This may very well be a mix of legitimate data and made-up data, and unless there is a way of proving the validity of each and every file by itself, those doubts will never go away.

        And that’s really the reason I’m casting my doubts over the data collection. I’m almost certain some of it is real. But that doesn’t mean all of it is.

    2. The V2 map data is definitely either fake or re-assembled using tileset from US Pokémon Red or Blue ROM. Because Japanese Pokemon Red & Green used different tileset graphics than US Pokémon Red & Blue. This is because US Pokémon Red & Blue uses the codebase and graphics from Japanese Pokemon Blue, the original “third version” to Japanese Red & Green. It would be impossible for a Pocket Monsters beta to have the US tileset of Pokemon Red & Blue (which is the graphics from Japan’s Blue version) because those graphics were made for the Japanese Blue version, which came out AFTER Red & Green we’re finished and released.

      This is why in US Pokémon Red & Blue and Japan’s Blue, Route 23 has a different tileset from the rest of the game. Because that’s actually the graphics used originally by Japanese Pokemon Red & Green for the entire game. When they updated the graphics for Japanese Pokemon Blue, they forgot to change Route 23’s tileset.

    1. Of course it’s true, that’s why it took so long to post in the first place, because we had to crosscheck and research the info provided. But I understand we’re not currently in the position to convince anyone. Rest assured that we don’t post all the rumors about proto-assets we get 😉 but only time will tell I guess.

  19. Did the Id list in this build end with Totartle and the event sprite Missingos or with the Oddish and Bellsprout lines? Was Mew already in Omega’s place or was he a Missingno? I want an idea of how just how late GF chose to scrub the betas out.

    It feels like this article jumps around the newly discovered build, later development, earlier development and the recreation effort indiscriminately. As someone whose read all the sites articles (which appears to be a prerequisite) I’m still confused by some things as they appeared in this rom…if this draft doesn’t get serious revision it’s going to be chaos. Maybe the team should revise this as the Helix Chamber community version and also draft a viral friendly version without the period theory, casual references the ‘Chief became Blaine’ theory, Spaceworld and the manga leak without presenting context, and all discussion of the recreated front sprites/types/names that’s not couched as ‘artist interpretation’ or something (and even then people will want to believe they’re 100% real…).

    1. Yea I agree, we need a more casual version as well as more in-depth articles, we’re beginning to work on them. Anyways the problem with these assets is that they’re all over the place, coming from different builds (at least 3) which were apparently endlessly tweaked by GF, so identifying the strata is … ultimately a waste of time for now 🙁
      The roster is as mysterious as it gets, again we think it’s a mix of the various strata, the hypothesis of a 190 roster, however, doesn’t seem plausible anymore, 032, PRE-MEOWTH and the fossils make us wonder whether they replaced some older Missingno., also the Oddish-Victreebel segment now it’s starting to look like a really late addition.

  20. to balance out all the complaints (which you are addressing graciously), I just wanna say, maybe you messed up, but whatever. there was obviously no malicious intent and this stuff is wicked cool. too bad the guy disappeared like that, but you all are doing a great service to pokefans everywhere, as well as game prototype documentation as a whole. hoping some day the rest of what’s hidden is revealed.

    1. thanks 😀 we can’t wait for someone to drop the front sprites, or even the rest of the 92/93 polls, those might hide more scrapped mons that didn’t even make it in the ID list!

    1. I mean, we wish. The important thing now is not to try and hunt the leaker down XD because such behaviour might discourage future leakers 😀

  21. Dunno if you approved the streamer or not, don’t care either. They were representing this, to the point of sheer annoyance, as “A PERFECTLY NORMAL COPY OF GREEN” and “NO ITS NOT A ROMHACK.” Already the bait and switch with Spaceworld was over the line, but then to insult people’s intelligence like this (and to have Twitch mods deleting romhack comments too) was garbage.

    And then, with these two frauds, you got people soliciting donations the entire time. I get it’s for a good cause, but that’s not the point. The idea was that one thing was going to be presented, then we’re told we’re getting something else, and then it comes out “oh… well… that wasn’t entirely the case.” The coy tweets while it was all happening just perpetuated scummy behavior. “We’ll tell you after, promise!” No, be honest and forthright to begin with.

  22. I’m left with… several questions. Considering how the “Pokeathon” was held in such of an impromptu manner, it’s beginning to make me question the authenticity of the “documentation”.

    Like, the sprites make sense, sure. But what about the trainer sprites? What about the maps? Are those just from that ROM hack? If they aren’t, where ARE they from?

    In this very article, you state that “Additionally, the formats in which we were given the files were non-negotiable (i.e. ROMs were off the table).” But what DOES that mean? What format were these given in? Was it said ROM hack, or was it from an actual Generation I prototype? After what happened with Pokemon Prism, I can understand why you wouldn’t want to release a prototype after building up a large-scale stream. However, what I’m after is: seeing these assets as they are seen in the source format. I want to see footage.

    PLEASE prove me wrong. I want to believe in these dreams, but uncertainty fogs the way.

    1. We can’t really be sure whether the assets came from a Rom/a prototype or if the leaker just had them as a sprite dump in the first place. They are of course in our opinion real, because we tested the leaker’s legitimacy in a couple of occasions since we luckily had a way to crosscheck its claims. The last proof came in November, when the second Capumon spritesheet was revealed on NHK and we found out that the “Capumon” Crylist we were given actually listed the never before seen proto-names Ambler, Wing and Omega associated to the correct Pokémon. I totally understand many of you are not satisfied with the explanations and won’t trust us because of the ambiguous way we handled this (sorry), I hope we will be able to find more valid arguments in the near future.

  23. So many of the historic Pokemon narratives are tied up in our childhood. The Beta-mon audience has a sense of ownership that stretches back decades. When my best friend invented non-canon details for our schoolyard games back in 1999, I would most often disagree with him. That was part of the fun.

    If you are positioning yourself as an archival service, the rules of archivers might apply. Collect, Preserve, Display. Not being community led like many of our favourite wikis sets you up to a higher level of scrutiny. This has its benefits and obligations.

