The definition of uber

[obi]

There are a few definitions and schemas that help determine what is uber. These are those that I find most common (as well as some stereotypical Pokemon to which they are applied):

Overcentralization

The argument here (most often used with regard to Kyogre and Groudon, the weather ubers) is that if we allow a Pokemon, there will be less usable Pokemon. "If you allow Kyogre, no one will use Luvdisc, Suicune, Milotic, Butterfree, Pikachu, or Geodude, and all teams will be forced to carry their own Kyogre and a Ludicolo, so there will be less usable Pokemon." This argument is entirely statistical. That means that if this is the reason for uberizing a Pokemon, we can use the Shoddy usage list to verify whether this argument holds water. This is the most objective of all the definitions of uber.

No counters

This is the main reason for the complaints behind Wobbuffet and, for different reasons, Deoxys-S. I'll start with Wobbuffet, because it's easier to discuss with regards to the "no counters" situation.

Wobbuffet's ability, Shadow Tag, means nothing can switch out of it. A counter by definition must be able to switch in. Unless the Pokemon out is holding a Shed Shell, has Baton Pass or U-turn, or has Role Play, Skill Swap, or is also Wobbuffet (or Wynaut), you cannot bring a counter in on Wobbuffet. Proponents of this argument must address the issue of Dugtrio, Trapinch, Magneton, Magnezone, and Probopass.

For other Pokemon, the "no counters" argument takes a slightly more vague reasoning. Imagine a Pokemon with 500 base Attack, Special Attack, and Speed. This Pokemon could OHKO anything it faces (and is faster than everything), so the only way to stop it would be with priority moves like Quick Attack, Extremespeed, and Sucker Punch (or an opposing version of itself that is faster). However, unless there were a similarly defensive Pokemon, you couldn't bring in any of these Pokemon to stop it, and thus you'd be forced to sacrifice something to bring in your "counter", and it would likely switch out and come in later to repeat the cycle. Here, the problem isn't that you cannot switch in at all, but that you cannot switch in safely. Obviously this is more extreme than any real example, but Pokemon like Deoxys-A and Mewtwo are similar to this scenario.

Alternately, consider the case of a Pokemon with base 500 HP, Defense, and Special Defense. Here, you can switch in safely (assuming it has low-average offenses), but once you're in, you can't do anything. A counter must be able to switch in safely and threaten the Pokemon it is said to be countering. I may be able to switch in Milotic safely on a Substitute, Calm Mind, Surf, Ice Beam Kyogre, but what am I going to do as it Calm Minds up?

Luck

One of the arguments surrounding Garchomp is that Sand Veil means that even if you have a counter for it, you may not be able to stop it thanks to Sand Veil missing. The more luck influences Pokemon, the less skill does.

Hyper-offensive Pokemon like Deoxys-A also have an element of luck. You are forced to make guesses as to what move it will use and what set it has as soon as it comes out. This decision is forced before you can possibly have enough information to stop "guessing" and begin "predicting".

This is one of the major arguments in favor of banning Double Team / Minimize and OHKOs.

This is probably the most subjective reason of all, as you have to make a judgment on how much luck is too much (or whether the increased element is even there to begin with).



Can anyone think of another reason to consider something uber / banned from standard play?

 

[Amazing Ampharos]

The second point simply shouldn't be in the definition. Garchomp has no counters. Name one Pokemon that can switch in on any move Garchomp can learn safely and then beat Garchomp. If you considered that he could learn moves like Toxic, you could even argue that Electivire has no counters as, when you see a totally unknown Electivire, there's no one Pokemon you can always switch to that will always, 100% of the time win. You can just figure what he's most likely to do and respond accordingly.

I am strongly of the opinion that overcentralization, and variations of it, should be the only criteria. For instance, if a Pokemon were so good that it were on 90% of teams, it would probably be worth considering uber. If it reduced the pool of useable Pokemon by half, it would certainly be uber. If it always forced you to use a Pokemon from a small pool of counters (which could be measured statistically, though I can't be bothered to define an example), it would be uber. Of course, defining the exact degree of centralization acceptable is fairly tricky, especially since you run into recursive issues (whereby unbanning Pokemon A is unacceptably centralizing but not if you allow the barely acceptable centralizing caused by unbanning Pokemon B first and measure A's centralization from that point).

