The Metagame - Centralization, Overcentralization, Diversity, and related topics.

There's been a lot of talk about the overcentralization and proposal about OU lately, and with it I noticed there are a lot of terms that are being thrown around undefined. I will try to be as objective as I can - none of these arguments necessarily reflects my own.

The purpose of this post is to illustrate what a Pokemon metagame is and to explore the definitions of overcentralization and the effects thereof. This thread also explores what everyone should be considering before they suggest a giant change in tiers.

A metagame is a competitive environment centered around a group of Pokemon. We use the term "centered around" - mostly because if a metagame is not centered around something, it is a metagame in development, not a stable metagame. We see that there are three stable metagames right now, Uber, OU, and UU.

We will talk about the OU metagame since that is the metagame that is the most controversial. We see that OU is centered around the threat lists (Offensive and Defensive) - and that it contains roughly 50 Pokemon, something that has not changed in some time.

Now we consider the term "overcentralize". What does it entail for something to be overcentralized?

One potential definition that I have seen on the forum was that "needing at least two Pokemon to stop a Pokemon". Yet this is a meaningless definition - while we can intuitively see that this might be the case, we see that it means nothing simply because we can't measure anything out of it. For example, one might be using Bronzong and Swampert to try and stop Garchomp - yet Bronzong and Swampert have a lot more use than stopping Garchomp.

Suppose that one could argue that "every wall/tank people use in OU must be able to deal with Garchomp in someway" - but even still, would removing Garchomp really simply stop making people use Gliscor, Bronzong, Skarmory, etc? How would one argue that they're just not just "solid" walls and that it's Garchomp's fault?

One example of an overcentralization factor is the use of "Toxic Spikes" in order to stop Garchomp - but does Toxic Spikes have so much more utility than stop Substitute Garchomp? Would one use a "Garchomp Killing Spike" on their every team if it hampered Garchomp when it switches in someway and does nothing else? Where is the line of "overcentralization" in this case?

We see that the line between "normal centralization" and "overcentralization" is hard to draw - it's like drawing a line in a gradient between black and white and saying "this is white, this is black". Would Bronzong, Skarmory, and Forrestress carrying Shed Shell to avoid Magnezone serve as evidence that Magnezone is overcentralizing? Is the fact that many teams have a EQ resist, or a fighting resist, or a Sleep Talker, or a Steel Type, a natural part of the metagame? When does this centralization become "overcentralization"? This is an extremely hard line to draw.

One fun potential definition is seeing the % use of a Pokemon and claiming that if a Pokemon use surpasses X%, it is overcentralizing. The argument is the Blissey argument - being the best special wall in the game, it has forced many special sweepers to find ways to deal with it one way or another. Here is an interesting Garchomp statistic.

November: 3.79%
January: 3.81%
February: 4.17%
March: 4.11%
April: 4.17%
May: 4.34%

This is the usage of Garchomp in terms of percentage of weighted points. (December stats are lacking but it doesn't seem too important) We see that the use of Garchomp is rising, and that the better players are using it. But even then, at what percent do we draw the line? 4%? 4.5%? 5%? Note that the maximum percentage a Pokemon will have is somewhere between 1/6 and 2/7 due to the nature of statistics.

There's also the "if the Pokemon is broken, it is overcentralizing" - and this is true, a metagame will center around the broken Pokemon.

However one thing that everyone should consider (or at least, try to dispute is the following) is that the metagame is centralized by definition. This does not imply that if we ban Garchomp something else will take the place as the #1 threat. We see that Garchomp was a natural part of our metagame from the start - so there is no doubt the metagame has already centralized around it. Then how does one define "overcentralization" in this case? One proposal I have made was to have a testing ladder on Shoddy that involves banning Garchomp and seeing the effects of the metagame after a month or two - and analyzing it and see if it's significant - but the problem with this is that there is no precedent and we cannot determine if it's TRULY significant. Perhaps it would be significant if some pokemon loses some points in ranking and the overall variety increases, but then again, how do you draw the line?

And even if we found that it is overcentralized, does it not matter? One comment I have heard is that "decentralization should be our goal" - but why should it be our goal? Why does it matter? Do we really want more variety? What would having more variety do? Would it truly make things more "enjoyable"? Essentially, is the current metagame so stagnant is all the battles really "just the same" that more variety is truly needed to keep things fresh? (If one person argues the "YEAH OU IS BORING/UNCOOL" hackneyed argument I will show them why I'm called Tangermean - do not attempt this argument unless you know what you're talking about and very well versed in the metagame.) Could you even argue this on a pure statistical basis?

In the end I feel as if everything just boils down to this - that overcentralization is a very arbitrary definition that has the potential to mean everything to absolutely nothing - and the overcentralization is defined based on what people want in the metagame - people who are fine with it and people who want "more variety". This is what I feel as if the argument waters down to - and it's a question that needs to be answered - that "What Makes a Solid Metagame?" - aka "What do you want in your metagame, how much variety is good?"

Discuss, answer questions, question the points, etc. Make sure you read the post carefully and fully before posting and making your point.

EDIT: This is NOT a Uber Discussion, this is NOT a Garchomp discussion. The Garchomp arguments were used because that's the most common arguments. You can use Garchomp as an example of something that's overcentralizing and why, but DO NOT make baseless and ignorant comments such as "Everyone uses HP Ice for Garchomp" and other comments that just proves you did not read this thread.

The problem with variety is indeed that, variety. Once a metagame centralizes you have a pretty good idea of what is going to be on teams, at the moment you know Garchomp is going to be on a large number of the teams you face, and you plan accordingly. Your team should counter Garchomp but also counter Garchomp's counters, which doesn't seem to be a major problem. But once you eliminate Garchomp and theres variety to contend with, no longer does every team have one of Garchomp's counters because it is not needed. What happens then is up in the air, what Pokemon will become the main threat you have to counter or will there be such variety it is impossible to build a decent team?