    Perhaps a positive step forward from this misstep would be a mission statement of some kind, a clear positioning of HC, separate from Bulbapedia, Cutting Room Floor etc.

    Might I also suggest a companion website for all extrapolation? I’d love to see those fan-sprites again, but I’d need to be able to ignore them if I wanted. The best part of disagreeing with my friend was seeing his creativity at work, then deciding if it fit with my interpretation.

    Keep up the good work.

  24. Obviously Shinjuku Jack(新宿ジャック) is inspired by Shinjuku Jackie(新宿ジャッキー) ,a famous japanese Virtua Fighter player.

    1. Indeed – came here to point this out based on a couple of tweets I’d seen, so I’m glad someone’s pointed it out already. Here are the tweets in question; the first points out the design similarities to Virtua Fighter’s “Jacky”:


      And the second clarifies the name’s connection to famed player “Shinjuku Jacky”:


      Hopefully the article can be updated in light of this info, as the connection seems pretty solid. It seems likely from this that, rather than a robot, Jack was actually intended to be a “virtual” polygonal trainer. Makes you wonder if there was any planned connection to a certain polygonal Pokémon…?

  25. hi there!
    sorry you`ve been getting so much backlash on this, but to be honest I was pretty upset too
    even if your intention was really nice, you basically turned real prototype pokemon designs into fakemon…

    the way this article was presented, with the made up names and the interpretations you guys had was annoying as well, I`m glad that is gone.

    it really was a very disrespectful way to handle the material that was given to you for preservation, i hope the guy who has the rom isn’t japanese, either way he’s probably really pissed off ( but if he’s japanese he’s probably not contacting again ).

    I think if you presented the material first then showed your interpretation it wouldn’t have been so shocking.

    putting that aside, a few interpretations:

    -balloonda probably evolves to jigglypuff, which would explain why it`s a balloon pokemon and why they bothered to even make a baby jigglypuff later
    -considering how exaggerated some front designs are, I`m sure zubat pre evo had room for a face
    -you talk a lot about cries, but none are in the article itself
    -marowak 2 is holding the baby and the baby is holding a bone
    -the pokemon you listed as gorochu kind of looks more like it`d be a dewgong evolution, I heard that at this point there was a “fire dewgong”
    -the pokemon you listed as blastoise pre evo seems more like a baby bear
    the bubble in front is either a pacifier or a sign of napping, it`s more likely a snorlax pre evo. ( looking at it again, it’s definitely a tauros pre evo, has a nose ring ( common in japan in cattle in anime and games ) and the fur around the neck. )
    -86 and 87 are probably what later became hitmonchan and hitmonlee
    -you guys didn`t make much of a deal about how the protag has a whip like in the concept art which makes it more justifiable why a lot of the trainers have them too, they were pokemon tamers.
    -i downloaded the assets but it doesn’t have that second world map from the pics, that look way more interesting with the hometown at the indigo plateau.
    -gyaoon probably became charizard, its front sprite even looks a lot like charizard in pokemon GSC opening when he uses flamethrower
    -you guys don’t mention at all how at this point the routes are literal roads with road signs, vehicles and everything in the image files or how the pokemon contests seems to be present in some form.

    since ape inc was involved, probably they changed this idea of pokemon as its own world into something more modern because of the popularity of the mother series
    so some stuff like the protagonists whip and dampe were gone and instead replaced with something more neutral and nice, probably nintendo requested it too since a lot of the ideas seemed a bit too dark/rough for kids ( from a western point of view )
    the pokemon world at this point was kind of harsh, and it probably had a lot darker stuff than the final game, but it was also a lot more lively and self contained, tbh this data makes it look like the pokemon we didn’t get was way more fun.

    1. hey! we also initially thought 175 was a walrus, but actually that pokemon has got evo data in the assets and it’s clearly stated it evolves from raichu. Where is the Fire dewgong info from? Perhaps Bomushika from the Spaceworld demo? 😀 cheers

    2. So That possible baby Gyoon could also be a pre-evolution of Aerodactyl. If you’ll notice Aerodactyl is the only fossil pokemon to not have an evolution line.

  26. After all this is analyzed, not to sound hasty, but it is extremely possible that another, more willing person has another copy of Gen 1 beta, and could share the front sprites, etc. Would it be possible to continue the hunt even more to obtain the info from another, more willing, collector? Not to sound hasty or anything, maybe it’s just cause I might be just a tad TOO hyped. lol

    1. of course! But not an “active” hunt for the leaker, they’re people who could get in serious trouble for doing what they do.

  27. I think how you went about things was wrong but I really respect that you owned your mistake and corrected it.
    However I can’t say I feel the same about this collector. I understand partly why they’d only show a few things but it’s still very selfish. They are lucky enough to own that beta and probably a lot of other things.
    They really don’t need to hide it away to make more money then what they already are lucky enough to have. Also they sound kind of rude expecting you to negotiate but they refuse to on their part.
    I’m sorry if that is harsh to say. I just feel is person is pretty selfish.

    1. thanks for the understanding :), let’s hope he, gf or whoever will drop more info in the near future, these assets sure raise a lot of new questions! cheers!

  28. There were five different pokémon that couldn’t learn Hyper Beam, right? There’s only two slots and unless I’ve been looking wrong neither Animon nor Madame appear anywhere there, so you’ve got 3 different pokémon vying for two different spots.

  29. I just wanted to thank you all for your hard work! While the reveal may have had its bumps (I attended the entire stream from platinum running late to the end of the “green” run), but I’m glad this information is out there now. Having the evolution data and backsprites (and more) is better than not! We’re closer than ever to these beta and proto- pokemon and to understanding the development of this amazing series!
    Before the Spaceworld leak, I never thought I’d see the real deal. After the SW leak, I didn’t think I’d see anything like it again. Clearly, dreams DO come true! I’m definitely not ruling out the collector, or a different collector, revealing more information or those front sprites. 🙂
    Thanks again all of you!