There seems to be some degree of backlash against defining uber statistically, and I suspect it is largely because people want to be able to ban things they don't like by making arguments that rely on opinions (I'm not even kidding when I say the most common argument for Wobbuffet being uber is "Wobbuffet is gay"). Regardless of the final conclusion of any definition we make, I think it's very important to rigidly define uber so there's black on one side, white on another, and no gray in a middle. An official stance on something like that could quell a lot of internet drama.

 

[obi]

The main question for centralization is what is measured. Do we only look at Pokemon usage, or do we also factor in item and move centralization? For instance, a Pokemon like Deoxys-A might not create as much centralization as one might think, because there are just so few defenses against it. It would, however, shift which Pokemon are used (maybe not decreasing how many are used a certain % of the time), and those newer Pokemon would carry priority moves a lot more often.

 

[Cathy]

I think there is an interesting argument to be made for considering only the pokemon used, and not the moves. Suppose we consider centralisation in the move set of every individual pokemon. The trouble here is that the lower-down pokemon in the usage aren't used enough to have any serious statistics, so the numbers for them are essentially random, so we'd expect a high degree of fluctuation between each month in most pokemon sets (all those crap pokemon no one uses). But suppose we only consider move usage in the OU pokemon. Then if the OU pokemon change, we can only compare to statistics that weren't considered reliable last month (and so probably still aren't reliable now). In practice this may not be much of a problem, and certainly the statistics on moves are going to be compiled so that we can actually see how dramatic the fluctuation is in underused pokemon.

On the other hand, we could just look at overall move usage. The various priority moves might comprise a lot of the top move usages after unbanning Deoxys-A.

I don't think "luck" should be used as an argument for banning anything because this is a game of statistics. Your goal should be to maximise your chance of winning, not to win every match. Things that involve "luck" might still end up being overcentralising though--perhaps you need to carry multiple Garchomp counters to deal with its Sand Veil and this creates a lot of centralisation. I don't think we need this as a separate criterion. (I also don't think this should be used as an argument against OHKOs and evade moves, but I realise the topic isn't about them. I'm just mentioning it because Obi mentioned them.) However, I realise that not everybody agrees with my long-term approach to this.

One distinction I think worth developing is whether being uber is an intrinsic property of a pokemon. The method I have advocated and that I give a brief description of on Stark leads to the idea that being uber is not an intrinsic property of a pokemon. I posted this in Curt's topic but it's not going to get much exposure and it fits better here, so it follows in the body of this post:

One objection that often comes up is: "Well maybe the pokemon isn't centralising because of X, Y, and Z and therefore it's still uber". Often X, Y, and Z take the form of social taboos. These people like to think of the Shoddy Battle ladder as just an approximation of a theoretical metagame where everybody is using the best sets on each pokemon, and the best pokemon. But the mistake here is that there is no reason to believe such a theoretical metagame exists: it is more likely that the metagame will in fact fluctuate endlessly, rather than approach a final "limit" as the best sets are discovered.

So how do we counter this problem? Instead of viewing "uberness" as an intrinsic property of pokemon, I consider it a function of the pokemon and the metagame it is being used in. There are some metagames where pokemon we now consider uber would not be uber. They are uber in our metagame, however. Rather than viewing an uber list as something that we discover and then let sit, I think it should be viewed as something that can be periodically revised as the metagame shifts.

Does the metagame actually fluctuate enough to warrant revisions? Probably not. But it might, and if factors X, Y, and Z are currently causing Wobbuffet not to be uber, then when those factors cease to exist it will be uber and it can be banned then.

Formally, I would define a metagame as a map of species : (usage, set usage). Suppose there exists a metagame where all pokemon are legal but -- for whatever reason -- only Magikarp is used. In this metagame, Magikarp is uber because if it is banned then people are forced to use other pokemon, so the game is very much decentralised. I've picked this far out example to stress my belief that the reason why the game is not being centralised or is being centralised is immaterial, since the list could in theory undergo revision anyway. Note that the act of banning Magikarp creates a new, decentralised metagame. Magikarp might no longer be uber in this new metagame. (We could test by unbanning it again and seeing if suddenly Magikarp was the only pokemon used again--or at least if it was more centralised.)

I think there are two questions left to address with the centralisation definition (which I think is the only workable one). The first is: how do we decide which pokemon are worth banning or unbanning to check for change in the centralisation of game? I think to a large degree this will have to be an intuitive decision, based on more specific concepts. For example, if there seem to be good reasons why pokemon X could be centralising the game, then that would make for a good test candidate -- but the existence of seemingly good reasons isn't enough to ban it if in practice the game isn't decentralised by unbanning it. I can't think of an objective way to determine what to test though. (AA also discussed this in his post.)