There is always a Pokemon you have to plan for in every generation, RBY had Tauros, GSC had Snorlax, and ADV had Tyranitar. Every team made in those gens HAD to counter that Pokemon lest they get sweeped, just like now. Garchomp happens to be that Pokemon of DP, counter it, or you will get sweeped. The removal of Garchomp might not change a thing, a different Pokemon becomes the main threat to every team, a testing ladder might show this, but to accurately see if something is taking over the game, you're going to need more than 2 months. We might need upwards of 9+ months for a trend to start. Then every team will be built around countering that Pokemon.

And the other option of variety, is it really to be sought after? Right now you know Garchomp and its counters are going to be on a large number of teams, with the removal of Garchomp you don't have that knowledge when team building, you then have to plan for a much larger number of threats. Even with the removal of Garchomp, variety may even only occur for a couple months, until a new major threat is discovered and once again abused and debated like Garchomp is currently.

I think variety is clearly what is desired in a good metagame.

I think we can all agree that we dont want the metagame to consist of 6 pokemon and 6 pokemon only. It totally eliminates the team building aspect of pokemon, and pokemon without a team building aspect is not pokemon. This is clearly an example of over-centralisation, so you have to accept that over-centralisation can exist and is a bad thing.

If we consider the possibility of too much decentralisation, the argument runs that this eliminates the battling aspect of pokemon as all battles are decided by the particular choice of team. But I dont think this is a certainty in the way that overcentralisation of the metagame eliminates the team building aspect. I think a metagame could be imagined that is entirely decentralised - having an infinite range of pokemon - in which battles are still generally decided by battling skill.

I think that Pokemon is a deep enough game to prevent the possibility of becoming paper scissors rock, at least for humans or any current computer. I mean any team can feasibly lose to any other, unless one of the teams was designed to lose, so given that there is a scenario one team may win (and that scenario may require a large chain of critical hits), then battling ability must come into play. I mean, the possibility of predicting exactly every single one of your opponents moves is unreasonable to expect, but it is possible. Therefore there is an almost unlimited possibility for improvement of battling ability no matter how decentralised the metagame becomes.

I think I have coevered everything.

Have a nice day.

But once you eliminate Garchomp and theres variety to contend with, no longer does every team have one of Garchomp's counters because it is not needed. What happens then is up in the air, what Pokemon will become the main threat you have to counter or will there be such variety it is impossible to build a decent team?

Click to expand...

The metagame will develop and turn into a metagame once again - giving you a good list of threats. Just because the one pokemon is gone doesn't mean that the others are suddenly not a threat - we still know what to expect and these will still be threats. If yes, one pokemon's lack of presence in the metagame does turn give the metagame a ridiculous number of variety such that the argument you have becomes true, then I think we can say that it is "overcentralizing", although this would make the definition of overcentralization a good one - a force that keeps the metagame healthy.

I think a metagame could be imagined that is entirely decentralised - having an infinite range of pokemon - in which battles are still generally decided by battling skill.

Click to expand...

It wouldn't be a metagame (at least by my definition). The idea is that it will always centralize around something, and the condition you set forth can be the "bliss point" - the point we are trying to achieve within the metagame. Yet how do we determine what that point is? Too centralized and we get the scenario you described above, and not enough centralization results in zerowing's scenario.

Therefore there is an almost unlimited possibility for improvement of battling ability no matter how decentralised the metagame becomes.

Click to expand...

The argument against this, if I remember correctly was this example

Imagine 5 fighting Pokemon, where one of them was clearly the best, and the rest are a mixed bag.
We ban the one that is clearly the best, and allow the other four to shine. The threats would be counterable simply because there existed a flying type that dealt with the one that was the best - so it can deal with the rest.

This just sounds like we'll be going into a more stallish metagame - and if we start banning walls such that we try to prevent this, then what stops it from entering an infinite loop, where we decide on the next best fighting pokemon and use that instead? Yeah, this may be a poor argument but I can imagine this being the case.

Or, perhaps that this would lead to each of the 4 Fighting Pokemon evolving into unique threats. Then how do we deal with them all? This is the concept behind zerowing's scenario - that we will not be able to - and that it will be a rock paper scissor scenario where someone who runs X Pokemon will win most of the time against someone who can't counter it because there are just too many threats.

I think that Pokemon is a deep enough game to prevent the possibility of becoming paper scissors rock, at least for humans or any current computer.

Click to expand...

It wouldn't be a true rock paper scissor situation - but I think if I don't carry anything that can deal with Mixape perhaps because of the hypothetical situational that there are too many threats, and my opponent has a MixApe, assuming equal skill, he will beat me most of the time. This is what I meant by the rock paper scissor situation - doesn't matter how "deep" the game might be, at least in my opinion.

I guess by infinite number of pokemon I really meant any arbitrarily huge number.

And I dont think Zerowings problem is a problem. It is just having to consider a larger number of threats. I guess some may be unhappy about having to do that, but perhaps those people should consider playing Ubers, or RBY whatever. Decentralisation is the goal of the OU, pretty much by definition. Anyone could argue that this is arbitrary, but that is why we dont pretend that OU is the only ruleset worth playing.

Have a nice day.

The question we have to ask ourselves then is: Does anything in the current metagame cause so much overcentralization that we no longer have any variety in our teams? So much so that it should be banned from the metagame?

Obviously if a pokemon forced us to have 6 counters for it such that every team consisted of those 6 pokemon or face being swept, that pokemon is overcentralizing but that is wildly exaggerated. On a realistic level, where do we draw the line when it comes to overcentralization and variety? 2 counters for a single threat on every team? 3? 4?

The obvious pokemon in question here is garchomp but I feel it comes down to more than just a single pokemon. If we apply standards for determining what is overcentralizing or not, the standards should remain the same for all pokemon. Garchomp may force us to have more than 1 pokemon on our team that can stop it, but the same can be said for many others. If we are to ban garchomp based on this, then it stands that we should ban these other pokemon as well.

I know in one of the other threads someone said it's overcentralizing if it takes more than 1 pokemon to counter a single moveset and gave SD yache chomp as an example. I'll just point out now that this isn't true as any given set can be countered by a single pokemon.

Does garchomp reduce variety? No doubt it does but if we're trying to counter everything then so do a lot of other pokemon. But does garchomp reduce variety to the point that it's so extreme it has to be banned? That's for you to decide based on where you choose to draw the line and your definition of overcentralization.

