A History of Pokémon through the Internal List – 1

The Generation 1 Pokémon Internal ID list might look random, but it can be broken down into 5 different “Periods” that we are now going to analyze along with first-hand sources and vestiges in the game data, in order to produce a guide to identify some of the 40 MissingNo.’s key characteristics. During our journey, we will also learn about the reasoning behind the design of the first Pokémon games themselves, and possibly discover connections that aren’t obvious in the final release, as many changes happened throughout the turbulent 5-year long development period.  Let’s start by analyzing the first Period of Pokémon development!

INTRO P1 P2 P3 P4 P5


Period 1 (circa 1989-1991) consists of Pokémon from ID 001 (Rhydon) to ID 037 (Slowpoke), and it is inadvertently one of the least cohesive pattern-wise. Since this period is not the best representation of the inner workings of the ID list in general when compared to the following ones, our method is not fully applicable here. The Pokémon from this period are diverse as well as unorganized; moreover, the ones in the range 017-037 seem to have been heavily reworked and moved around during an unspecified point in time when they had the opportunity to reorganize things properly – this happened much earlier than the hurried 150 cut. As a result, many original P1 monsters with low scores on the first polls of 1992 could have disappeared without a trace.

This batch of monsters was probably designed between 1989-91, between the time the early Pokémon pitch was first drafted and pitched to Nintendo with the name Capsule Monsters, and its initially scheduled release in December 1991, before Shigeki Morimoto and Motofumi Fujiwara even joined the design team.

Capsule Monsters was pitched along with an illustrated booklet by the same name – yet to be revealed in its entirety – detailing the early plans and art for the first Pokémon games. Dated 1990 (both on the cover and on the Kanto layout map), this document is, however, most likely a collection of concepts that Tajiri and Sugimori had been developing for quite some time, as some inconsistencies in the contents might prove (namely plans and designs coexisting at visibly different stages of development).

Game Freak wasn’t able to meet the December 1991 deadline, mainly because they were busy developing Jerry Boy (Smart Ball) and Yoshi around the same time.

Tajiri: Right after Quinty, I thought of something similar to Pokémon. So Shigesato Itoi made something called APE with Nintendo, a collaborative company in Tokyo, And he went: “Anyone who wants to make a game with Nintendo, here’s your chance.” So I took the idea of Pokémon to Mr. Itoi and they were going to make it but they said it would take some time to complete, so they went: “Mr Yokoi of Nintendo plans on making a game about Yoshi” (…), they decided on a due date and I did the best I could in half a year. I wanted to make the best product I could despite the short amount of time.

So we might assume these 37 Pokémon were, in fact, all they had done by the end of 1991 despite Tajiri’s wish to have “more than  200 monsters” right from the start as he writes in the pitch. (read the translation by GlitterBerri)

This makes Period 1 one of the most enigmatic and intriguing, and the reason is simple: Satoshi Tajiri and Ken Sugimori instinctively put their passions and experiences into Capsule Monsters with complete freedom, taking inspiration from their favorite TV shows and video games like Ultraseven, Final Fantasy Legend (SaGa), Earthbound, and Dragon Quest.

Tajiri: Everything I did as a kid is kind of rolled into one–that’s what Pokémon is. Playing video games, watching TV, Ultraman with his capsule monsters–they all became ingredients for the game.

This makes it quite tricky to split further, but we seemed to find two distinct patterns nonetheless:

Period 1a (ID 001-030, Rhydon – Tangela) specifically refers to the earliest stage of the games’ development, quite well-sourced thanks to various media and principally the aforementioned Capsule Monsters booklet.

Period 1b (033-037, Growlithe – Slowpoke) is an outlier that we will discuss later on.

Missingno. 031 and 032  could belong to either of these patterns; some evidence suggests that they were a part of P1b, although it’s impossible to tell with the sources that we currently have.