The second question is: should we try to maximise the number of viable pokemon, or should we try to get it above a critical threshold, while minimising the number of pokemon on the ban list? I am going to argue that it should be the latter because this way the lower tiers (e.g. UU) will contain a lot more pokemon, and there is a higher chance of a pokemon being viable in some metagame. If we make standard as decentralised as possible then many pokemon will have a place in no metagame.

I also think that the decentralisation method lends itself nicely to crafting a new standard by starting with all pokemon unbanned, and gradually choosing pokemon to ban and observing the effect on the number of viable pokemon, but we might need to think a bit about my first question (how to decide what to test).

 

[X-Act]

My definition of uber is the following:

A set of Pokemon which, when removed from the standard metagame, produces the highest amount of usable Pokemon.

This basically corresponds with whoever is in the 'overcentralisation' bandwagon.

I would start from the premise that no tier system is 100% balanced, as I said elsewhere (quoted underneath):

...I am of the opinion that there is no tier system that is 100% balanced, no matter how hard we try. Say we remove Garchomp and Tyranitar... then Salamence and Heracross would be the #1 overcentralising forces. If you remove those as well, you have things like Lucario. And so on. You simply cannot create a balanced metagame...

The real test would be to find the tier system that has the largest amount of playable Pokemon. And this is no easy task. Even if you would test out all the possible metagames containing every subset of the 498 Pokemon (the number of possible metagames has 150 digits, to give you an idea, although testing only the fully evolved ones would leave you with around 100 digits less... still a daunting task), you still need to answer the question: "how am I going to prove that Pokemon X is overcentralising the metagame objectively?".

Because of the problems outlined above, I'd suggest to let the metagame be, for the following reasons:

1) We don't have any ways and means to prove objectively that removing or adding Pokemon X to the metagame would improve it from the point of view of balance.
2) To test each and every Pokemon would take forever.
3) Every metagame is not 100% balanced anyway, so we'd just as well stick to what we have.

Of course, my opinion stems from my definition of uber, which may or may not be the same for everyone.

 

[Mekkah]

I prefer X-Act's way of looking at this (though it is, as far as I'm concerned, merely a perspective, and doesn't mean that I disagree with Obi, Colin and AA at all). Mostly because it shows light at the end of the tunnel. Amount of usable Pokemon is something we can work with and compare.

While it would be a suicide for the social lives we have left to test them all, we can start with some theorymon on what a certain Pokemon would do to the metagame. Some things would be doubtful and could be tested, others can be determined merely by looking at stats (for example, Mewtwo would make Alakazam about completely pointless).

There's a few factors that makes ubers render other Pokemon useless:
- outclassing: Rayquaza makes Dragonite about as obsolete as he can be, so unless you want two Dragon Dance/Outrage Pokemon on your team without breaching Species Clause
- countering to the extreme: If a Pokemon is so good that he is very commonly seen, that makes the OU/BL/UU Pokemon he is countering much less viable. I would raise Tauros as a rather hard to play force if Giratina stands in its way all the time.

My brain is kind of overloaded now to think up more, but you probably get where it's going.

That said, here's a few points I've been stressing every time this came up:
- On top of what X-Act said about there always being an overcentralizing force in the metagame, people will always be divided on what is uber and what is not. Stark retards, Stark intellectuals, Policy people, Smogon Staff, Shoddy regulars, you can bet they all have a different opinion. That means there won't ever be a full concensus.
- Testing different sets of 'suspects' gives you different results, which makes the picture incomplete. If Kyogre countered Rayquaza, that's nice and maybe balanced, but Kyogre may still be too strong. Still, if it's all that holds back Rayquaza, both should be banned. This may in turn cause other suspects to be stronger or weaker. It's ever-shifting and hard to measure.
- The Deoxys-E problem: by allowing ubers one at a time people will be using it a lot, and therefore trying to counter it a lot. This doesn't have to result into an extreme shift of usage to Deoxys-E counters - after all, Metagross and Spiritomb among others were already alive in kicking. The problem becomes apparent when you can't think out of the box far enough without getting pulverized by Deoxys-E, because you'll need a counter to it and that is limiting your pool.

That's all for now, I guess, for this thread is plenty tl;dr.

