Would banning Garchomp from standard play even solve the problem?

[Cathy]

Before we can decide whether Garchomp is overcentralising the metagame, we need to establish whether the metagame is in fact overcentralised. There are approximately fifty OU pokemon on both the current Smogon list and the list of just February (that would include Deoxys-S and Wobbuffet--which did not cause more centralisation). Fifty seems like a pretty big number to me. If the metagame is not currently overcentralised then no pokemon can be causing overcentralisation, and the focus should be unbanning pokemon without reducing the number of viable pokemon (i.e. simplifying the rules), rather than banning pokemon.

But it comes down to whether fifty really is all that big. Compared to previous generations, I am told that it is.

 

[IggyBot]

We also need to look at the ratio of OU compared to the number of available pokemon when comparing previous generations. DP has 493 pokemon available to use, which is more than any other generation, so I would assume that all tiers grow in size.

 

[Cathy]

I disagree. The number of total (fully evolved) pokemon does not matter at all. Only the actual number of viable pokemon does.

Let's suppose we are comparing two metagames: DP and DP+. DP+ is the same as DP except it has two hundred new new piece of shit fully evolved pokemon that are so bad that no one is ever going to use them. DP+ has a lower ratio of viable pokemon than DP, but I would say they are both equally decentralised, since the newly added piece of shit pokemon don't actually change anything.

 

[Mekkah]

It does indeed matter for tiers like OU that use percentages, depending on what algorithm you use. If you use TopX, then of course it won't matter. For the rest, it just means NU grows a ton, and perhaps some things in NU would have to go to UU as they become relatively overpowered, but we haven't really explored that yet.

Anyway, I think we should stop babbling and get a move on. So if Colin could announce that Garchomp is going to get banned from the Ladder soon for a testing period and then a week later ban it, that would be great. The announcement would give people time to adapt their teams and play style a bit, plus they won't get shocked all of a sudden when the ban does happen.

 

[Cathy]

The OU tier is being defined in terms of percent of cumulative usage frequency, not percent of pokemon. DP and DP+ (see above) would have the same OU tier. When measuring centralisation, it is the number of number of viable pokemon that matters, not the percent of (fully evolved) pokemon.

I am not interested in banning Garchomp on the ladder right now. So far few people in this topic have even explicitly stated that they believe the current game is too centralised (which perhaps should have been the first thing discussed). Since we have not even agreed on this, there is obviously nothing close to a consensus on whether Garchomp should be banned. Even most (if not all) of the people who supported banning him have not stated that they believe the current game is in fact too centralised. This could be implicit, or it could just be that we have not give it enough thought. We need to decide what the target number of viable pokemon is, or else we will have no way of knowing whether the game is overcentralised and whether banning pokemon is in order. And the ladder is not just a place for testing whims. I advocated a different method for the bulk of the testing in the "What is Uber?" topic.

Note that with unbanning pokemon, the goal is different. The goal there is to simplify the rules without decreasing the number of viable pokemon. This is something that does not raise any unanswered questions.

Nevertheless we could try banning Garchomp as part of the programme I outlined in the "What is Uber?" topic, but we'd still need some idea of a target. Fifty viable pokemon seems like quite a few, and I would support fifty as the benchmark of "sufficiently decentralised" for DP.

 

[X-Act]

Note that when we say 'there are x Pokemon in OU', what is meant is the following:

If a Pokemon is used in a battle for the first time, then it is one of the x OU Pokemon 3/4 of the time.

That's the definition of OU that I settled on, and that presumably was agreed upon (I'll assume no response to proposal means acceptance of proposal at this point).

So yeah, if x is small then the metagame is indeed overcentralised. However, x was always greater than 45 for DP for every month (it seems to fluctuate between 46 and 49). I think that is a good number, although we cannot compare with the other metagames since we never had such detailed usage lists for the other metagames.

 

[Hipmonlee]

Over-centralisation is relative. If removing Garchomp causes the game to be less centralised, then the current ruleset is over-centralising. If not then perhaps it isnt.

Have a nice day.

 

[Cathy]

I disagree that overcentralisation should be treated relatively. For one thing, there are questions whether anybody would want to play that game (in that "The definition of Uber" topic). There is also the question of whether the game will actually reach a point where banning any of the remaining pokemon will not increase the number of viable pokemon. There is also the concern that we cannot detect whether a pokemon should be unbanned. If the ban list contains every pokemon except for Pidgey, Magikarp, and Kakuna, then there is only one viable pokemon. If we ban Pidgey, there are now two viable pokemon. Obviously the right course of action should be to unban some pokemon instead, but how can we tell with a goal of making the game as decentralised as possible? It almost seems as if we would only stop when the ban list exceeds the number of legal pokemon.

Also, the idea of treating overcentralising relatively is equivalent to minimising the number of pokemon who are not viable. But why should we care how many pokemon are not viable? If GameFreak adds five hundred new shit pokemon I don't see why any effort should be made to accommodate them, even though the metagame could be be greatly decentralised by banning most of our current pokemon.

So instead I suggest constructing the uber list so that it contains as few pokemon as possible such that the number of viable pokemon is above some critical threshold. Then we would define "overcentralisation" as if the number of viable pokemon fell below this threshold, and we would say a pokemon is "overcentralising" if banning it would increase the number of viable pokemon above the threshold. Then it would be fair to say that if the metagame is not overcentralised then no pokemon can be overcentralising it.

It does not seem to me that there is a tradition of attempting to make the metagame as decentralised as possible either, which is why I did not address the notion above. I don't think we should be banning pokemon if we do not even agree on the purpose of making such a ban. Maybe there is no disagreement, but there have not been that many posts in "The definition of Uber" compared to this topic, even though here we essentially need the results of the other topic.

 

[Hipmonlee]

Well, I disagree with you entirely.

Obviously we cant test every possible metagame, but no matter how decentralised a metagame is, if we have reason to believe that adjusting the rules in some way will result in an even more decentralised metagame, then I dont see why we wouldnt try it.

And really, now that I think about it, surely the fact that this forum exists, and the fact that I have been arguing for 2 pages about how broken Garchomp is sorta proves that I feel the metagame isnt decentralised enough in its current form.

Have a nice day.

 

[Cathy]

Has anybody stated that he does not believe that banning Garchomp would decentralise the game? I have no doubt that it would decentralise the game. In fact many people (e.g. X-act) in this topic have stated that they don't dispute this fact, but still dispute banning it. So obviously there is not an agreement on the purpose of such a ban. The fact that Garchomp centralises the game is hardly in dispute, yet there are several pages with regard to whether it should be banned. If anything the existence of this lengthy discussion shows that it is completely not evident where the various people stand on the purpose of a ban. Rather than try to read things into these posts that aren't there, I am asking that we explicitly consider the issues that underscore these debates.

In any case my point here has just been that the issues in "The definition of uber" need to be decided before there can be any action taken. If we don't agree on what "uber" means then we are all using the same word but there is no common understanding--that would not be much of a discussion.

Ultimately the question of whether anything needs to be done to make the game less centralised probably will have to be decided by a poll, because I don't see any compelling arguments on either side, and it is really just a roadblock on the way to discussions about whether anything should be banned.

 

[Hipmonlee]

Honestly, I dont see any issues coming from the definition of uber thread. Decentralisation is our goal, it always has been our goal. I guess the only issue I could suggest is the definition of decentralisation. Which actually is quite a big issue, and probably warrants a topic of its own.

Also I cant see a poll ever happening. I mean, there really is no way for it to be conducted fairly. Nor do I trust the majority of pokemon battlers to even understand what is being asked of them.

Have a nice day.