PERIOD 1a (001-030)

ID 001-016 (Rhydon-Nidoqueen) forms the core of this period. We can see them all printed on the early sprite sheet revealed by the Satoshi Tajiri manga (along with the scrapped “No.1” proto-Gyaaon). This sheet comes directly from Capsule Monsters, and it’s some of the most compelling evidence in favor of the chronological Internal ID list index method. It’s safe to assume that this was the earliest Capsule Monsters roster and some kind of confirmation is also found in a 2007 video interview, though Tajiri may have exaggerated a bit.

Tajiri: It was really hard to come up with things that people would want enough to trade for them. At first, we had only made about ten, so you would get bored right away. So then it was a question of how many Pokémon you’d need for someone to be able to play without getting bored, and it felt like there was an endless road in front of us. We spent five long years making those Pokemon you’d want to trade for, until there was a whole mountain of them, and that became Pokemon as we know it now. The whole process really took a long time.

These early Pokémon, or rather “Capumon”, were heavily inspired by Tajiri and Sugimori’s childhood fantasies, enthusiastically translated into their project. The title and basic concept itself was inspired by the capsules used to summon the Kaiju in Ultraseven, a detail that ultimately led to the trademark controversy that compelled Game Freak to change the title of the game.

It’s easy to spot the parallels between our bulky Capumon and the Ultraseven Kaiju and we also get to meet the Capumon version of many traditional RPG mimics and monsters as Voltorb, Grimer, Tangela, and Exeggcute, along with some cute original intuitions like Kaiju-mom or having a baby rabbit-kaiju evolving into the menacing bunny eared Nido-royals that we all know and love today.

Sugimori: It first started with designs of dinosaurs and monsters in a super-deformed style. they had fangs and long tails, and spikes jutting out all over their bodies. It started with Pokémon like Rhydon and Nidoking, for example. And when we looked at them all together, the designs felt rather uniform (…) as the games’ development progressed we added the idea of “types” to the game system.

Surprisingly, typing wasn’t a thing yet:

Sugimori: The idea came to us during development. Battles would get monotonous if there were only strong and weak Pokémon, so by affixing types to the Pokémon we were able to give the fighting more depth.  

The absence of typing is reflected by the fact that, even in the final, many P1 monsters have a versatile TM movepools. One surprising example is Lickitung, who can learn 35 of 55 machine moves in the game; Clefairy, Nidoking and Nidoqueen also benefit immensely from vast TM movepools.

However, the Capumon pitch written by Tajiri gives us some insight about an early sorting in “categories”. Worth noting are the “Mirage monsters” (幻のモンスター) and the mysterious  “Power Kings” (パワーキング).

I asked if he wouldn’t mind adding 5 Powerkings, monsters with the strength you might expect to see in soldiers, and we finally struck a bargain. Powerkings have a high encounter rate, and you’ll quickly run into them if you wander around dungeons. They’re strong, though, and the theory is that, if you have a lot of them, you can put them to work as soldiers when you run into other monsters.

More interviews reveal that Capumon were designed with specific roles in the human society, like “pet” or “transportation”. Those early categories, very close in style to classic RPG classes, might give us some context to understand the original purpose and conception of the first 37 Pokémon like Lapras and Clefairy.

Sugimori: The first Pokémon were Rhydon, Clefairy, and Lapras. At first, we’d planned to have Pokémon living alongside humans, making their lives easier. So, during the early stages of development, many of the characters we came up with had clear roles, like carrying things around or sailing across the sea with people on their backs. We also thought up cute pet-like characters like Clefairy.

Evolution, another defining Pokémon characteristic, seems to be a mechanic initially implemented as just a test. This concept would be thoroughly exploited only a couple of years later in P3. We can’t possibly tell for sure what the original relations between Gastly and Gengar, Exeggcute and Exeggutor or Rhyhorn and Rhydon were; in this phase, we can confirm evolution just for the Nidoran family as their connection can be easily inferred from their looks.

The other obvious evolutions (Growlithe and Slowpoke) appear among the last additions to the period, and we’ll talk about them when discussing P1b.

While types and evolutions were absent or just roughly outlined, the creators seem to have untimely fantasized about some concepts that wouldn’t be fully realized until much later – specifically gender (evidenced by the Nidoran), and perhaps even breeding as hinted at by the fetus and eggs visible in Capsule Monsters. 