 

[Amazing Ampharos]

I think the biggest problem with defining the rules that maximize the number of usable Pokemon as the right rules is that it would pretty clearly mandate a sweeping set of bans. Garchomp seems to pretty clearly limit, not expand, the metagame as it forces you to have some degree of a specific answer for him. Dugtrio is certainly doing many low defense Pokemon weak to ground a great disservice. I think if we wanted to maximize the number of useable Pokemon, we'd just want to ban Blissey and all the best offensive Pokemon (the dragons, Tyranitar, Metagross, Heatran, Gengar, Weavile, Heracross, Gyarados, etc.). Blissey being gone would prevent special from being obsolete, and removing all the best offensive Pokemon would make the metagame really slow. Basically, the goal would be to recreate a GSC style metagame except without a "curselax" to keep everything in check; with a plodding progression in every match, a wide array of Pokemon would be able to be used to great effect, especially since the incredibly low level of offense would make Rest a viable move again (allowing all sorts of crazy things to stick around). Just using the non-banned best Pokemon remaining (the walls) would leave you open to a Taunt + stat boost plan, but the Pokemon left to do that would be weak enough so that a lot could deal with them.

I doubt anyone actually wants that, but that's my guess at what the way to maximize the number of usable Pokemon without any regard to anything else, including how fun the game is to actually play.

 

[Cathy]

I'd suggest to let the metagame be

There's no reason to believe that the current set of bans produces a particularly decentalised game compared to another one. And if the current game doesn't include, say, Deoxys-S, it's already apparent that the game with Deoxys-S isn't anymore centralised, so why not take advantage of that data? I agree that it's a problem to decide what to test, but I don't see a problem with allowing a bit of intuition here.

If the ban list can't be modified at all then it really has no connection to decentralisation. The definition of "uber" is just "what is on the uber list".

It definitely seems that most people don't want the metagame to change from the one they already enjoy, so this definition would probably be popular, but it seems to make all the talk of creating a balanced metagame more or less facetious since the initial ban list was essentially created through intuition rather than any data on centralisation.

The way I think implementing the centralisation definition could become practical is making a series of bans based to some degree on intuition, and seeing the effect on the game, with a week in between each check or something. We could probably arrive at a new standard in just a few months with this method. I agree it's not practical to test every metagame, but that doesn't mean we should test none. How do we know when to stop since there is no "balanced" game? I would say to stop when the current number of viable pokemon matches the old standard's, but the ban list is shorter, or, if that is never reached, when the number of viable pokemon exceeds the current number.

 

[X-Act]

There's no reason to believe that the current set of bans produces a particularly decentalised game compared to another one.

That is exactly my point. There's no reason to believe that changing the set of bans produces a particularly decentralised game either. Whatever the metagame is, there would be Pokemon that would centralise it. Hence my suggestion not to bother about changing it.

 

[obi]

Just because there is some measure of centralization doesn't mean that all levels are equal. The whole point of a test is that it isn't permanent. If we find that we don't like how it turns out, we can always go back.

 

[X-Act]

Granted. But my two biggest worries are these:

1) We need at least two months to have a hint of the effect of one new Pokemon in the metagame, which is too long.
2) We won't agree on what the effects of the new Pokemon in the metagame are.

 

[Jumpman16]

I think there are two questions left to address with the centralisation definition (which I think is the only workable one). The first is: how do we decide which pokemon are worth banning or unbanning to check for change in the centralisation of game? I think to a large degree this will have to be an intuitive decision, based on more specific concepts. For example, if there seem to be good reasons why pokemon X could be centralising the game, then that would make for a good test candidate -- but the existence of seemingly good reasons isn't enough to ban it if in practice the game isn't decentralised by unbanning it. I can't think of an objective way to determine what to test though. (AA also discussed this in his post.)

I agree both that the centralization definition is the only viable one and that the decision will largely have to be intuitive. The "No Counters" suggest Obi made can be valid but in my opinion only really applies to Wobbuffet (and Wynaut) because the other pokemon mentioned can only trap a specific subset of pokemon (Steel-types, non–Flying-types and non Levitaters). We could possibly try to enforce "no counters" for Wobby but then we would actually have to see just how overcentralizing Wobby is anyway.

I think a great way of testing this would be to just test Wynaut, since it has the same "ability" (lol) as Wobby but obviously worse stats. I mean, we can even use "intuition" to come to the conclusion that its worse stats will keep it from "overcentralizing" the metagame, though it still has "no counters".

