You are not logged in.
As I am here, its pretty easy to guess I program. I also program in part as my day job; I deal with a lot of web scripting, dealing with data, market prices, etc. by the thousands.
I also write D&D modules as a side gig and DM a campaign; its been a bit, but game design work all the same. Given both, I also hate tedious details, NPC generation, world detail generation, etc. I like to explore myself as well, and writing all of it...takes some of that away. So I code a lot of tools to help with my world building.
I just rolled out my first release of my pokered project with the basic engine but very light on details for new monsters. Fiddly bits, like the world building details. SO, I'm writing some tools to make that easier.
First part; data collection and parsing. I used Smogon's site to pull data and classifications to build a "standard array" for several groupings. D&D uses a Standard Array for character building, which can be modified further by choices at generation but it ensures the play experience is more balanced. Goal is to do the same here.
Smogon's classifications are:
"Little Cup", mostly 1st stage/Basic Pokemon of two and three stage evolutions and even then mostly three-stage evolutions. Comprises roughly 35-38% of the base 151 Pokemon pool.
"Under Used", are mostly 1st Stage/Basic Pokemon of two stage evolutions, 2nd stage of 3 stage Pokemon and mid-tier evolution-less Pokemon. Comprises roughly 44-47% of the base 151 Pokemon pool.
"Over Used" are more heavily used "good" competitive pokemon; 2nd stage of 2 stages, 3rd stage of 3 stages, and strong evolutionless pokemon, Legendaries, etc. Comprises roughly 9-12% of the base 151 Pokemon pool.
"Uber" are the unstopables; in RBY that's just Mew & Mewtwo, essentially 1% of the base 151 Pokemon Pool.
Playing with the spreadsheet, eliminating some outliers, and tinkering with the data, I came up with the following arrays of stat values. The basic calculation followed the D&D idea, which worked well here as well: 1 poor value, 1 average value, 2-3 values increasing from the standard deviation (STD), and 1 high value. I have some concerns about the Over Used array right now and will play with it more, but its a solid base.
The idea here is to create a baseline of generic valyes, and then you/I can place the values as appropriate and then tweak them based on the roles, skills, and character traits of the intended Pokemon. If you look at the spreadsheet, for example, you'll see some of the ignored stats belong to pokemon like Magikarp, who is deliberately lower and weaker. Including these stats would throw the overall averages, but it makes a fine character choice. It makes up for it on the other end by Gyarados having a much higher speed for its class than the average should really allow. Likewise, Chancey has a SILLY high HP score and abysmal Atk and Def for its class; most of Chancey's scores were considered outliers for the purposes of that classification's averages and data.
SO, here are my results; it goes Class, Low, Average, Average + STD, Average ++ STD, and the High.
Little Cup: 20, 50, 60, 70, 92 (STD 10)
Under Used: 32, 76, 87, 98, 118 (STD 11)
Over Used(a): 46, 87, 104, 121, 130 (STD 17; that's the concern)
Over Used(b): 46, 87, 99, 111, 130 (STD 12; experimental tweak)
Uber having two instances, Mew & Mewtwo, is sort of pointless to do a similar break down on. Roughly 100 across the board, with a low of 90 and a high of 154 on Mewtwo.
Next step is using this and some tweaks to allow variance in the actual stats themselves in a webscript to generate files. Ideally we can set some details and have it order the stats with a preference using things like the type and role to generate the order (IE psychic higher/est scores in Special, ground/rock/ in Def/HP, fighting in Atk/Speed, etc)
Until then, here's my spreadsheet. Tweak and play on your own if you disagree with my "outliers" or want to resort the classifications; I just riffed off Smogon's because they do the competitive play and had them broken out in a few tiers.
http://projectneo.retrogaminggear.com/f … t_avgs.xls
Pokemon: Project Neo A Pokemon hack 15 years in the making...