The second half of P1a (ID 017-030, Cubone – Tangela.) is suspiciously less sourced outside the in-game data, leading us to some unusual speculations.

Early sketches of Lapras, Gastly, Staryu, and Blastoise appeared in Capsule Monsters, and Sugimori himself stated that Lapras was among the first Pokémon ever designed along with Clefairy and Rhydon (curiously enough, Lapras wasn’t granted a sprite among the first 17 though). Some other P1a Pokémon are visible or referenced to in the two poll documents available (the first October 1992 poll and the Tajiri manga poll) – specifically Cubone, Scyther, Pinsir, and Tangela. Also, Slowpoke and Rhyhorn’s name might be partially visible in the list from the GameInformer video, although not dating from the Capumon period specifically as its title “PM” suggests (probably 1992 – 1993).

This leaves us with a bunch of P1a Pokémon yet to be documented in any form as actually pertaining to this first period: Arcanine, Gyarados, Shellder, and Tentacool, as well as pretty much the entirety of the Period 1b Pokémon. We might argue that Shellder is somehow present on Slowbro’s tail; furthermore, Shellder’s final sprite is among the downsized ones, hinting at an established presence in the games before the existence of Cloyster.

The three allegedly re-designed P1 sprites.
It’s also a bit suspicious that Growlithe or Arcanine weren’t mentioned among the favourites in the 1992 polls; their final sprites look surely newer than P1, and Arcanine’s and Gyarados’ (along with Lapras) look pretty much made by the same author at the same time (probably Sugimori).
This is not final evidence that they weren’t included in the original 37 Capumon, because we know for sure that Lapras changed appearance (hence the sprite) between Capsule Monsters and the final game. The same could have happened with Arcanine and Gyarados, meaning that they could have been present from the beginning with a different design. 
Talking of Gyarados, Sugimori claimed that at some point the rival was supposed to be a Gyarados trainer and his name was supposed to be Gyarao. This statement is as mysterious as it gets and can’t prove Gyarados was originally among the first 37 because it’s lacking a timeframe.
Another defining characteristic of a P1 Pokemon is that the vast majority of them have default base cries (pitch and length values of 0 and 128 respectively, we discuss that topic further in our article about cries), however, a bunch of base cries are not assigned to any Pokémon in the final build hinting towards lost P1 owners. How many Capumon are missing? Why are there 38 base cries and only 37 P1 Pokémon? Sadly we don’t have enough documents to answer these questions, but we will sum up our musings in the Missingno Summary and the Lost Pokémon dossier at the end of the series. 


In this Capumon mockup, you can clearly see three “capsules”, is a Voltorb among them?

Typical RPG monsters, Grimer and Gastly are two reiterations of very common RPG monsters, namely the slime monster and the gas cloud/ghost, along with the “Green Dragon” only mentioned in the pitch.  Mimics and other typical RPG monsters are also present among the 37 Capumon, like Voltorb, probably added as soon as the early Pokéballs (“Capsules”) were implemented in the gameplay as an item-holding icon, in order to mimic them. Even Tangela, Exeggcute and Exeggutor might have been envisioned as mimics in some peculiar locations. Gengar embodies the concept of mimic quite literally, seeing as its name and design is inspired by doppelgangers.

Kaiju and dinosaur-like monsters, “Gyaasu”, Nidoking, Nidoqueen, Kangaskhan, “Ivysaur”, Rhydon, Rhyhorn, Staryu, and Blastoise are all descendants of very important Kaiju in the Godzilla and Ultraman franchise.

The common fauna monsters, Cubone might’ve inhabited the Desert location, Gastly’s habitat would have probably been the cemetery. Spearow is a common bird, while Tentacool and Shellder were supposed to inhabit the sea, rivers and lakes of proto-Kanto.

This proto-Kanto landscape was revealed in the preface to the 2018 manga about Satoshi Tajiri’s life. You can clearly see many different habitats that didn’t make it to the final (like the desert, lakes, and rivers) as well as weirdly familiar locations.