  30. Yeah, I’m finding it really difficult to imagine a scenario in which this is real. For it to be true, someone has to possess the following:

    At least three early Red/Green ROMs
    An early Blue ROM
    An early Yellow ROM, or information directly from the developers as to what the unused song was for.

    Additionally, the supposed donor is clearly not Japanese, raising serious questions as to how they managed to obtain multiple prototypes that were never even shown off to the public, and should logically have never left GAME FREAK’s offices.

    The amount of technical knowledge they supposedly possess also makes little sense. They were apparently able to rip some assets and data from the ROMs, but not all of it? It’s quite strange how they were apparently incapable of ripping all the sprites and text, but were able to rip the more complex map data just fine.

    One explanation for this would be that the donor only had the data they sent you and not the ROMs themselves, but that just raises further questions, and IMO is even harder to believe.

    If this really is a hoax, there is really only one group of people outside of GAME FREAK themselves who knew enough about the development of Red and Green to pull it off. And it would explain how it aligns so well with your theories, and how the hoaxer possessed corroborating information that you also had but hadn’t made public yet. It wouldn’t even require the whole group, just one rogue person attempting to destroy your reputation.

    Regardless of its authenticity, you really shouldn’t have published this as 100% real prototype stuff with the information you had. You should have released it with the caveat that it’s credible, but not 100% confirmed.

    Seriously disappointed in you for this, and this isn’t even getting into the Pokeathon hack shenanigans.

    1. Regardless of the Pokethon hack, however, we took our time to analyze and crosscheck the contents, and we believe it’s real. We get we’re not in the position to say to trust our judgment right now, but we don’t think the assets are fake, that’s why we didn’t put a disclaimer on it.

  31. I’m glad you guys are being receptive of criticism. The work you do is important; historic even. I want you guys to be able to recover from this. You gave us SpaceWorld, one of the most monumental finds of gaming history. I don’t want you to be brought down by this mistake; I don’t think anyone does. I think a lot of the criticism you are receiving is coming from passionate people who are angry, yes, but angry because they know you can do better because you have before. You have a lot of work to do to make up for this. You have severely damaged your reputation to the point where I can’t imagine that another someone with a prototype or even that same person would want to approach you with the information they have and it’s a shame. I personally believe these assets to be real, but I understand why other people don’t considering you outright lied for hours on your stream. I hope you can fix things because the information you have given to the community is unrivaled. Best regards.

  32. Hey guys. I was wondering if you could clear up some information in regards to the files you were sent. If I am not mistaken, the zip the collector provided was unaltered from what you released, correct?

  33. So, to be clear, it seems one of the big reasons you believe this to be real is that you obtained this data which contained the names Ambler, Omega, and Wing, before they showed up on the official NHK broadcast showing off capsule monsters pitch?

  34. Hello, I haven’t watched the broadcast, and “read the documentation” (where? no link), I only discover this news today, sorry.
    So I haven’t read the former version of this article you claim went through several edits.
    But believe me, it’s a little confusing to read now for a beginner.

    I assumed you recived this zip file recently, like a month it was for spaceworld beta gold. After your last article “how capumon stole christmas”, let’s say.

    But how in such a short time they could have made a rom hack ?

  35. Quick theory: ID 121 could likely be a middle stage between 63 and 122. The uncovering of 127 as a middle stage between Psyduck and Golduck shows that a 3-stage line during that period isn’t out of the question.

  36. The comment posted by Anonymous at 7:00 pm on February 19, 2019 are very important points.

    How did it leak out about the prototype version which The Pokemon Company, Nintendo, Creatures Inc, Game Freak, all of these four companies, or one of them possess? The donor should have explained it.

    Agatha’s party listed on the prototype trainer list is made up of Pokemon other than Ghost-type. As with the product version, Agatha’s party is located next to the Channelers. Because both agatha and channeler are ghost type users, this arrangement is natural. However, the prototype agatha is not a ghost type user. We can not rationally explain that Agatha is next to the channeler. I have no choice but to judge that it is a fake made based on the setting that it was a former rival of Oak.

  37. Hi, Gabe here. I would like some quick confirmation.

    Was it true that the collector outright refused to give the front sprites, beta ROM and Capsule Monsters ROM to you, or were there other reasons?

    I am eager to see incoming info as well, and I hope that you 100% do indeed succeed in finding the ROMs and front sprites, so we can piece everything together.

    I’m rooting for you!

    –Gabe S.

    1. thanks gabe! What the collector said has been reported, but there’s no point in believing he had more than what he gave us, unfortunately 🙁

  38. Something is fishy about the V2 maps – they all use the tileset from the International release of Red & Blue, which was a different tileset that was used from the first release of Japan’s Red & Green.

    Pokémon Red & Blue uses the codebase and graphics from the Japanese Blue version. The Japanese Blue version came out almost a year after the original Red & Green. It altered the sprites of the Pokémon AND the graphics for the overworld tileset (except in Route 23 – this is why Route 23 in International Red & Blue uses different trees from the rest of the game, because that was actually the trees used by Japanese Red & Green for the entire game).

    The fact that the V2 maps uses tileset graphics from Japanese Blue version instead of the original Japanese Red & Green versions that preceded it is very suspicious – because those graphics weren’t made until AFTER Red & Green we’re finished and released (yeah that’s how you can tell the romhack was also fake). It leads me to believe that either the maps are fake, or that the map data was assembled using the tileset graphics from a US Red or Blue game, which might explain why a lot of the maps look corrupt.

  39. 135 couldn’t by any chance the baby Doduo from Spaceworld could it? It looks a bit like it and is next to the baby Meowth, also from Spaceworld.

  40. Just wanted to say that this is really cool.
    I really want to believe that this is real but I also hope that nothing will happen to this just because of the rom hack situation.

  41. Here’s something interesting regarding some of those deleted evolutions. Pokémon get stat increases when they evolve, but for gen 1, Satoshi (who I’m pretty said was the one in charging of determining how strong a Pokémon would be) seemed to have had a pattern.