The second question is: should we try to maximise the number of viable pokemon, or should we try to get it above a critical threshold, while minimising the number of pokemon on the ban list? I am going to argue that it should be the latter because this way the lower tiers (e.g. UU) will contain a lot more pokemon, and there is a higher chance of a pokemon being viable in some metagame. If we make standard as decentralised as possible then many pokemon will have a place in no metagame.

Agree with latter as well, but can you explain why having standard as decentralized as possible would adversely affect the number of viable pokemon in any metagame? The way I see it, if we're going to let the number of viable pokemon in UU (the other metagame besides the ones people care about most in standard and uber) affect our definition of uber, then we should remember that NU does still exist as a metagame, even if not too many people care about it. I mean, I for example, do not give one fuck about the NU metagame and care only slightly more about UU, but I also don't really care about the uber metagame that much either. Isn't it fair to say that we should strive to decentralize the standard metagame first and foremost? I think so.

Also my "argument" stems on the fact that there is indeed considered to be an NU metagame—if there weren't your concern would be well-placed.

I also think that the decentralisation method lends itself nicely to crafting a new standard by starting with all pokemon unbanned, and gradually choosing pokemon to ban and observing the effect on the number of viable pokemon, but we might need to think a bit about my first question (how to decide what to test).

We can decide what to test using intuition, a process for which Obi has attempted to lay the groundwork, but that seems to be the hardest and most-time consuming part of this process. We do need to try and start somewhere, though, and sooner rather than later, which is why I've decided to underline your points ("centralization only", "minimizing pokemon on ban list") and try to start to reach and establish a common ground so we can initiate whatever it is we need to initiate.

 

[Cathy]

Agree with latter as well, but can you explain why having standard as decentralized as possible would adversely affect the number of viable pokemon in any metagame?

When I said "any metagame" what I was referring to was any "balanced metagame". So if a pokemon is banned from standard it is necessarily viable in no balanced metagame.

Another way of looking at this is that the metagame without Deoxys-S and Wobbuffet and the metagame with them have approximately the same number of viable pokemon (as measured by the number of usages comprising the top 75% of the usages). So I would rather choose the metagame with the simpler rules (i.e. fewer bans), given that they both have approximately the same number of viable pokemon.

 

[Aldaron]

Before we get all muddled with the specifics of the "overcentralization" argument...can we even say that that argument is what we want to use when defining ubers?

Is objectifying the standard used to determine what is Uber or not even what is desired from the majority of battlers? I might support the "overcentralization" argument (though admittedly, I have recently been running away from it), but does everyone else, or at least a large majority of everyone else?

Before we get into this overcentralization argument at all, we should determine if we want to make the standard for determining Ubers objective, or if we want to keep it subjective. I'm not talking about counters, luck, or overcentralization at all here.

All I'm asking is whether or not it should be objective or subjective.

 

[Dragontamer]

In other games such as Magic the Gathering or YuGiOh the card game, the creators continuously churn out new cards with new techniques every few months. And thats what got me thinking: Perhaps some people simply want a changing metagame. Clearly, the TCG model shows that there are people out there who like the fact that some 50+ cards are banned every few months.

Anyway, we can take advantage of this situation. Some people like change, while many people want a stable OU metagame. And finally, we need someone to test these metagame before we make any additions to the OU metagame. So here's the proposal.

Every two weeks or so, we can create vastly different metagames on a Shoddy Official Test Server. We can ban or unban any pokemon without complaints... and at the same time we'll collect usage statistics, and any other objective or subjective information we think is useful.

The problem is of course: who is going to administrate it, etc. etc. But just tossing the idea out there.

 

[Cathy]

A separate server is not required (I am planning on just making multiple ladders or whatever you want to call them) and I already outlined a testing method above (that I will be setting into motion any day now) that is similar except that the metagames being tested aren't "vastly different".

I think your analogy is flawed though. In a trading card game, the set of cards to consider is actually changing. The set of all pokemon is not changing with time in Pokemon.

 

[TheMaskedNitpicker]

I've read over all of the arguments in this thread and the "Would banning Garchomp from standard play even solve the problem?" thread, and I'd like to weigh in with my take on the situation.

The crux of the argument here, as has already been established, is that different people have different priorities when creating a play environment. The two main goals are:

Goal A: Create an environment in which there are as few Pokémon banned as possible while maintaining at least X viable Pokémon (about 50 seems to be the going benchmark at present).

Goal B: Create an environment in which there are as many viable Pokémon as possible.