Giant bug pokemon, Pinsir and Scyther. Though there was a strong bug collecting inspiration, P1 bugs are weirdly scarce, one would think the classical Japanese rivalry between Stag and Heracles beetle would’ve resulted in two Capumon, but there’s no evidence that this was the case. Bug catching might’ve just served as a base inspiration: “What if you could catch Kaiju like bugs?”.

The helpful monsters, Lapras and Clefairy, were often described by the creators as monsters with a specific role in human society. There’s also a Rhyhorn-looking monster depicted carrying some barrels on its back around a town.

“Rival” Pokémon you’re not supposed to catch, “Rival monsters”, maybe “evil” counterparts, to defeat/tame. Possibly Gengar, Gastly and Grimer? Could this be the earliest origin of Dark-type Pokémon? You were not supposed to collect every monster, the “Catch ’em all” logic was a later evolution of the concept.

Power Kings パワーキング – described as “strong as soldiers” by Tajiri, they’re the most mysterious early category, we can’t possibly list any Pokémon.

The folklore inspired monsters, Slowbro, Lickitung, Exeggutor, Gyarados and Arcanine.

The “Mirage Monsters” 幻のモンスター(Maboroshi no monsutā) sound like a predecessor to the concept of Mythical and Legendaries as Mythical Pokémon are still called “Maboroshi no pokemon幻のポケモン in Japanese. Assuming Arcanine and Gyarados were actually coded in P1, the two share some similarities and possibly a common inspiration, namely, they’re both based off legendary Chinese creatures. Arcanine is even infamously referred to as “Legendary” (でんせつポケモン) in the Pokedex, and it was placed among the legendary birds in the anime and other Pokémon related media. Gyarados is less obvious, but its high Base Stat Total and lore make it into another candidate for an early legendary Pokemon. Perhaps these two were designed with this “category” in mind.  This identification of Arcanine and Gyarados as early legendaries is pretty speculative, mostly because P1 is the least organized period of all, and doesn’t always follow a pattern structure. However, it might be a hint as to the identity of a MissingNo. that would appear between them had Mew not been added at the last moment.


Notably, on the manga chart, P1a Pokemon are the only ones missing their index numbers (Kangaskhan is even totally misplaced), instead, they are marked with circled numbers that match up with their decimal Cry ID + 1. We explored this topic in our study of early development documents.
As we pointed out in that article, there are a couple of puzzling exceptions in the poll chart indexing: Blastoise and Tangela’s current cries are 18 (0x12) and 19 (0x13), which would suggest they should be labeled 19 and 20, instead they are labeled 35 and 37 respectively. This could hint at a temporary re-indexing of the early portion of the ID list.

Perhaps Blastoise and Tangela (along with Pinsir who’s most commonly grouped together with them) were actually placed in slots 035 to 037 in the index list at some point, but they were moved back after a couple of earlier monsters had been deleted.

Our hypothesis is that, in the final game, the resulting empty slots might’ve been partially re-filled with Pokémon like Growlithe, Fearow, Pidgey, and Slowpoke.

PERIOD 1b (033-037)

That brings us to Period 1b and the consideration that some of these 5 Pokémon indeed feel like a later addition. If anything Onix and Slowpoke are suspected to belong to the earliest P1 revision, Onix because its sprite style looks similar in style to Blastoise’s and Pinsir’s, already finalized by P2, and Slowpoke being tied with Shellder and Slowbro concept-wise and listed in the GameInformer list with its final Cry ID position – that would make it one of very few Pokemon in Period 1 with a clear evolutionary relation, along with the Nidoran lines. 

Another clue that might prove most of them took the place of pre-existing Capumon is that unlike the vast majority of Pokémon in the ID range 001-037 that have a unique base cry with default parameters (00, 128), these Pokémon have different values or a shared base among the first 37. By cross-referencing the sources we can separate the exceptions (i.e. Voltorb and Gengar which probably had modified base cries from the beginning) from the actually dubious ones, in fact, turns out Onix, Growlithe, Pidgey, and Fearow are nowhere to be seen in the documents dating 1989 – 1993. This later re-fills might’ve happened as late as P4 as Growlithe could suggest.