    For Pokémon that evolve twice, the first evolution gets around a 75 stat increase (15 avg), and the exceptions still follow that 15 pt increase for most of their stats (the big exceptions are Dragonair, and the two cocoons, and those make sense because they’re special being the dragon and two very early evos.). For the second evolution, that stat increase will either be still 75 pts (15 avg again pt) or 100 (a 20 avg; this is true for the starters, the nidos, and the Pidgeot). The dragon line and the two bugs are exception again (Vileplume only increases by 70, but most of its stats beside speed increase by 15 pts, so it follows the pattern for the most part).

    For Pokémon that evolve once, the stat increase is 125 pts. But there are exceptions, and when I look at them I find something interesting because quite a few are Pokémon with removed evolutions.

    Rapidash and Marowak have stat increases of 75 pts in the final version. That’s low for a two stage evolution line, but it fits the pattern of a three stager. Ninetales and Golbat have an increase of 175, and that makes sense if Vulpix and Zubat were given the stats of their pre-evolutions (Zubats stats are particularly low as is). Golduck increases by 150 pts, meaning that two 75 stat increases are possible, as if the middle evolution between it and Psyduck was cut out. Seaking has a stat increase of 100, like the final stage version in a 3 parter. If those back sprites are fake, then the person making them did their homework.

    Granted, Magneton and Persian follow the pattern for a 2 stage line and they have some other forms it seems. But Magneton and Persian are odd because while other lines were inserted next to each other, these two aren’t. My guess is that either these other stages were never given stats to begin with or they were deleted rather early and Satoshi made adjustments. I can’t explain Gorochu either because Raichu gains 135 more points upon evolving.

    There are also a few also have stat increases that suggest a deleted form but likely don’t have one for the missing nos. Fearow, Arcanine, Tentacruel, Exeggutor, and Starmie, and I have no answers for them besides guesses.

    I’m making a lot of assumptions, but if my hypothesis is correct, than Game Freak had already had a lot of stats written done before they started removing many Pokémon, and its kind of disappointing you weren’t able to get them.

    Also that first encounter table, if it is real, might the oldest thing you got given it only contains monsters from the Capumon pitch.

    1. aw wow this is a fascinating analysis! I wish we could’ve included in the previous research to unmask the missing third stages. It’s still very interesting, as Gorochu was probably designed in P3 but put on hold til quite late, and of course we think baby-meowth overwrote someone else at the last moment. Also the others you mentioned are among the “capumon” and might follow older patterns of growth. What is even more interesting is that 134 and 135 might still hide some scrapped p1 evolution, hence the research isn’t over. Thanks!

  42. Although the GCCX images are very blurry, the back sprites of Kotora and Raitora in the .zip don’t seem to match up with the GCCX sprites.

    The big detail to me, is that in the alleged beta sprites, we see their tails rise up behind their left ears. The pixels in the GCCX video screengrabs do not seem to reflect this.

    Of course, they could have gone through revisions between builds. But this does have me skeptical.

    1. The backsprites surely went through revision, just look at Kadabra and Slowbro’s in the GCCX clip. Also you may consider the enhancement process is prone to produce artifacts. Anyways we just asked for “159 and 160” and the leaker produced exactly what we expected identity-wise, it’s either real or a freaky coincidence XD

  43. So, one of the most baffling things about this to me is the whole “Silph Co. League” thing. The article claims the donor provided information to you about it, what it was intended to be, and the development of it, but nothing besides some maps and encounter data is included in the folder that was provided to the public. Would it be too much to ask what exactly they shared? If the donor provided significant amounts of information to you outside of the folder, why have the relevant logs of their communication not been shared (with any personally identifying information removed?)

    In regards to what HAS been shared, while the whole “Silph League Dungeon” thing does answer some questions (such as why Moltres and Mewtwo are so out of place, the blank maps around Lance’s room, the displaced Silph Co. maps, and the blank maps around the Cable Club and three non-Lance Elite 4 member rooms) it raises further questions too. And many of these revolve around our good friend, CHIEF.

    In regards to Blaine being given the Silph Chief’s design… while we do know that Blaine was given his final design late in development, in the final game, the scientist fire-type gym leader Blaine is pretty well connected to the rest of the science-themed Cinnabar Island, with his gym based around quizzes, the lab, and the burned mansion full of fire types where Mewtwo was created all being found on Cinnabar. The only things that seem “off” are the lack of the volcano (despite Cinnabar being described as volcanic) and how out of place the whole “get the key to Blaine’s gym from the Pokemon Mansion” thing is. (FRLG and LGPE fix this somewhat by adding an NPC to the Pokemon Center actually hinting what you’re supposed to do, and establishing that the mansion belonged to Blaine’s friend, presumably Mr. Fuji)

    Regardless, either Blaine and the content surrounding him was originally completely different to fit his military design (gym, mansion, lab), the content was the same but he had an unfitting design for some reason, or he inherited a lot more from Chief than just his design (such as the Pokemon Mansion). Or a combination of the first two: His part of the game was originally very different, but was changed to the final one relatively early in development, with his design not being updated to match until close to release.

    One other problem with the whole “Blaine was given Chief’s sprite” is that one of the very few things we previously knew about the Silph Chief is that he was intended to be found in the Safari Zone very late in development, to the point he is still mentioned by invisible NPCs there in the final release of Red and Green. So much more of the Chief seemingly survived to the final release than of original Blaine, or any of the Silph League stuff for that matter. The stuff provided by your donor also doesn’t seem to have anything about his connection to the Safari Zone.

    This got more rambley than I intended it to.

  44. Also, I noticed in the version 1 teams that it says in the final rival Battle (“Rival 3”), the rival has a Blastoise, which he evolved from his Squirtle. This should not be possible because of the scrapped Wartortle evolution. The rival should have had that instead of Blastoise.

    –Gabe S.

        1. that it’s a fake? We don’t know have a way to know when that rival team was made, or when the two evo lines were merged :s

        2. The trainer list and evolution settings in the ZIP file were obtained from the same prototype ROM.
          Isn’t it?

