Pokémon Power Creep: A Statistical Analysis

Alex_Dino

Haven't been here in a while, ever since the whole Salamence fiasco I kind of left here, but with the B/W metagame coming up, I'm really stoked for it.

This DOES contain some B/W information, so if this is in the wrong board or I need to put HIDE tags on the stuff on Gen V please let me know. My thoughts are that since it's not speculation, but rather statistics, this board would work.

Anyway, what this thread is about. In the last few days I was looking at the base stats of Gen V and, to me it seemed obvious that there is power creep going on, with all these base 120+ Attack and SpAttackers everywhere. So I decided to find out if this was true. Nothing more than taking averages and such, but still, statistics intrigue me, and so does Pokémon, so I wasn't going to let my question go unanswered.

What I found was pretty surprising. I don't have time to go into detail, as I have class in less than twenty minutes, but my basic question was answered: the Pokémon series IS having power creep, however, contrary to my thoughts, the biggest power creep to date was from Gen III to IV, not IV to V. In fact, Generation V, for the most part, counteracted the creep experienced by IV, but not back to pre-Gen IV levels.

Again, I don't have time to state everything, and in fact, I'm still not done, I still have to compute the averages for all fully-evolved Pokémon as separate generations, but when I get done with that I'll post an updated file. Here's the link to my work so far, it's a 2007 Excel spreadsheet, and I'll try to upload a pdf file as soon as I can:

http://www.2shared.com/file/cmx6Kapj/pokemonpowercreep.html

I don't know if these findings will help at all, or even if they're findings, if you already knew this information. I just wanted to contribute this in the chance this hasn't been looked at. Feel free to post your own findings.
 

Coldcut

This is really interesting. While browsing through the base stats, I definitely noticed that speeds especially seemed to be shifted toward the higher end, with many well above base 100. Just keep in mind that Pokemon is a multi-dimensional game, and it takes more than stats to make a good pokemon. It would also be interesting to do a similar analysis on the new moves, though again, base power is not the only consideration, and I have not found a reliable catalog of the new moves.

One little nitpicky thing: Would things change significantly if you used the Special stat from Gen I and counted it as both SpAtk and SpDef? This is not a huge change for most, but, for example, Alakazam would be granted an extra 50 base SpDef and Chansey gets an extra 70 base SpAtk. This raises their BST significantly and would change the averages for that generation by itself and the combined averages for all generations.
 

Alex_Dino

Yes, Pokémon is a multidimensional game, therefore these findings only hold true when everything else is held constant, which isn't the case, due to abilities, movepool, typing, and other things. I guess I could cross-reference this data with the current metagame and other things like movepool statistics, ability statistics, etc. to find relationships between them, and likewise get a somewhat more accurate view of the metagame from a mathematical/statistical standpoint.

But yes, these findings are based solely on stats and nothing else.
 

scorchedsky

Perhaps there's a way to link this to the average base power of "staple" moves such as Earthquake? Gen IV introduced 120 BP CC, 120 BP SE...maybe averaging the base power of attacking moves commonly used by OU pokemon from Gen I to Gen IV would reveal a similar phenomenon.
 

Coldcut

I finished the Fully Evolved, Separate data set and fixed a few inconsistencies (you counted Tyrogue four times in the Fully Evolved, Combined data, for example), but I'm not sure how to upload it. The trends are somewhat different from what the other data sets show. For example, Defenses dropped dramatically after their gains in Gen IV; Speed and HP continued to rise as before, SpAtk fell off a bit, and the largest change is that Atk rose sharply. BST also dropped a bit.
 

Alex_Dino

scorchedsky said:
Perhaps there's a way to link this to the average base power of "staple" moves such as Earthquake? Gen IV introduced 120 BP CC, 120 BP SE...maybe averaging the base power of attacking moves commonly used by OU pokemon from Gen I to Gen IV would reveal a similar phenomenon.
Click to expand...
Due to Explosion being in Gen I, and iirc it's still the strongest move (concerning base power) in the game. Due to this the range of base powers wouldn't change. Another thing is that a lot of moves such like Flamethrower, Ice Beam, Surf, Earthquake, Elemental punches, Outrage, etc., again, were around since Generation I, so there shouldn't be much difference since, for the most part, we still use those moves heavily. However an analysis of this, as well as all moves, would be interesting. I'll put both on my to-do list.

The thing is that, the more variables you add, the more specific you get, i.e. only listing OU Pokémon in the data, only using ___-type moves for the base power analysis, etc. This could lead to warped, or more prominently showing statistic readings, for example, you can see clearly that by just keeping all fully-evolved Pokémon the averages jump considerably due to higher base stats, and the differences stand out more due to a lower number of data sources. Another example would be doing only Fighting type moves in the base power analysis, as DynamicPunch and Close Combat's effects would show more than if it was all type moves.

Either way, it's good to add variables to the mix, which I do plan on doing, but it's good to start off with the basics to get a foundation of what you should expect.