One thing to keep in mind is that if we are to ban something, the same standards apply to everyone. Let's not do the same thing we do with Double Team and ban it just because we don't like dealing with it. I mention this as no doubt Garchomp is not unrelated to DT if you know what I'm getting at.

Decentralisation is the goal of the OU, pretty much by definition. Anyone could argue that this is arbitrary, but that is why we dont pretend that OU is the only ruleset worth playing.

Click to expand...

I think there's already a metagame with a ridiculous amount of decentralization - which is PBR, where you take 6, you see your opponent's team,and you pick 3. With 3 vs 3, where you will not be able to counter every threat, thus allowing for a ridiculous amount of decentralization and becomes "who pulls off their strategy better wins" - and I think this is what you're hoping OU would be.

That kind of environment for "OU" would be fun indeed to play in, but I'm curious if it's even possible in 6 vs 6 because of the likelihood for 6 vs 6 to centralize in one way or another constantly (the infinite loop argument), or stallish (walls not being banned).

Is it ok for me to give up and wait for the smart people to decide what the tier list looks like?

Quite frankly, through reading all the arguments, I've come to the conclusion that there is just no way to decide this well. We argue back and forth over 1 or 2 pokemon in a tier of 50+, and in end even the people at top realize there's no objective way to do it.

It's going to all come down to opinion at some point. I myself want to do 1 of two things:

1. Wait for the smart people to tell me which pokemon I can use. At this point, I really don't care. Arceus can come down from ubers and I don't mind so long as there is a solid list of "this you can use, and this you can't."

2. Suggest we organize a subjective way to decide.

A straight out vote/poll probably is not the best way, but I belive we have to acknowledge it comes down to subjective decisions as to what is tested and/or banned. I think we should organize to decide accordingly.

Well firstly that's an unfair comparisson. 3v3 has less depth than 6v6 pretty much by definiton.

Likewise there is a metagame with a great deal of centralisation - RBY.

The threat of paper scissors rock is exaggerated greatly in my opinion, and it extends mostly from use of stall teams. People who build a stall team, then battle lazily - because stall teams are effective against poor players even when playing lazily - and then complain when they meet a pokemon their stall team does not adequately handle. I mean I am being unfair here, and talking in generalisations. But basically if you design your team with only countering pokemon in mind, then, as you cannot counter everything, you are creating a paper scissors rock effect almost by definition.

Think of it this way. If teams did not have an advantage over other teams, then there would be no need for team building. There will always be some level of disadvantage in pokemon based on team building unless everyone uses the same team. So I guess given that you dont know your opponents teams, you can either counter a specific team, and hope your opponent uses that team, or you can try and cover every possible team equally, or you can do something in between. But no matter what you do, a good team should always have a hope of winning, so battling ability always influences the outcome of the battle.

I think maximum decentralisation ought to be our goal in terms of levels of decentralisation. Anything else is arbitrary and subjective.

Have a nice day.

Sirlin said:

A multiplayer game is balanced if a reasonably large number of options available to the player are viable--especially, but not limited to, during high-level play by expert players.

Click to expand...

Sirlin words things so much better than I do, here's his full article here for the interested: http://www.sirlin.net/archive/game-balance-part-1/

If the only way to "dominate" the game of Pokemon is through Garchomp, then, that's not "balanced" by definition even though every player has access to Garchomp. If there was a single strategy discovered that countered 99% of other strategies out there, the game would then evolve to the point where, the average Joe, having discovered this strategy after being beaten by it, joins in either with the strategy, or the counter team, and the whole game becomes that single strategy. I think maybe that's what we're looking for when explaining terms like "balance" and "centralization" in making sure that the game remains varied.

As you mentioned, overcentralization is a term which is thrown around a lot without much of a consensus on definition. I know the way Colin determines overcentralization on shoddy battle is: if the pokemon comprising the top 75% of usage (aka OU) falls below a certain number (I don't recall exactly, but it's around 50), then the metagame is overcentralized. I think this is a good method to determine overcentralization, since it involves pretty much zero subjectivity beyond the arbitrary %usage and the number of desired pokemon. If we accept this definition, then there is currently no need to change the OU metagame in any way on account of overcentralization.

This method implies that variety is important, but it also takes into account that it shouldn't be the only factor involved in the creation of OU. If variety was our only concern, then the largest metagame would probably be some combination of NU and NFE; however in actuality variety takes a back seat to the players' desire to use the strongest pokemon possible. We do not seek variety; we do however seek variety above a certain level.

It is also worth noting that overcentralization is not the only reason to ban a poke. A lot of people (myself included) support the removal of Wobbuffet from Shoddy's OU not because he is overcentralizing (for this would be a faulty argument), but because he makes both players' skill irrelevant as soon as he is sent out--a first time player instantly becomes as good as any veteran.

An argument for banning garchomp which was not mention in the OP is that he has several unique and powerful traits which, when combined, make him far too powerful considering how little prediction is involved in his use.

The first is sand veil, which arguably raises the luck factor of the game above a desirable level. Put more directly, Sand Veil is too rewarding of an ability for too low a cost on such a deadly pokemon.

Another factor to take into account is his 102 base speed, a total slap in the face to every other bulky sweeper in the game. Every pokemon which can take his hits is too slow to dish them back; every pokemon which is fast enough to strike first is too frail to risk switching in. Choice scarf remedies this somewhat, but choice scarf is a very risky item to use due to common stat-ups and (on shoddy) Wobbuffet.

His final trait is something Tangerine mentioned, the "it takes two pokes to counter him" argument. However, unlike Tangerine, I do not believe that this argument is faulty. Garchomp is unique in the sense that no single pokemon with any remotely reasonable can always safely switch into it--even if the set and item are known--and KO it before garchomp takes them down. In my opinion, a pokemon which can not possibly be defeated by any poke switching in should not be part of a healthy metagame.