          It means that I can point out contradiction. This prototype is fake.

          1. the assets are a collection of “snapshots” of Pokémon development from 1990 to 1995, apart for a few of them there’s no way to say with certainty which one co-existed with the other, only educated guesses.

  45. That means, taking into account that the background of the leak is unknown, there is nothing to guarantee the reliability of this trial version.

    If possible we would like to believe, but we have to evaluate at the present moment like this.

  46. Mr.Pean:

    UPDATE: As far as I can understand, everything provided was obtained from multiple builds, so we don’t have definitive way to call it fake or not.

    –Gabe S.

  47. Satoshi Tajiri reveals in the autobiography that when the development machine crashes the prototype data has been lost.

    It is Junichi Masuda that restored the machine and it seems that after restoration, it is written that they have only prototypes close to the product version in addition to the product version.

    Because of this background, I think that it will not happen that the initial prototype that has not kept its original form leaks in the form of data.

    1. I must be misinterpreting this because “development machine” singular makes it sound like there was only one computer with a dev kit at Game Freak and everyone had to take turns using it.

      1. Maybe some of them had a dated personal computer of their own, but the company itself likely just owned one. This is not unlike most small game companies started. It is not impossible that some assets survived in somebody’s elses hard drive, but the main thing died at the incident.

  48. Well, we did have the early shark Pokemon, and the elephant, and the deer, and Crocky, and Barunda, and the cactus from the Satoshi Tajiri Manga.

    If you say that things only close to the product survived, that could be the reason why the teams in the assets have only the final 151–these were the only teams that survived.

    As for faking the other sprites/Pokemon, I doubt it. It would take way to much work, and way too much time.

    1. Also, the electric cats being here supports Miyamoto’s interview where he said there were 190 Pokemon, and the rest were “saved for later” (the cats later came back in the Spaceworld ’97 Gold and Silver demo).

  49. There is no need for the leaked person to reveal the true identity, but he should have revealed in what way he got it in what way.

    Pokemon Red and Green were made in the 1990s, it is said that it was not anticipated to be released in countries other than Japan.
    Even now, GAME FREAK and Nintendo of Japan have possess it secretly. perhaps.

    So, who can steal their prototypes?

    1. Well, we don’t exactly know. Somebody who sold it to the black market, then a collector got ahold of it? Most likely. That’s the same thing that we are still asking ourselves for the Spaceworld demo as well.

  50. In Japan, leak of prototype data and source code from within a major game company has not happened to the best of my knowledge.

    Therefore, It is difficult to trust as real if there is suspicious points.

    As I mentioned earlier, the early prototype has been lost and there are things relatively close to the product version remaining.

    However, map images included in ZIP are far from the product version.

    It is unnatural that this perfection degree is at the time of development late.

  51. I am curious as to where the version 1 maps indeed came from. If they actually are fake, the collector would have had to spend a ton of time, as well as money, making a custom tileset. At the long shot that it’s real, this information is huge and could change the video game industry.

  52. Perhaps 32 was an older back sprite for male Nidoran, it doesn’t look too different from the final and may have been bigger because they planned for it to evolve straight to Nidoking. Number 137 might be one of the Pokemon on the cover of the Capsule Monsters booklet as they both have two horns, similar head shape and two small limbs on the front. Another possibility is that it’s the hatchling shown in one of the CM illustrations.

    In each mention of a cry you guys should’ve added them because only the hardcore are going to remember their order and how they sound.

  53. I spent some time tinkering with the V1 maps. They are in various states of incompletion, with the start of the game being the parts with most work on, and the last cities (Saffron, Fuschia and Cinnibar) being clearly placeholders.

    Here’s something kinda interesting I noticed:

    * Rather than having a PokéCenter and a PokéMart, cities have a Pokémon Center and an Inn.
    * There’s no maps for Inns nor V1 Pokémon Centers included in the assets.
    * There is, however, tileset and blockset for inns. It’s very similar to the final Pokémon Center tileset, but it features tiles for what seems to be bars for a cage, and roofs in a disctintive, otherwise unused style. It also features a “PokéBall” poster, similar to the one present in all Pokémon Centers, but with the letter CAP rather than PC (or PMC, going by V1 graphics), suggesting these assets are from Capumon rather than Pokémon.
    * There’s also, tileset, blockset and maps for V1 PokéMarts, but these are in a very early state, mostly empty and incomplete, especially compared to final.
    * Pokémon Centers lack both a tileset and a blockset, unlike PokéMarts and Inns.
    * In the overworld blockset, The CVS block meant to be used as the sign in front of PokéMarts, is not used correctly anywhere in the maps. It is used, though, but it seems to be standing in for a different block, suggesting the CVS block replaced a different block in the blockset, and is a late addition.
    * There’s also a misplaced block that seems to correspond to a building wall. This one got replaced with a block full of garbage.

    All of these facts lead me to believe that, at some point, Inns had both stores and healing places within, pretty much like the Victory Road PokéCenter, alongside other stuff like trading/buying/selling monsters that got discarded once the idea of separating Inns into Pokémon Centers and PokéMarts was made. V1 overworld data seems to be from around the time this was being worked on.