Pidgey’s base cry wasn’t even originally assigned to it, but was originally used by Clefairy before switching it to a sweeter sounding base cry, that – we now know – was originally designed for Cubone. This used to be one of the two instances of a shared base cry in P1a, the other one being Staryu and Mew, another clue that Mew, if even present among the final 190, wasn’t coded at 021.


Pattern 1a is the most diverse and includes all sorts of different monsters, however, the various inspirations could be grouped as follows and the 3 MissingNo. in this period should’ve been designed along these lines:

  1. Pokémon designed by Sugimori and Tajiri As the early Capumon documents were illustrated just by Ken Sugimori, should we assume Period 1 MissingNo. were also designed by Sugimori with Tajiri’s directions? The sprites indeed show some early Sugimori traits, such as big eyes and a Toriyama-ish style.
  2. No clear typing P1 Pokémon were not designed with a specific typing in mind. The typing mechanics were, in fact, added later.
  3. Mostly one stagers Most of the Pokemon are single-stage at this point and there are no three-stage lines. Pokemon currently in evolutionary lines were not necessarily meant to have one at this point.
  4. Kaiju and dinosaur-like monsters “Gyaoon”, Nidoking, Nidoqueen, Kangaskhan, “Ivysaur”, Rhydon, Rhyhorn, Staryu, Blastoise.
  5. Typical RPG monsters Mimics and other typical RPG monsters like Voltorb, Tangela, Exeggcute, Grimer, Gastly.
  6. Helpful and pet monsters  Lapras, Clefairy, Rhydon.
  7. Folklore inspired monsters Arcanine, Gyarados, Exeggutor, Likitung, Slowbro.
  8. Common fauna Pokémon created along with the proto-Kanto map, designed to populate the different environments like Cubone, Gastly, Spearow, Tentacool, Shellder, Pinsir, Scyther.
  9. Mirage Monsters Legendary or Mythical Pokémon were a thing even in the concept stage, perhaps not meant to be caught or even seen. Possibly Arcanine and Gyarados?
  10. “Rival” Pokémon you’re not supposed to catch maybe “evil” counterparts. Possibly Gengar and Grimer? The “Catch’em all” slogan was a later invention.
  11. Monsters with a base cry Most of the Pokémon in this period have cries with pitch and length parameters set at 0 and 128 respectively, which makes us think that these were the default values – and in turn, that each base cry was designed for a specific monster. Most of these monsters appear in Period 1, suggesting that cries were, for the most part, programmed during that period.
  12. Monsters with a versatile TM movepool Many of the Pokémon, especially within the first half of P1a, have access to a vast number of machine moves, overshadowing many Pokémon added later in that regard.


Some Capumon and placeholder monsters featured in the earlier documents don’t appear in the final list, like Dragon4 or the many “Rhino” monsters you’ve seen in these pages, find out more about them in our Lost Pokémon dossier.



Onix, a P1b Pokémon (so suspected to be a later addition) is one of five fully evolved Pokémon in the Generation I games that cannot learn Hyper Beam. The other four are: Hitmonchan and Hitmonlee, Farfetch’d and Ditto.
It would be strange for this oddity to be a result of deliberate design. Onix’s stats are very low as it is, contrasting with its appearance. Hyper Beam would be an awfully weak move with Onix’s Attack stat, but why take away that option?
What’s notable is that – apart from the Hitmons – all of the aforementioned Pokémon received evolved forms in the Gold and Silver prototypes. Animon and Madamu were never fully realized, but Steelix survived until the final release of the games.
Presence of these Pokémon in a game that directly followed Red, Green and Blue could suggest that they were supposed to appear in Generation I in some form – of course, it would be a stretch to say that Steelix was there as a Steel-type, but an Onix evolution can’t be ruled out to have occupied the index list at some point.




INTRO P1 P2 P3 P4 P5