(I do not want to turn this solely into a garchomp discussion, however it was mentioned extensively in the OP and I felt I should address it)

As I conclude I would like to draw further attention to ChouToshio's post, which is a few spots above mine, because he is absolutely right. Pretty much every Garchomp argument has been made; in fact this thread already sums all of them up pretty nicely. I know it seems a bit arbitrary, and a simple majority might not be enough for a change, but perhaps the best way to go would be to simply to take a vote among badgeholders? If nothing else it would quell the onslaught of "X poke should be OU" threads and establish what I think is a reasonable way to determine the validity of a tier change. Obviously this isn't a perfect method, but I would like to hear others' opinions.

My response to Colin's statistical method.

Basically if I had to summarise your argument, it is that Wobbuffet and Garchomp are overcentralising, but that it isnt shown because of flaws in Colin's method, some of which are mentioned in the linked post.

And about ChouToshio, we pretty much all agree. Unfortunately the debate that has occured has been fragmented and disorganised and a lot has been missed. That is why there is this push to finalise the arguments, with threads such as this one, to help figure out how best to proceed.

Have a nice day.

Is it ok for me to give up and wait for the smart people to decide what the tier list looks like?

Click to expand...

It maybe okay for you, but it's no longer okay for any of us - this thread is aimed to address the bottleneck that is stopping any action from being taken - the arbitrary definitions that are being thrown around by both sides of the party.

I think this is a good method to determine overcentralization, since it involves pretty much zero subjectivity beyond the arbitrary %usage and the number of desired pokemon. If we accept this definition, then there is currently no need to change the OU metagame in any way on account of overcentralization.

Click to expand...

The problem with any statistical method is that because the percentages are arbitrary, we find that hard to accept as the definition of overcentralization. You may draw the line and say "this is overcentralizing, this is not", but what does that mean in terms of the game? What does that mean in context to the metagame itself?

imperfectluck said:

Sirlin words things so much better than I do, here's his full article here for the interested: http://www.sirlin.net/archive/game-balance-part-1/

If the only way to "dominate" the game of Pokemon is through Garchomp, then, that's not "balanced" by definition even though every player has access to Garchomp. If there was a single strategy discovered that countered 99% of other strategies out there, the game would then evolve to the point where, the average Joe, having discovered this strategy after being beaten by it, joins in either with the strategy, or the counter team, and the whole game becomes that single strategy. I think maybe that's what we're looking for when explaining terms like "balance" and "centralization" in making sure that the game remains varied.

Click to expand...

Hmmm, this post made me think of one potential definition of overcentralizing. For this, we assume TAY's counterpoint to the point in the OP is correct - that people *are* using more than two Pokemon per team because of Garchomp's presence.

A Pokemon is overcentralizing if it prevents the metagame trend from changing.

A trend, in the case of Pokemon, is the most popular strategy around the times. Trend assumes centralization around a strategy, and of course, if we get to the point of decentralization that Hipmonlee wants, then the trend has the potential to change all the time - simply because it's the best tactical decision - and not because one is "forced" to - whether it be due to sheer power or because the Pokemon is just that centralizing, which is the case for Garchomp right now, I feel.

Let's explore this definition, again by using Garchomp since he is by no doubt our biggest centralization force.

http://www.sirlin.net/archive/yomi-layer-3-knowing-mind-of-the-opponent/ This is the general concept, I think.

How long has the metagame been simply about Garchomp, and countering Garchomp? Sure, one can argue that Wobbuffet has changed this, but with Tickle being removed, could we say that Wobbuffet will be as effective anyway?

Is the presence of Garchomp then preventing the metagame trends from changing? There are many ways people attempted to try dealing with Garchomp - would there be a counter strategy that would arise for this?

Essentially - is the presence of Garchomp preventing new strategies from forming, other than trying to find new ways to deal with Garchomp? (Is a "trend" of Pokemon attempting to counter Garchomp's counters failing because of Garchomp's presence?) Thus, when playing to win, does the presence of Garchomp just prevent this from happening?

From the testimonials I have heard from players and my experience, this has the potential to be very, very, true. This, and assuming Shoddy Battle statistics mean anything, we see that Pokemon usage points are barely changing - other than some arbitrary changes in places. "More Variety" then can be solidly defined as not simply as "more decentralization" - but a metagame constantly in development, never quite centralized and "fickle".

Or perhaps this is a definition way too simple for Pokemon - where there are six Pokemon in a team.

An even more interesting statistic is that Blissey had a 4.7% usage at one point (I think it was December), and nobody was saying for it to be banned...

I'm a bit hurt about saying that "your definition of OU is arbitrary", as I was the one that defined it, Hip, and not Colin. :(

So hereby I'll explain again why the definition of OU is not arbitrary.

OU are simply the Pokemon that are used most often. That is something everyone agrees about. The question is "what does 'used often' mean? How much is 'often' exactly?" And here is how I answered these questions.

Every game contributes to at least 7 Pokemon usages and at most 12. That means that an average Pokemon game will see 9.5 Pokemon, and this corresponds with Shoddy's statistics. Since a game will see at least 7 Pokemon for sure, I defined OU as being the minimum amount of Pokemon that appear in 7 out of 9.5 games. Since 7/9.5 = 0.7368 and this number is very near 0.75, I decided that OU should consist of those Pokemon that appear in 75% of games. OU currently consists of 47 Pokemon, but it must be said that this was once as high as 49 and 50. Interestingly, the number of OU Pokemon lowered from 49 to 47 at the time when Wobbuffet and Deoxys-S were introduced to OU.

Because of this, Colin defined the amount of centralisation in OU by the number of OU Pokemon it contains. The higher the number, the less centralised the metagame is, and vice-versa. I agree with this definition. If a metagame is centralised, it would contain few viable Pokemon, and hence only few would be overused. A less centralised metagame would allow more Pokemon to be viable, and hence more would be overused.

This means that knowing if a Pokemon provides centralisation or not is quite simple, in my opinion. There are two scenarios: either the Pokemon was banned and is introduced (like Wobbuffet and Deoxys-S), or is allowed and then banned (like what some people want of Garchomp). In the former case, if after the new Pokemon is introduced, the OU list is shortened, then the new Pokemon centralised the metagame more. This actually happened for Deoxys-S and Wobbuffet, but since both were introduced at roughly the same time, it can't be known exactly whether one was the culprit or the other (or both). In the latter case, if after the Pokemon is banned, the OU list is lengthened, then the metagame was more centralised with it than without it, which would confirm it that it was a centralising force.