    Here are some more things I noticed:

    * Speaking of the overworld tileset, there’s two unused tiles: a very simple pattern of dark spots and, oddly enough, the tile used for the floor in your house in R&G
    * Tons of maps use what seems to be roads meant for vehicles. I’m not saying you were meant to drive around, but rather than asphalt with painted lines and vehicles were meant to be part of the landscape at some point in Kanto routes. This trend carried on to V2, as you’ll notice a lot of the maps seem to have what seems to be either two roads, or a big road on the exits connecting them to routes.
    * Related to the last point: there’s no place in V1 Vermillion where a Machop could’ve been stomping the ground to build an apartment. There is, however, a place that looks very much like an empty lot with lots of tall grass. There’s also what seems to be a place “under construction” in Celadon, with a bunch of very rudimentary fences and a lot of signposts around.
    * The same “rudimentary fences + signposts” pattern can be seen in Viridian, where a fence is blocking the path towards Indigo Plateau. This is weird because GF tends to use actors blocking your path to halt your progress when the place you’re trying to go is out of reach, and rarely modifies maps to block players from reaching a place (only instance I can think of is Spaceworld and north from Silent Hills, where it connected with Kanto)
    * “Oak’s School” data seems to be from V2 rather than V1, as it already features a table for three PokéBalls, something we know was implemented very late in development. There doesn’t seem to be an Oak’s Lab building in Pallet Town. There is, however, a school in final Viridian City, but I don’t think it means anything.
    * Most worked on routes are 1, 2 and 19, oddly enough. 19 features a house. The rest of the routes are all placeholders,but seems like at least signposts were already being coded (See Route 25)
    * No Viridian Forest, also. Route 2 reminds me to a very rough and unfinished G/S Route 2 with the mini-maze standing in for Viridian Forest. There’s also a cave at the end of V1 Route 2. There’s an entrance to the Diglett’s Tunnel on Route 2 in final, but I don’t think this cave was meant to connect Pewter with Vermillion like that in V1.
    * Weirdest town award goes to Lavender.
    * Lots of towns use long grass as
    * No cuttable trees. Considering how much did GF toy around with HMs and overworld abilities (the “surfboard” unused item in Gen 1 was, according to this data, meant to be called “Lapras”, which was already a Pokémon and even had a front sprite… Maybe at some point the ideas was that you carried your monsters in your bag, like every other item?)
    * Not map-replated, but the item list shows three fossils: Dome, Helix and Wing. The Wing Fossil is obviously meant for Aerodactyl, and this wouldn’t be a problem but there’s also Old Amber in the item list, suggesting the Old Amber was meant for something else. Considering how many changes to the whole Oak/Silph’s Chief/Mr. Fuji/Rocket Team storyline were made during development, I’m guessing the idea for a “genetic Pokémon” was there since the beginning, and the Old Amber was originally meant to be related to Mew/Mewtwo
    * Also that theory about Erika being a ghost trainer because of her clothes does lose a lot of credibility when you look at the Gym 5 map (she’s marked as 5 in Sugimori sketches) and see it full of flower pots. Sure, the Gym maps are most likely V2 maps (they use the “Oak School” tileset, ust like Oak’s Lab uses the Dojo/Gym tileset in final, and since the Oak School tileset is most likely V2, the maps are also most likely V2) and it’s weird that she’s using the thing backwards and that her PokéBall is floating and all, but at least by V2 she’s already a grass trainer. Since she seems to have been created specifically for V2, I can’t help but feel that maybe the idea was to make her a ghost or psychic trainer at some point, but she was quickly repurposed to be a grass trainer. We only have her sprite to go with this like of thought, so I’m more inclined to think that there’s something weird going on with her during the development, but she (as in, Erika the Gym Leader) was never meant to be a ghost trainer, let alone an actual ghost.

    I noticed some other stuff also, but that’s the most important/interesting things I noticed that doesn’t sound like I’m going through a Pepe Silvia situation. Hope this helps someone!

    1. Is there any indepth information on the original storyline aspects? That’s the kind of beta stuff I’m most interested in, but my googling hasn’t turned up much outside of the invisible NPC text boxes in the Safari Zone and some of the earlier tweets regarding Blaine’s original design.

      1. GlitterBerri had a translation of the few pages on the Capumon pitch. One of the first pages is this pitch in the form of a first-person narration about a kid that explains what Capsule Monsters’s about. In this text, the unnamed kid describes the game’s world as a “dungeon-covered overworld”, and trading/dealing in monsters seems to be one of the most important aspects of the game, as most of the narration revolves around it. This is where the “Green Dragon” monster’s mentioned, which was later speculated to be the identity of Dragon4 and, lastly, Strike/Scyther. It also mentions other monsters, such as fireflies and “powerkings”. This is all very, *very* early info. From 1990, on the first pitch document if I’m not mistaken. Other than that, we just have a couple snippets of information…

        Check it out! https://web.archive.org/web/20180305065713/http://www.glitterberri.com/pokemon-red-blue/early-concept-art/2/

      2. Unfortunately the proto-story line(S) are still a mystery .-. you can make your own speculations based on the assets though, we’ll be glad to hear them 😀

  54. As far as I can see, did you manage to get the actual music and cries for the beta, or was it just the text data? Because I’m really interested in seeing what was listed as “Unused theme: Giovanni” in the music and cries text document. 🙂

    –Gabe S.

    1. Might refer to the actual unused songs in the final games.

      “Unused RGBY Theme: Trading” might be the song you can find by searching “Pokémon Red/Blue Unused Song” in YouTube. This song is very damaged/in a different format to the rest, so it sounds totally off. There are a bunch of restorations.

      “Unused Y Theme: Giovanni” might be the song you can find by searching “Unused Theme [Pokémon Yellow]” in YouTube. This song is usually described as sounding very similar to the boss battle theme on Bushi Seiryuuden, another Game Freak game. It apparently got used in Kalos’ Victory Road.

  55. There’s another, generally overlooked detail in the data of the final versions that suggests the veracity of this info. There are 9 gyms in the dump, and there are exactly 9 unused badge names in the final Japanese version. TCRF translates them as Thunder, Shell, Bodhisattva (?), Falcon, Chill, Friendship, Rose, Fireball and Gold. Some of these correspond to final gym types, and Falcon sounds like it would make sense for Giovanni’s badge when he was still a Flying-type trainer. I wouldn’t be surprised if the colored badge names come from a later period, probably once Super Game Boy compatibility was part of development.

  56. For the names on the picture file names, such as Bloonder for Barunda, how did those names come to be? Did Helix Chamber staff name them or?

  57. One thing that has me skeptical about this. If Mew was originally added as sorta last minute addition. How can he exist within this list with all of the other Missingno. I would assume that if Mew was added after the debug info was taken out, wouldn’t they NEED those debug features to be able to remove all of these other Pokemon at the same time?