Coldcut said:
I finished the Fully Evolved, Separate data set and fixed a few inconsistencies (you counted Tyrogue four times in the Fully Evolved, Combined data, for example), but I'm not sure how to upload it. The trends are somewhat different from what the other data sets show. For example, Defenses dropped dramatically after their gains in Gen IV; Speed and HP continued to rise as before, SpAtk fell off a bit, and the largest change is that Atk rose sharply. BST also dropped a bit.
Click to expand...
Thanks for the help as well as fixing some of my errors, I had some studying to do last night so I couldn't finish it.

As for uploading it I use http://www.2shared.com/ because it's simple and you don't need to sign up or anything. You just need to upload the file, and copy and paste the link wherever you need it.

Your findings are interesting. You could accurately assume (barring all other variables) that Gen V's metagame will be even more physical based than Gen IV, with higher attackers coming into play. Blissey usage would drop more due to this, and more physical walls could see more usage, such as Skarmory. Higher Speed would mean more need for revenge-killers with priority moves. BST usage is just signifying that there will be more mon's that are tailored to do specific things, as well as more "base stat efficiency," as in, utilizing stats in the best way possible. In other words, Gen V sounds like it is slightly more tailored to competitive players. :D Again this is just an assumption though.

Also, to answer your question about changing Gen I's stat to reflect the old Special stat, I could do that, but it would take another table listing those stats... or I could just add them together and omit the Sp. Def stat from that generation, that's all it was, correct? I never payed attention to their stats in Gen I, since I was too young to realize their implications.

The best part about these findings is that, again, barring all other variables, and assuming the game mechanics stay the same, you could approximate what Gen VI's averages could be based on previous data.
 

Coldcut

Here is the link to my updated spreadsheet:

http://www.2shared.com/file/Cg3nD7t3/Power_Creep.html

Interestingly, every stat has increased from Gen I to Gen V except SpDef. Also, the Fully Evolved, Combined stats are probably inaccurate and less useful for two reasons. One, there are too many individual cells for Excel to handle at once. Two, Many pokemon that were initially the final evolution got additional evolutions in later generations (Electabuzz, Piloswine, Rhydon, etc.), so what is considered Fully Evolved changes from generation to generation. Unfortunately, this is also the most relevant for competitive battling.

I added a separate data set for offensive stats (Atk, SpAtk, Spd) and defensive stats (HP, Def, SpDef) and graphed them. It's interesting to see how the average stat changes affect the metagame for that generation. In Gen II, when all offensive stats dropped and all defensive stats rose, stall ruled the metagame, and the stats predict this without knowing anything about the addition of the Steel type. Based on this analysis, Gen V is going to heavily favor offense over defense.

Edit: There is a subforum for B/W data, so might this be better off there even though this is data analysis of all generations?
 

Alex_Dino

Coldcut said:
Here is the link to my updated spreadsheet:

http://www.2shared.com/file/Cg3nD7t3/Power_Creep.html

Interestingly, every stat has increased from Gen I to Gen V except SpDef. Also, the Fully Evolved, Combined stats are probably inaccurate and less useful for two reasons. One, there are too many individual cells for Excel to handle at once. Two, Many pokemon that were initially the final evolution got additional evolutions in later generations (Electabuzz, Piloswine, Rhydon, etc.), so what is considered Fully Evolved changes from generation to generation. Unfortunately, this is also the most relevant for competitive battling.

I added a separate data set for offensive stats (Atk, SpAtk, Spd) and defensive stats (HP, Def, SpDef) and graphed them. It's interesting to see how the average stat changes affect the metagame for that generation. In Gen II, when all offensive stats dropped and all defensive stats rose, stall ruled the metagame, and the stats predict this without knowing anything about the addition of the Steel type. Based on this analysis, Gen V is going to heavily favor offense over defense.

Edit: There is a subforum for B/W data, so might this be better off there even though this is data analysis of all generations?
Click to expand...
That offense/defense graph is very interesting. And you're right, the stats reflect why Gen II was mostly stall-oriented.

The Fully Evolved, Combined graph is probably inaccurate yes. I guess we could count the Pokémon considered previously as fully evolved, but that could possibly make it more inaccurate. I don't know if there is a simple solution to that.

Consequently, I'm going to work on an OU table soon, so that should help.

Edit: As for moving the thread, I'll make a thread there too. My original idea was that since this mostly an analysis on all generations it would be best to put it in this board. I mean, speculating about B/W's possible implications on the current metagame is great and all, but there are other questions that I would like to answer with this (and other analysis), such as: Why was Gen IV was so heavily offense oriented when the average defense shot up past Gen II levels? In theory this should of had made Gen IV just slightly offense oriented, but the fact that Hyper Offensive teams can do well in the Gen IV metagame kind of shows how offensively oriented it is.
 
You must log in or register to reply here.

Users Who Are Viewing This Thread (Users: 1, Guests: 0)

  • ... and 1 more.
    • Top