I think the above method would be good enough to know if a Pokemon provides more or less centralisation to the metagame.

I know I'm not exactly the most intelligent person of the bunch around here, but if I may I'd like to make some statements about this.

I'm going to say that about these "Garchomp discussions" because those are the ones that appear to pop up the most. What the problem is is that many people look at Garchomp at a defensive standpoint (by this I mean how to actually "counter" him) and how it affects teams. Yet even though I highly disagreed with a lot of KapKap's discussion about the other four Pokemon, it sort of snapped into my head that when people look at a counter they look at every set possible; even the niche situations such as Metagross carrying Grass Knot for Swampert. For Garchomp, we look at all the sets that are possible (and for an outline I'll post them here):

- Swords Dance
- Earthquake
- Dragon Claw / Outrage
- Fire Fang

This would be the #1 common set with items varying from Yache Berry, Salac Berry (the EV'd version to survive an Ice Beam), and Life Orb.

- Substitute
- Swords Dance
- Earthquake
- Dragon Claw / Outrage

This set would usually be used with a Salac Berry (or Brightpowder if you allow hax items). This would be the next set that I find to be extremely effective as it even surpasses Deoxys-S after the Salac activation (and if you're Jolly of course).

- Outrage
- Earthquake
- Fire Fang / Fire Blast
- (The last slot is really filler but let's assume Dragon Claw)

This would be both the Scarf and Band set. And finally:

- Swords Dance
- Draco Meteor
- Fire Blast
- Earthquake

The final set that was [somewhat] created by Surgo. The "Chain Chomp" set.

Now understand that I'm not making this topic to become an overall "Garchomp discussion" as the topic is about overcentralization and such. I think to understand overcentralization, you must look at THE MOST COMMON SET THAT MOST PEOPLE WOULD USE, NOT DIVERSITY-WISE. In the examples above, the people who assume Garchomp counters assume that you should always be able to counter everything Garchomp tosses at the opponent. However let's ask ourselves: when will I ever use Choice Band Garchomp outside of Ubers? When will I use Chain Chomp over the Swords Dance Chomp? How "overcentralizing" is something that has Sand Veil? How does this make the metagame look unbalanced and why? I know this sounds a bit complicated, but there is more to finding out how a Pokemon is considered "Uber" than by stats, statistics, and "diverse" movepool. By this we should look at two things:

1) The most common set used on the Pokemon.
2) The considered most dangerous set used on the Pokemon.

We can all agree that the most common set AND the most dangerous set on Garchomp is the Swords Dancer by no doubt. That's how we should define, in my opinion, how Garchomp (or some other Pokemon for that matter) are truly "Uber" or not. I guess in Garchomp's case it goes back to Obi's topic on "What is Uber" and such. I also can agree that it can be determined how the Base stats are spread but we could also argue on that one all day.

But to fully answer about creating a Garchomp-less server is where I'll agree with Zerowing completely. It will take more than 2 months to even accomplish what the next "great threat" will be. Hell we can even argue that it took us THIS LONG to figure out how much of a threat Garchomp is and it's been over a year. I think that testing is something that needs more than just effort and that it also includes time, and perhaps in this case a lot of it. And to add on his agreement, we've had threats like this before in other generations and despite how we bickered about them being Uber they've stayed in the OU metagame ever since. I guess to me it's more than just balancing that I'm worried about as far as the metagame goes. What I'm worried about is what will happen with the absence (sp?) of Garchomp and how the metagame will then become. Will it still stay offensively? Will it curve into another "stall episode?" Will it even make us revert back to using a more balanced team? Will a possible new threat go through this same episode of "it could be banished to Uber" again? Those are the questions that should also be factored in as well. Hopefully I don't sound completely stupid posting this and that I've at least hit a few points here...

Just to put things in perspective, the number of Pokemon that are considered 'OU' in the ubers metagame is 14 (both for April and for May). Comparing this with 47 to 50 in the standard metagame, it shows how the ubers metagame is much, much more centralised than OU is.

Okay, let's determine what would be a "good metagame" before we get into the Garchomp thing.

"What is a good metagame?"

I believe a good metagame is one that allows a wide range of strategies to be viable; yet not such a wide range that it cancels the role of skill and overemphasizes suprise.

For example, a good metagame in my view, would allow for the use of a wide array of team strategies - offense, defense, balance, Trick Room, hail, etc, mostly to equal effect. That makes it possible to win with different strategies, without having one strategy being significantly better than another, which emphasizes the skill aspect of pokemon. I, as a skilled player, should be able to win as much with an offensive team as a defensive stall team.

"What overcentralizes the metagame?"

In my opinion, statistics aside, "overcentralization" to me occurs whenever a team consistently has to devote a great deal of resources specifically to countering one pokemon. Not only is that having "two" counters to that pokemon, it's also having to have means of countering that pokemon with your non-counters.

For example, Hippowdon and Forretress may be your Garchomp counters, but you may also need Gengar to revenge kill with HP Ice, Celebi to cripple it with Reflect, etc. Or you may need Bronzong to take its NVE/resisted hits, but also a bulky water to take its Fire Fangs.

Also, I think a large part of overcentralization is not having "alternate means" to beat a pokemon. For example, take Gengar. While Gengar theoretically has no counters, there are other way of dealing with it - Surfing it with Starmie, having something like Heatran take a Focus Blast then hitting back with Flamethrower, bringing out a Choice Scarfer to take it out, etc. Salamence is the same way - while it's very powerful, it can be beaten via alternate means - it's still somewhat fragile, it's weak to Stealth Rock, and it doesn't have the sheer power to 2HKO it's counters.

"How does/doesn't Garchomp accomplish the goal of overcentralization?"

Essentially - is the presence of Garchomp preventing new strategies from forming, other than trying to find new ways to deal with Garchomp? (Is a "trend" of Pokemon attempting to counter Garchomp's counters failing because of Garchomp's presence?)