    1. hey! Actually we don’t believe these assets all came from a “proto-green” build, It’s more than probable that they are snapshots of various development phases and builds that range from late-Capumon to early-Blue, 😀

  58. Awesome post but there is something that puzzles me. In the second Early Kanto-map, there are tiles used for signs and doors that GF used when developing the Blue version. In the original Red&Green-versions they used a different tileset for signposts and doors. It seems strange to me that they during developement of the original games used a tileset that got scrapped and then later re-use the scrapped tileset in the Blue version.

    Something seems fishy about that. HC: did you ask the provider of thos leak about that? What was the response?

    PS. I love your work!

    1. Hey! Well in general there are a few cases of scrapped-and-reused things in GF history, however in this case I think one answer could be the RG assets were probably replaced with the blue ones in the renderings because they were somehow missing from the leaker’s set! cheers! thanks for the love

  59. As others have pointed out, even if everything in the .zip file is authentic, the data has been dumped and encoded in such a way that makes verification impossible. In other words, we cannot say with certainty that it’s real; we can only say that it doesn’t outright contradict what we know so far. Even that might be a stretch, as other posters have raised some pretty compelling points. But the most frustrating parts of this reveal are:

    1. The actual carts are inching closer to bit failure every day.
    2. The leaker is aware of point #1 but still gave you this instead of the full contents.
    3. In doing point #2 he made it less credible because the modern encoding of supposedly old files is the equivalent of putting “Unregistered Hypercam 2” in the corner of every frame.
    4. He contacted Helix Chamber specifically yet handed over the data in a format worthless to the stated purpose of HC, unless you take this Mystery Leaker at face value.

  60. Ichitarou could have worked as Brock’s original Japanese name too, since all of his siblings in Japanese have their names number-based. Maybe Takeshi was a late name change?

    1. Nothing’s impossible, but I have a hard time seeing the anime being influenced by beta stuff, it’s not like it was a planned tie-in from the start. It’s more likely his siblings have number-based names because that’s a quick and easy way to name nine minor characters in Japanese.

  61. Thank you for this! I enjoy every bit (pun intended!) of it.

    When I saw 63 it reminded me of Slowbro’s famous “Shellder” tail. Considering it was made a Pokémon – Turbann – in the gold and silver betas of 1997, I feel it would be plausible.

  62. I’ve read somewhere that Rhydon’s original index was 00, and Gyaoon (protoTyranitar?) was 01. Rhydon was later indexed as 01, replacing Gyaoon, which later resurfaces as index number 31.
    My guess is that index number 31 and 32 were supposed to be the original spots for Nidorina and Nidorino, but was later adjusted to accommodate the reindexing of Rhydon from 00 to 01, and Gyaoon from 01 to 31.

    With regards to index number 21, I hypothesize that the slot went through multiple reiterations before eventually scrapped and being the first MissingNo slot where Mew was later indexed.

    My theory is that Ho-Oh was intended to be the final recipient of that slot, which ties in to the bunch of Pokemon indexed around it: Lapras (Genbu, cross between a snake and a turtle, as seen by Lapras’ serpentine body with turtle like flippers, tail, and shell); Arcanine (Byakko, tiger, evident by the stripes on Arcanine’s body and general resemblance to a tiger); and Gyarados (Seiryuu, dragon, which could explain the change from Gyarados originally resembling a worm monster into one with draconic origins). By extension, Gyarados’ design origin could also explain why it received the Dark Type after Mega Evolution.

    O from the manga poll, later revealed to be the robotic monster known as Omega, could have been a temporary holder, just like how Lapras, Arcanine, and Gyarados didn’t resemble their current forms either. O has a phonetic resemblance to “Ho-Oh”. I further theorize that Ho-Oh was later deliberately postponed later in development after the legendary birds were conceptualized, which would explain why it was publicized in the anime early in Generation I’s cycle, as well as a possible translation of HoOh was not meant for release in Red and Green during Zog’s impromptu interview w/ Morimoto. This then created the empty slot that Mew would later reside in.

    On a side note, I feel that Omega was meant to be Blastoise’s evolution, considering that we see a possible evolution for Wartotle later on, and both Omega and Blastoise sharing the same posture and body type.

    1. From what we know it is indeed possible those mno. 31 and 32 were both some proto-nidos and when Gyaoon was moved, it just overwrote one of them and the other slot was simply left unused. This of course would kinda justify two of the missing slots before 37 in the Kaiju Zukan (assuming they also skipped 21) . However it would also raise the question, why not restore mno. 32 and add just the other one in p5 when the mid-nidos where supposedly re-implemented?

      About the “missing fire bird” that formed a trio along with Arcanine and Gyarados (or possibly a quartet with Caravaggio/Lapras), as much as I liked that theory Omega’s reveal and gyarados’ original concept, kinda made me lose all hopes aahah. Cheers! Thanks for following us!

    2. I’m a bit late to the party at this point, but Morimoto saying that Ho-oh wasn’t made for Gen 1 matches up pretty well with what we know of the GS development timeline. They started development in Spring 1996, and originally aimed for a 1997 release. By the time the anime started in Spring 1997, Ho-oh would have likely been finalized; It was one of the few Pokemon formally announced that year, with a name and design identical to the final release as well. Realistically, Ho-oh showed up in the anime to market Gen 2. It was supposed to get kids talking about that mysterious Pokemon that they saw on TV, the one that they could find if they brought the new games.

      … But, for countries that got Gen 1 at the same time as the anime, and had no access to Corocoro, Ho-oh was a mysterious Pokemon that had no explanation. Why would a Pokemon that debuted in 2000 have a cameo in 1998, before the games even came out? The “cut from Gen 1” theory makes sense in that context! It just falls apart when looking at the bigger picture. We have confirmation that Ho-oh was not, and was never a Gen 1 Pokemon, and the evidence continues to back that up.

  63. The back sprites for the kotora line look way to detailed for gen 1 standards, the pixels arent enlarged like other gen 1 backsprites

  64. How does 172 remotely resemble a potential Blastoise pre-evo? Looks more like a panda-type black/white tapir-like creature. I don’t see how the white part could be mistaken for shell/cannons.