These two viewpoints have been outlined quite well in the "Would banning Garchomp..." thread by Colin and Hip, respectively. Back in late January, I actually created a poll to gauge which of these goals was the most popular. I'm not claiming the poll was accurate, since it only got 128 voters and I'm not sure how many of the voters understood what was being asked (as Hip fears), but I think we can draw a couple of tentative conclusions from it when combined with the discussion here in the Policy Review.

• Goal A is the majority stance. However...
• There is a significant segment of the population that prefers Goal B.

One problem, as I also outlined in that thread, is that our current system probably doesn't reflect Goal A and demonstrably doesn't reflect Goal B. Obi made a very good argument in the Stark version of "What is Uber?" that the unbanning of several Pokémon may not further centralize, and may actually decentralize, the current metagame, so it seems quite likely that Goal A isn't being met. The current UU metagame has arguably much less centralization than standard play, so we already have a counterexample for the assertion that Goal B is being met. What we have now is a system based on tradition stemming from previous generations of Pokémon games, intuition, and tests of a small number of individual Pokémon (Manaphy, Wobbuffet, etc.).

I am of the opinion that Goal A and Goal B are completely irreconcilable with the current pool of Pokémon. There is no good compromise between the two. So the way I see it, we have (at least) four options:

Option 1: Go with the majority view (Goal A) and test a 'standard' environment with a shorter ban list. Let BL, UU, and NU follow as per usual.
Option 2: Create two play environments, one for each goal, with no correlation between the ban lists for each.
Option 3: Leave things as they are.
Option 4: Leave the current tiers alone, but add the Goal B play environment, whose ban list has no connection to any existing tier.

Option 1 and 2 would both result in the same Goal A play environment. At this time I'd like to ask those who play Ubers how centralized the current Uber metagame is. Shoddy just implemented an Ubers ladder and I assume that statistics are being collected on weighted usage, etc., for this environment, so if enough people play Ubers that should give us some good data. It may be that the Uber metagame as it is satisfies our threshold for Goal A. Jibaku and Great Sage's guide to Uber battling, for example, already lists about 30 Pokémon that can perform well in Ubers, not including various Chlorophyll and Swift Swim Pokémon and others that didn't quite rate a specific mention.

If it turns out to be the case that we don't need to ban very many Pokémon to achieve Goal A, then the next tier 'down' might look similar for options 1 and 2, but would probably play very differently. Option 1 gives us the usual tiers where we automatically ban everything in standard play from UU and then remove a bunch of other Pokémon that would unbalance the remaining pool, relegating them to BL. Option 2 would involve a set of bans that have nothing to do with what's popular in the higher, 'standard' tier, but would be put in place in order to maximize the number of usable Pokémon. I prefer Option 2 (or 4) because I like a very decentralized game (Goal B), but I don't want to enforce that view on those that prefer Goal A.

Of course, both Options 1 and 2 have the two disadvantages of Whining and Work. If drastic changes are made to the tiers, there may be a significant backlash against Smogon by the masses. People in general dislike change. The other problem is one of work, meaning that all the analyses would have to be revised as well as some of the articles. I personally think that the eventual improvements to the game will outweigh the extra work, but I'm not the one who put in all the hard work to write the current analyses, so I don't think my opinion regarding this issue should count for very much.

Option 4 is a bit of a compromise that would keep Goal B people (like Hip and I) happy but has disadvantages of its own. For one thing, we start to get overlap between 'tiers', which makes the new environment not really a tier at all. Secondly, there may or may not be enough people that want to try this new environment, although I might be proven wrong and it doesn't really hurt to test it for a period of time. The other disadvantage is that we'd have to decide on a starting point for this tier that would be based mostly on intuition (backed up by experience, of course). Agreeing on this might be difficult, but that remains to be seen.

So, here are the questions I have:

• Are we willing to go through the work and put up with the whining required to change the current standard environment in order to shorten the ban list?
• Are we willing to create a separate environment for the people that do want a highly decentralized game?

 

[gengod]

Option 1 and 2 would both result in the same Goal A play environment. At this time I'd like to ask those who play Ubers how centralized the current Uber metagame is. Shoddy just implemented an Ubers ladder and I assume that statistics are being collected on weighted usage, etc., for this environment, so if enough people play Ubers that should give us some good data. It may be that the Uber metagame as it is satisfies our threshold for Goal A. Jibaku and Great Sage's guide to Uber battling, for example, already lists about 30 Pokémon that can perform well in Ubers, not including various Chlorophyll and Swift Swim Pokémon and others that didn't quite rate a specific mention.