Click to expand...

I think it is.

JRrrr and imperfectluck have referred several times to "making certain combinations" unusable in reference to Deoxys-E and Wobbuffet, but I think this concept applies far more to Garchomp.

Garchomp is such a huge threat that you simply have to devote at least 2 spots to beating it or you will lose to it. Furthermore, because of its unique type coverage, there are only a handful of pokemon that can take its hits and survive, and even those are not absolute counters, while because of its stat distribution and typing, it's difficult to "revenge-kill" as it can take hits easily and survive hits from its counters. (forget the 4x weak, having only two weaknesses PERIOD is amazing - that's why Flygon was good in ADV; it was decidedly mediocre but it's typing + Levitate makes it highly effective.) Because of that, it limits what a given team can do strategically, because your strategy ALWAYS must involve having something to switch into Garchomp.

but even still, would removing Garchomp really simply stop making people use Gliscor, Bronzong, Skarmory, etc?

Click to expand...

No. However, the presence of Garchomp makes those pokemon neccesary. It should not be neccesary to need 3 solid walls, and 3 solid walls devoted to countering one thing.

As big a threat as Salamence/Tyranitar/Metagross were in Advance, they were not the same manner of threat that Garchomp represents.

Tyranitar was the #1 threat in Advance because it was unpredictable, with many, many equally effective sets. However, once one knows Tyranitar's set, a good team was usually able to adjust. Metagross suffers from a bad attacking type, Salamence suffers from not having the power to 2HKO things while getting ohkoed by random Ice Beams, etc. They also had to go out of their way to beat things like Skarmory, and had trouble beating bulky waters at all.

Contrast with Garchomp which after a Swords Dance "can 2hko every pokemon in OU" with the ability to survive at least one hit from its potential counters, and often having a perpetual Double Team effect on it. While this is not "unbeatable" - obviously we've all beaten Garchomp - it forces everyone to spend a disproportionate amount of time preparing for it, and with a very limited pool of pokemon because of its type coverage, it's stats, and ability to stat up.

What I'm worried about is what will happen with the absence (sp?) of Garchomp and how the metagame will then become.

Click to expand...

While you lose one big offensive threat, by not having to devote so much effort to countering one pokemon, it expands the amount of viable strategy in the game (Instead of always spamming Bronzong and Hippowdon just to beat Garchomp, I might use something else that nobody's ever seen before!)

How Ironic that I'm slacking off studying for my stat final while I make this post...

Okay, I'm going to expand some things right now. I was talking to Ancien Regime and I think my definition really has potential now.

First of, lets define a novelty set. A novelty set is something that takes advantage of a centralized metagame, or a set that takes advantage of a trend.

If a novelty set is found to be useful, then it becomes a trend. The perfect example is Tentacruel - a novelty when it first entered, and soon it became a trend.

If a novelty set is found to be useless, then it becomes a gimmick. Then a gimmick set is a failed attempt to take advantage of a centralized metagame.

The maximum decentralization point should take into consideration the the proportion of new trends and novelty. If there's far too many novelty sets that work such that a new trend cannot form, then we can say that that's "too decentralized"

The minimum decentralization point should take into consideration the number of new trends and gimmicks. If there are no new trends but a billion gimmicks, it's a sure sign that new trends are not forming.

The question of course - is "where we draw the line". This is something that can only be debated through theorymon and experience - perhaps we need people journaling about the state of the metagame more often - mostly because at this point, statistics is meaningless.

We have Colin's mathemtical definition of "amount of centralization". It's nice and all but what does it mean in terms of how we play the game other than "we expect to see these pokemon"? To be honest this tells us absolutely nothing about how the game is played - Pokemon is far to intricate to be dumbed down to such a simple statistics.

Yes, Pokemon is a game statistics but the statistics and the analysis needed is far too intricate to be contained by the simpleness of Colin's formula on centralization.. There are far too many variables in Pokemon that it's impractical to have a pure statistical analysis and impossible at this point with the data we have right now, not to mention we don't eve know what kind of data we want. This is why we currently have to rely on theorymon and testing the theorymon instead of holding on to "well this is what the statistics say" because that means nothing since we have no precedent and in the end what you guys say with statistics are exactly the same thing as what we're arguing with actual battling experience.

What should be done right now, IMO, is a lot of experimenting for the reasons I mentioned above. This is the only way we'll ever be able to draw solid lines through anything and until then it's "theorymon".

Sorry if I come off as too harsh to anyone - it is not my intention and I was just criticizing the concepts.

I don't want to deeply involve myself in the discussion as of yet, nor do I want to repeat what someone has previously said, which is understandable when there aren't many areas that haven't been covered already. There are no ideas in particular that I have thought of to voice out, therefore I will hold back from entering the in-depth area of the topic.

Quickly, this ladder test that is under consideration with regards to extinguish Garchomp's presense for a total of 'ammount of' months to find out whether he does indeed overcentralise the metagame has my support, so long as we can gather atleast one convincing reason as to why a decentralised metagame is for the better and not for the worse, if it is to have an affect at all. But that has yet to happen, I believe - unless 'a wide range of strategies' are introduced to the metagame we plan to strenghen with a solid end product that is forever developing, is a viable response.

Currently, we cannot accurately see exactly how centralized the metagame is without a metagame to compare it with. Thus, we need to test how many Pokemon would be in OU without those which currently are the closest to be called overcentralizing. Months ago, Blissey was the closest to be called overcentralizing; now it is Garchomp.

Testing the game without Garchomp would help us measure exactly how much the game is centralized around him. Perhaps OU would shrink to 45 (random number), or maybe it would include 60 (random number) Pokemon.

tangerine said:

The maximum decentralization point should take into consideration the the proportion of new trends and novelty. If there's far too many novelty sets that work such that a new trend cannot form, then we can say that that's "too decentralized"

Click to expand...

That definition has potential, but how do we measure it? We can't accurately measure the different sets that may be used.

X-Act's method is one where we can accurately measure the level of centralization.

tangerine said:

What should be done right now, IMO, is a lot of experimenting for the reasons I mentioned above.