    Also, are you saying that there’s only 181 Pokemon planned for Capsule Monsters? Seems to contradict with Masuda’s repeated claims that there were originally 190 (which implies that the ghost/fossil sprites replaced the Pokemon that once resided in 182-184 like how Mew erased Omega).

  65. Could Mr Mime have originally been in index number 115 or 121? It would fit right in with Period 4a giving evolutions to Period 2 Pokemon, as Mr Mime’s preevolution would have been index number 42.

  66. The Pokémon League was originally going to be in Saffron City? That explains why it’s the last city in the fly screen, with Indigo Plateau being the second to last! It may also explain why it’s initially walled off from the rest of Kanto until via the gates; perhaps you originally wouldn’t be able to enter until you had all badges. In the final game they kept it walled off, but changed the entrance requirement to merely having a drink in your inventory.

    Any idea what purpose would the Indigo Plateau have served? Perhaps it would have been home to the 8th gym?

  67. Would it be possible to crowfound a bribe to the owner of this ROM?

    I want so much to play this demo… please..
    So dissappointed with the let’s go pikachi/evee… I want to play the real Kanto… with pockets monsters, no pets.

  68. Why my previous comment didn’t make it though? 🙁

    Just asked if it’s possible by crowdfunding to convince the leaker to release the actual ROM…. we are to many wishing to play it.

  69. Why my previous comment didn’t make it through? 🙁

    Just asked if it’s possible by crowdfunding to convince the leaker to release the actual ROM…. we are to many wishing to play it

    1. Sorry, we manually approve most comments due to the regular influx of spam. Your comments have all been approved now.

        1. IDK what the laws concerning crowdfunding would be like, but it does feel like it would be of dubious legality (I’m pretty sure referring to it as a ‘bribe’ is right out), and I don’t think HC wants to take that potential risk.

          That being said, if it is in fact legal, I’d happily donate money… though if the deal fell through I’d want my money back.

          1. Scratch this, unfortunately; from what I can glean from the leaker going public, they apparently don’t even have the front sprites, or anything related to 115, 121 (tho I’m thinking it could be a mid-stage between 63 and 122, but that’s just a possibility) or 135, much less any sort of prototype.

            We could still petition and/or offer money (which would be crowdsourced) to GameFreak

  70. Im surprised that nobody has mentioned this, but 86 and 87 seem to me like they could be early designs of Makuhita and Hariyama. The two black bulges in 86 would just be Makuhita’s boxing gloves, and 87 has long hair similar Hariyama. 86 and 87 also seem to fit the sumo style with the way their hair looks.

  71. Very neat stuff! One of them looks like some kind of 3-evolution chain fa of a tiger-like Pokemon or something? Only the back sprites so I can’t tell if that was a thing from another generation.

    1. Technically they were in another generation; the first two (Kotora and Raitora) were in the Spaceworld Demo for Pokemon Gold. Beyond that , I dunno; the Shinx line in Gen 4 had the general theme of ‘electric big cat’, but they were lions so IDK if whoever developed them had the Kotora line on their mind when they were designed.

  72. I love all the beta stuff in pokémon!! I´d love to know the pokémon type of Yujirou`s gym, or the type of the betamons! ^^ I hope you´ll find something! thanks for your efforts! ^^

  73. I discovered Farfetch’d’s original name – it’s Okupān. I did a thorough search on a Katakana chart and the O symbol matched up perfectly. I also found that the objects on #86’s face could be boxing gloves.

  74. In my opinion, numer 80 looks like Gigan from godzilla series. Expect this sphere-like thing behind it? Is that maybe hand?

  75. Not only the design proposal of Pokemon, but also the design proposal of characters appearing are released. Animation at the beginning of the game, When you go to Pokemon s world! Let s go! Is like this.

  76. Hello! I am ending the game but I need the HM04 -force- in order to pass the last cave. I had been looking for it in Fuchsia City (the teeths of the Guardian, just like in the original versions of the first gen) but I did not find it. Please, help. Anyone knows where do I can get the HM04.
    Thank you

  77. missingno 61 looks a bit like coufant and copperajah from sword and shield. we all know that the copperajah and coufant entries state they came from another reigion. so that means there is another whole reigion to explore!

  78. The Pokemon Reverse Engineering Team (PRET) have managed to pull off a marvellous achievement, they have fully reverse engineered Pokemon Red and Blue back into Z80 assembly code

  79. Hello, HC,

    I want to let you know that I played this game complete and really loved it, I truly admire the work you did there, and with this whole site of course (I’ve read every article, some of them multiple times haha).

    I also went ahead and posted the game to archive dot org for preservation, along a fix I made for the Strenght HM problem. You can find it as “Pokemon Green Helix Chamber”.

    I want others to fully enjoy your hard work as much as I did. I guess you didn’t want it to be spread around, but I still want it to be preserved, so I’m just being forward about it. I don’t have a twitter so this seemed like good way to tell you.

    Thank you for all your amazing work. Have a great day!

  80. It just dawned on me that the flying/fighting fish line (79 and 80) might actually be the first Pokemon line made with the concept of evolution in mind! (At least when evolution was decided to be a proper mechanic.) Both designs are right next to each other in the index, their designs are similar, and there was even evolution data for them at one point in development. That would also mean the Ninetales line (with Mikon included) would have been the first three-stage evolution as well! It’s a shame that both fish got scrapped and Mikron was pulled out if they were responsible in developing the concept of evolution.

    That also makes me wonder where the fish actually do fit in in terms of periods. Your article has them in Period 2d when they were still Missingno.’s but they could very well be in Period 3a, when new Pokemon lines were made with evolution in mind. They even form an elemental trio with the Ninetales and Raichu lines! On the other hand, I can’t tell if they were Pokemon that Nishida herself designed and having only their back sprites doesn’t help either.

Leave a Reply to ScorpioTHK Cancel reply

Your email address will not be published. Required fields are marked *