I currently play Ubers very often so i will just give some insight on the Uber metagame's status. Currently the Uber metagame is in my opnion centered around 3 pokemon. Kyogre (not rain teams to be exact more so than different Kyogre variants the most common are Rest Talk with Calm Mind, Scarf and Specs in that order), Dialga variants (Bulk Up Dialga and Mix Dialga also in that order) and Calm Mind Lati@s (with Soul Dew obviously). On almost any uber team your going to see at least one to two of these if not all three. Obviously I can't give hard numbers or speak for everyone but we will see all of that on Shoddy within the upcoming months.

 

[Yagami Light]

I'd also like to add that Rayquaza is pretty much the Ubers Dragonite/Salamence/Garchomp , and as such you see a whole ton of him. It's quite common.

 

[CardsOfTheHeart]

After Obi's arguments in the Wobbuffet "vote" thread, I decided to look this one over again. I can't remember if this thread started before or after I got my badge, but at the time I didn't have much of a quality opinion on what was Uber. However, Jump's Wobbuffet thread jarred something loose in my brain and allowed me to form an opinion on what was Uber, which was the following:

Here is Smogon's current official stance on Wobbuffet:

Wobbuffet's May 30, 2008 movement from the Limbo Tier to the OU tier is a direct result of no conclusive evidence of it overwhelming the standard metagame on Shoddy Battle's Ladder arising after three months of testing. Its effects on the standard metagame will continue to be closely monitored, and it will be as tiered as fairly as possible as the metagame evolves, just as any of the other 492 Pokémon.

The impression that I am getting from that statement is that Wobbuffet has not been an overcentralizing figure in the Shoddy metagame--that statement few will argue with--thus it is not Uber. One problem: have we even defined what is and is not Uber?

We can make a logical explanation for Pokemon going from Uber to OU; we have done it with Wobbuffet and Deoxys-S. 'If it doesn't overcentralize OU, it shouldn't be banned from OU.' That, I believe, is our explanation. However, we haven't been able to do the same for Pokemon going from OU to Uber. Therefore, until we establish that definition, we just go with 'If it doesn't overcentralize OU, it shouldn't be banned from OU.'

The problem I have with that: is that the BEST reasoning? Decisions about whether certain Pokemon need to be banned are subjective in nature. No matter what we use to help us make those decisions, the decisions themselves are ultimately subjective.

As far as I'm concerned in making those decisions, objective evidence should only support the subjective decisions, not actually make the decisions themselves. That's why the Wobbuffet reasoning seems "backwards" to me. We seem to be looking at the objective evidence to make a subjective decision.

As for the subjective decision to be made in the case of Wobbuffet, is overcentralization even the right question to ask??? If we, as a community, are trying to develop a metagame that "promotes skill," shouldn't the first question to ask be if the Pokemon in question actually promotes skill?

From what I understand (since I did not play Pokemon competitively in RBY), evasion moves such as Double Team and Minimize were originally banned because they promote luck instead of skill. As far as I know, we didn't need objective evidence to tell us that.

The Wobbuffet argument I am seeing most often is that it does not promote skill. Many people are making this argument, including several that have used Wobbuffet to great effect on Shoddy. We are seeing logs that seem to support this argument. From what I am seeing, the way to use Wobbuffet can be outlined in a few simple steps almost regardless of the situation. Does that really promote player skill?

I don't think it does. I would want Wobbuffet banned for that reason.

...

...

Since the discussion shouldn't all be about Wobbuffet, I should go into the case of Garchomp, shouldn't I?

I defined above two subjective decisions that I believe we should make in determining whether something should be banned from OU:

1. Does it promote skill? If not, ban it.
2. Does it overcentralize OU? If so, ban it.

First, does Garchomp promote skill? Well, I can't claim to have actually battled in the last few months, but I don't reasonably see how it can just Swords Dance + Outrage through absolutely everything. I believe it takes skill to use so I'll answer that question "Yes." Next question.

Does Garchomp overcentralize OU? I suppose a certain strategy is necessary to defeat it as TAY has been the latest to elude to, but the current objective evidence doesn't seem to show any overcentralization. I'll answer that question "No."

However, I would support a temporary removal of Garchomp from ladder play to see if Garchomp would be an overcentralizing figure. I like to use, as a general rule, the premise that if the removal of a Pokemon from a metagame increases the number of Pokemon in the metagame OR the addition of a Pokemon in a metagame decreases the number of Pokemon in the metagame by a material amount, then the Pokemon is an overcentralizing figure. I also like to use a 10% figure for materiality in this rule. In other words, if the number of Pokemon changed by at least 10% as a result of the addition or removal of a Pokemon, then that Pokemon is overcentralizing.