Click to expand...

Agreed.

so long as we can gather atleast one convincing reason as to why a decentralised metagame is for the better and not for the worse

Click to expand...

That's actually a good question, and deserving of a thread in and of itself. We're so worried about overcentralization; but what is undesirable about overcentralization? Is having a "wide range of strategies" a good thing?

Arguably, moderate overcentralization accomplishes the same thing as moderate undercentralization - if everyone's using roughly the same teams, that means winning is a matter of skill, as opposed to team advantage. While if all team styles are relatively equal, then winning is still a matter of skill, as there is no inherent advantage - it comes down to what individual players are more comfortable with - player x may favor stall, player y may favor all-out offense with minimal defenses, player z may use a specialty strategy like Trick Room or Hail, and so forth.

If that is the case, then is it right to assume that "more viable pokemon" > less viable pokemon? Is more viable pokemon more fun than less viable pokemon? I think so, and I think there are many who would think so.

X-Act's method is one where we can accurately measure the level of centralization.

Click to expand...

One of my problems with that method is what IPL and others have been speaking to - it only measures in terms of individual pokemon, rather than teams. Because of that, while one may not see a difference in the OU list, (again, Skarmory/Bronzong/Gliscor/Hippowdon are good pokemon in their own right), they make a difference in the types of strategy used in the metagame, because ALL strategies must be devised with dealing with the centralizing pokemon in mind.

X-Act said:

OU currently consists of 47 Pokemon, but it must be said that this was once as high as 49 and 50. Interestingly, the number of OU Pokemon lowered from 49 to 47 at the time when Wobbuffet and Deoxys-S were introduced to OU.

Click to expand...

Last I checked, this datum was batted around, but is false. The OU list only consisted of 47 Pokemon vs. 49 if you don't count Wobbuffet and Deoxys-S, which is illogical for fairly obvious reasons. The fact is that you can only use 6 Pokemon on a team, so every use of one Pokemon must decrease the use of another, in a manner of speaking. They essentially just bumped a few Pokemon down the list, but didn't actually increase centralization.

For an alternate way of looking at it, imagine that we had a game with only 3 Pokemon unbanned: Unown, Caterpie, and Magikarp. We then unban the entire current OU list, and we get what we have now. If we use the method of counting all Pokemon that would fit the definition of "OU", including the "new" Pokemon, the list would consist of about 50 Pokemon. If you only look at the old Pokemon that remained, it would appear as though OU was centralized to a point where there are now 0 Pokemon in OU, which is the same thing as not counting Deoxys-S and Wobbuffet in the OU statistics.

I haven't read the rest of the thread entirely yet, just enough to check to see if this had already been addressed.

Obi said:

Last I checked, this datum was batted around, but is false. The OU list only consisted of 47 Pokemon vs. 49 if you don't count Wobbuffet and Deoxys-S, which is illogical for fairly obvious reasons. The fact is that you can only use 6 Pokemon on a team, so every use of one Pokemon must decrease the use of another, in a manner of speaking. They essentially just bumped a few Pokemon down the list, but didn't actually increase centralization.

Click to expand...

Then why is it that unbanning everything (i.e. playing ubers) would give you only 14 'OU' Pokemon?

And the OU list consists of 47 Pokemon if you COUNT Wobbuffet and Deoxys-S. That means you have 45 Pokemon, plus Wobbuffet plus Deoxys-S.

I think there is something we need to consider here: Why is it we want a less centralized metagame? This has been discussed here somewhat, but I think it is talking too much about creating a healthy metagame? Why is it that we want a healthy meta? I think that a thorough look at our motivations will help us make a better decision.

I feel that the reasoning for decentralizing the metagame is due to the fact that we want different/newer threats. Now, I understand that many of the pokemon that would rise to the forefront are/have already been used, but they would be more "fresh" so to speak than Garchomp. People want different pokemon to build their teams around. Team building is a core aspect of pokemon and one that many people greatly enjoy. We enjoy the creativity sparked by trying to design a new strategy or path to victory. So, I think one of the core questions here is, "How much restriction on team building does Garchomp's presence place?" This is pretty much completely tied to Tangerine's idea of, "How many counters are too much?" I think we need to find how many slots we are donating to Garchomp-related purposes when we build a team. Most teams are going to attempt to include Garchomp since he is such a dominating threat. Now, you can't just use Garchomp and not have an answer, so most players will likely devote 1 or more slots to a counter. Now, some teams may choose not to use Garchomp, but instead just carry counters. So, I think that it would average out to about 2 slots per team in direct relation to Garchomp. (1 Chomp + 1 Counter, 2 Counters, etc.) However, we must also consider how other pokemon one the team are restricted in their movesets/EVs/whatever because of Garchomp. Do you have HP Ice on a pokemon just in case your normal plan fails? Does a pokemon on your team have Reflect to allow your counters to truly counter Garchomp? I think the top players should take a look at their teams and take a look at how many slots we are actually devoting to one pokemon, and not just counters.

A good counterpoint is, "The removal of Garchomp will just mean another pokemon will take its place." This may indeed happen and is extremely valid. What I think we need to look at before deciding to ban a pokemon is how many options become viable as a result of the banning. Where will Garchomp's now vacated usage be distributed. Many feel that it will eventually end up in 1 different threat that will become just as overused. Then what have we accomplished? Pretty much nothing. But, what if Garchomp's place as the top threat was now shared by three other threats? Now people have variety on what to center the teams around and what to counter. Team building will become more difficult. It may actually be harder to come up with a good stall team now that you have to handle three more common threats rather than one extremely common one. Is that what we want? I think we need to look at why we are doing this again. I, personally, would prefer it. It becomes harder to handle, but prediction of what you are facing becomes much more important, and more variety is available. It excites my creative side.

Let me take a timeout to compare this to something I am very familiar with, Magic: The Gathering. I'm sure many of you know what it is, but for those who don't, it is a trading card game. Now, Magic has a large competitive environment with tournaments worldwide. Magic requires a good deal of skill to play, much like Pokemon. It requires people to make the highest percentage play the majority of the time, but also allows for bluffs and carries a bit of luck involved. Also much like Pokemon, players need to construct a deck (i.e. team) in which to compete with. It is another major part of the game. Magic also has different formats in which you play. Legacy is much like Ubers in that everything is allowed. This metagame becomes somewhat centralized in the fact that the best cards ("best pokemon") will be used far more often. Magic also has the extended format (OU environment), which restricts these cards in which many of the extremely powerful effects are limited, so different options are available. Magic also has standard, where this is restricted yet again, which becomes similar to UU. Now, granted, many of the restrictions are based on when the cards are printed, but each format carries its own banned list. Now, each format supports different decks. People play more powerful decks more often, but "rogue" decks often arise that are designed to attack the stronger decks. Now, you can't beat every possible threat, but you can attempt to predict what will show up the most at your specific event and attempt to attack those strategies. Magic has a very healthy competitive environment that generates frequent innovation, which is what I think we want for Pokemon. So how do we translate this to Pokemon?

What we are looking for is to create a metagame that has several "strong" strategies, but also includes counters strategies that can attack the top competitors and actually win. This would create leading trends (A), which then become countered by other trends (B), forming new leading trends. When those trends become dominant, new strategies (C) form to attack them. Now, the with the first set of counter-strategies (B) being weakened by the new counters (C), the original threats (A) can now make a comeback. This pushes an ever evolving and shifting metagame that requires not only great players, but great inventors as well.

Now, the question we need to look at when looking at deciding to ban Garchomp or any other pokemon is how many options is opens up? Is Garchomp suffocating the metagame to the point that we cannot come up with a B trend that would allow the meta to change? If that is true, then I think we need to get rid of Garchomp. Doing so alters the leading strategies somewhat, but may allow other strategies to push to the forefront and keep shaking things up. I think what people are afraid of is that D/P is rapidly becoming what RSE was ending up as. Towards the release of the 4th gen, several people were rapidly becoming bored with the ADV meta, saying that it was becoming "stale." Now, with the increasing pool of competitive players and researchers, the metagame advanced more quickly than expected. I think people are worried that D/P may become stale and are looking for something to stop that from happening.

So, in my opinion, the real question is, "How many options does the removal of Garchomp open up?" Now, the best way to find this out would be testing. Some people have said that it will take a long time to discover what will rise after Garchomp's removal, (should we decide to do it) and that is almost certainly true. But isn't that what we want? This is why we need to look at why we want Garchomp banned. We want to have FUN, and that's what it all comes down to. If we are looking to shake things up, banning Garchomp is certainly going to change things for while. I guess the point I'm getting at is that we need some innovation, but I don't want us to do something that is just a quick fix. We should really evaluate what is opening up and if that is the direction we want to take.

So, what become more viable when Garchomp leaves? Don't say defense. That isn't necessarily true. Like I said earlier, three common threats may prove harder to counter than one very common one. For example, let's say 66% (this is completely arbitrary) of the threats you expect to face are Garchomp. Now, having a good answer to this one threat will earn you a fair number of victories, especially if you can answer the other threats decently. Remove Garchomp. Let's say, for instance that you now are looking at 25% Gengar, 25% Tyranitar, and 25% Salamence. Now, even if you use less resources for each one individually, you are going to still need to devote more slots to cover a similar percentage of threats. The meta would then still be more offensively oriented since it will be more difficult ot counter a good number of things. I'm not saying this will definitely be the case, but it is a possibility.

I feel that Flygon will make a return with Garchomp gone. Not because he fills the same role as 'Chomp, but because the only real reason he didn't make much of an appearance is because Garchomp pretty much outclassed him. Flygon's typing will once again become unique, and it is a very effective typing with a good set of resists and STABs. Flygon will be a different type of Pokemon than Garchomp though, likely utilizing Roost, U-Turn, and good type coverage. So, we get a new pokemon, but Garchomp's void still isn't filled. A new bulky physical sweeper will gain popularity to fill up that spot that many teams need. Dragonite will now probably have more viable sets since it won't be outclassed, but it still doesn't have the ever-useful EQ STAB, is slower, has less useful typing in general, and needs to use DD over SD. (since it doesn't have it) So, other bulky sweepers besides Dragonite will also gain more popularity. (probably more so than Dragonite) This is completely theory here, but I think this is what we need to look at. If Garchomp getting removed makes SEVERAL other things viable, it might be a good idea. If almost all of its absence goes to something else, we really haven't accomplished much.

Do we want to take on the difficult task of balancing the metagame? It would likely take multiple moves until we could achieve a balanced state that would allow the most different options possible. I think people fear the slippery slope. We might overdo it and just end up eliminating a metagame. This is why we need to be careful and making this decision is taking so long. Everyone needs to be patient and wait for the best solution to come forth. If we dive right in, we could take a poor direction and just have to backtrack.

Sorry for the EXTREMELY long-winded post, but (I think) I covered some useful topics. Also, sorry if any of this got posted while I was typing this up, it took me a good while to create.

The question of course - is where we draw the line

Click to expand...

More like when we draw the line and not where. Better yet,how are we going to determine the line? The metagame is always imbalanced. Especially since how we have issues on 600 BST Pokemon that want to become OU. Either way, drawing the exact line somewhere is too much to ask for now, as it only brings to more arguments and tier based threads. Statistics aren't alays accurate either

What Makes a Solid Metagame? - aka What do you want in your metagame, how much variety is good?

Click to expand...

Variety is a thing that many seek for when playing. Playing Garchomps every day can be a huge pain in the butt. However, this does not say Garchomp makes the metagame not variable at all. In fact, this is the type of situation in which people should be inventing new sets, gimmicks or not. Despite how long the game has been out, I believe there is potential, and I believe we can achieve it by working together as a group and stop ridiculing ideas.

Another way to achieve this desired metagame is to set different servers with bans and allowance of certain Pokemon. Every once in a while, the ban list changes according to what the majority wishes, and then they go try it to see how better/worse it becomes. This process may be long and tedious, but it's certainly an action compared to what we have in the official server right now-people simply flowing with the metagame.