Okay, I'm starting to ramble like I usually do on these topics. I'll stop now.

Don't expect me to come up with a good-enough overcentralization definition, though, as the one I posted here is very arbitrary. I'm not so sure if I would seriously use it.

 

[imperfectluck]

Along with the creation of the maximum # of useable Pokemon, I'm also interested in the maximum # of viable team makeups/strategies.

 

[Tangerine]

The only reason we should ban something is because it's uncompetitive. I'm assuming this is what "uber" is, not something that's simply "too powerful" or "too centralizing" but just anything that takes away from the nature of the game and makes it less competitive in that aspect.

There are two things that I believe that has been brought up that promote uncompetitive play thus far. This is of course, "promoting guessing" and "being overpowered"

Pokemon is a game of managing risks, something I have probably repeated a million times in the past few days. A player gets information as they play, and make decisions based on their information. It is a battle of information - if it wasn't team scouting wouldn't be such a huge issue in tournaments. With this we can define "promoting guessing" as something that strips this aspect of the game.

This can happen in some way. Suppose that GameFreak makes a weather condition such that we can't see the opponent's Pokemon that you are battling against. Would you still play this game with that weather condition on, or would you ban that weather condition because it is uncompetitive?

This definition is why I'm against Wobbuffet - as soon as it enters, it turns the game into a guessing game considering as soon as it enters most of what you know about your opponent's team doesn't matter since they'll get one free turn out of it (or kill your only counter to the Pokemon). Either way you're in a lose lose situation. If we had a way to measure what would happen with Wobbuffet and be able to do something about without overpreparing for it(hint: you can't really do something about it without overpreparing for it but that's retarded in it's own right) then I wouldn't care about Wobbuffet being in OU.

The definition also applies to Deoxys E. Deoxys E's presence means attempting to "play around it" without a dedicated counters quickly means doom unlike every other OU Pokemon I have faced because of it's speed and it's movepool. Once you start dancing with it you realize that you have to continue to switch and hope that they don't predict and maybe you'll predict that and stay in but they might have read that and decided to play it safe etc etc until you realize that it comes down to pure dumb luck whether or not it tries to overpredict or not and you have to hope that it doesnt if you want to take it down to enough health so you can kill it with a priority move.

The second definition of 'being overpowered', I believe is best defined by my trends definition.

If we define a metagame as something centralized, we realize that it centralizes with trends. Trends are sets of Pokemon used for a specific strategy. Species Clause and the 6 Pokemon Limit naturally limit the number of trends that can be used in one team.

We call a Pokemon overpowered, if despite the countertrends developed in order to stop it, the usage continues to rise. We can connect the dots here and see that this situation means that this pokemon is counterable or people are trying to counter it but it's so good that people will use it anyway. Needless to say Garchomp falls under this definition, and I have no doubt when Deoxys E reaches it's full potential this will also happen. This definition also accounts for fully "unpredictable" pokemon that a standard team can't deal with (unlike Gengar, Lucario, TTar, etc) no matter how much they try, so I believe that this is a very good definition since it ignores the "no counters" argument and gives us substance behind that argument.

These are what I believe what make Pokemon Uber - I do not care for "overcentralizing" and what it might mean in favor of a more balanced metagame (although overcentralizing is probably a side effect of "being overpowered" but that's not always logically the case so I'll leave it here)

If I made anything unclear feel free to ask, I'm more than willing to clarify and I would really love comments/criticism on my theory, especially since there's a lot of subjectivity left to be dealt with (I'm counting on you Obi ) Yes, I know that I used statistics in my theory but this is how I believe they should be used since I believe my theory does a good job reflecting the current state of the metagame.

 

[Ancien Régime]

We call a Pokemon overpowered, if despite the countertrends developed in order to stop it, the usage continues to rise

I agree with this. Basically, the crux of the argument is that "a large amount of people will still use pokemon x in their teams, despite the fact that every good team is prepared for it to the maximum possible degree."

What countertrend (barring the unbanning of ubers that potentially counter it) can possibly exist for Kyogre (to name an example). Maybe heavy Shedinja/Ludicolo usage, but Kyogre itself has ways of getting through those counters (maybe put HP Flying on its moveslot?) or the rest of the team can work to eliminate those highly specialized counters.