Portrait of an Uber

[Jumpman16]

This topic will serve to finally give an accurate enough definition or at least appraisal of what makes an uber an uber. At once the notion of "uber" is lofty and vague, seemingly as timeless as the Mona Lisa herself. The picture may indeed be worth a thousand words instead of a few dozen, but in the interests of having an accurate appraisal for the sake of competitive battle, and not just to impress some artsy-fartsy scene chick with your own reasoning behind why Mona Lisa is or is not smiling, a definition should be as clear and succinct as possible.

As with art, we shouldn't kid ourselves and believe that a hard-and-fast definition of "uber" exists, one that everyone versed well enough in art will doubtless agree upon. As soon as we accept this, we can safely shoot for the sky and end up somewhere that is good enough for a definition of uber to be understood and accepted by the majority of people who understand competitive pokemon.

So without further ado, I think we should start to list the characteristics of a pokemon that make it undeniably uber. I first brought this up in #insidescoop with the intention of continuing the discussion here, so I will repaste a few things.

[23:57] <@Jumpman16> well we actually could start by listing the quailties that make Rayquaza so obviously uber
[23:57] <@Jumpman16> assuming we all think it is
[23:57] <+Obi> Yeah
[23:57] <@Jumpman16> then see how many of those apply to other obvious ubers

Two such quailties began to be chiseled away at, removing the fluff in the interests of succinctness.

[00:02] <+Obi> I think it has to do with having a decent answer to its common set

[00:03] <@Jumpman16> also "common" is a good thing to remember
[00:03] <@Jumpman16> Sand Stream is common
[00:03] <+Obi> Well
[00:03] <+Obi> I didn't mean it that way
[00:03] <@Jumpman16> and make Moonlight Cress not viable
[00:04] <+Obi> I meant it in the sense of "The only counter to SD Chomp is Starmie. Starmie cannot switch into Scarf or CB. If I let SD Chomp get a free turn, I lose." somethihng like that

[00:04] <@Jumpman16> i think "common" needs to be remembered in whatever definition we're trying to come up with
[00:04] <@Jumpman16> what about yache, lol
[00:04] <+Obi> I was just using that as an example
[00:04] <@Jumpman16> i know
[00:04] <+Obi> =p

[00:04] <@Jumpman16> that's actually the point though
[00:04] <@Jumpman16> without yache a lot less people think its uber
[00:05] <+Earthworm> what are the most common circumstances in which you face the most common garchomp?
[00:05] <+JabbaTheGriffin> sandstream, sr, sd yache chomp
[00:05] <+IggyBot> Yache Berry
[00:05] <+IggyBot> Stealth Rock

[00:05] <@Jumpman16> at least one of SR, Sandstream, Spikes, T Spikes
[00:06] <+IggyBot> Stealth Rock is almost a guarentee in every match
[00:06] <+JabbaTheGriffin> it didn't used to be :/
[00:06] <+JabbaTheGriffin> stupid suicide sr leads
[00:06] <+IggyBot> lol
[00:07] <+JabbaTheGriffin> but in that kind of situation what makes a pokemon uber?
[00:07] <@Jumpman16> at once the idea of "if the pokemon does not have to greatly hinder itself to dispel would-be counters to its common set sweeping, it is uber" hits me
[00:07] <+JabbaTheGriffin> because in that scenario no pokemon can switch into chomp and win 100% of the time
[00:08] <+JabbaTheGriffin> hmmm


[00:23] <@Jumpman16> maybe if we append "how easily a pokemon a can sweep a team without risk to itself or its team"

Right now, I would further modify this "uber quality" by saying something like:

How easily a pokemon a can sweep a team in common battle conditions with little to no risk to itself or its team.

This rules out, in getting at the definition of uber pokemon like Gyarados and Salamence, which have the ever-so-common Stealth Rock with which to contend, and pokemon like Dugtrio and Slaking, which, while being the very embodiment of "easy" and "power", bring risk to their teammates due to their respective nature (Duggy needs a CB to "sweep" at all, and Slaking's Truant Ability risks the safety of both it and its teammates, despite Slaking's own power).

"Easily" is subjective and instantly problematic, but we can chip away at this and get a much better definition.

[23:58] <@Jumpman16> to start, ray is "capable of sweeping standard with one turn of set up"
[23:58] <+Earthworm> which?
[23:59] <+Obi> I don't know what can switch into Rayquaza
[23:59] <@Jumpman16> remember, this is in regard to standard, not uber

[23:59] <+JabbaTheGriffin> well
[23:59] <+Obi> But Kyogre has a good set of switch-ins
[23:59] <+JabbaTheGriffin> groudon and lugia can switch in to ddray
[23:59] <+JabbaTheGriffin> any other version and they're fucked
[23:59] <@Jumpman16> we have to remember that this is in terms of OU pokemon
[23:59] <@Jumpman16> not ubers
[23:59] <+Obi> Rayquaza is just a beast
[23:59] <@Jumpman16> ray rapes standard and is uber beacuse _______
[23:59] <+Obi> One Dragon Dance and nearly everything is dead

You get the idea. To start at another characteristic:

Capable of sweeping standard in common battle conditions with one turn of set up or less.

Later, there was:

[00:20] <@Jumpman16> also, if it does indeed have both offensive and defensive uber qualities, that's all the more for it being uber
[00:21] <@Jumpman16> so "borderline ubers" like mew and manaphy would have those added in their "favor"

which we can get to in time. Right now I'm going to go ahead and open the floor for other "characteristics of an uber" and for people to further chisel away at. I really feel that a list with enough succinct characteristics will aid us in determining what is obviously uber and what isn't, even if we'll never come to a perfect definition so we can go forward with our many Suspect tests. Because hey, at the end of the day, it doesn't matter whether or why the Mona Lisa is smiling—what does matter is whether you were able to get into that scene chick's pants (if you're into that kind of girl).


edit: the fruits of our labor should be on the first page:


Offensive Characteristic
A Pokémon is uber if, in common battle conditions, it is capable of sweeping through a significant portion of teams in the metagame with little effort.

Defensive Characteristic
A Pokémon is uber if, in common battle conditions, it is able to wall and stall out a significant portion of the metagame.

Support Characteristic
A Pokémon is uber if, in common battle conditions, it can consistently set up a situation in which it makes it substantially easier for other pokemon to sweep.

 

[Syberia]

I'm not sure Rayquaza is the pokemon you want to use as your example of something that is clearly uber, as people can always make the argument that you can sacrifice a pokemon to it, or get your Weavile/Mamoswine in against a set-up move, and revenge kill it with Ice Shard, much as can be done with Garchomp. I'm not saying I agree with this argument, nor find it viable, I'm just saying that with your choice of pokemon, the argument can be made.

Perhaps what I consider the biggest argument to be made in favor of a specific pokemon being uber is that, despite knowing exactly what it's going to do, it can still do massive amounts of damage to a team before going down. Take, for example, Swords Dance/Earthquake/Stone Edge/Fire Blast (maybe Fire Punch, if OHKOing Celebi after a Swords Dance interests you) Groudon. In OU, that's a very predictable set for it to be running, but still damn near uncounterable unless you want to sacrifice two to three pokemon to bring it down.

The same can be said, to a slightly lesser degree, about Garchomp. At least 75% of the Garchomps I've seen on Shoddy have been of the Swords Dance/Yache Berry variety, however, despite that, two to three pokemon are still required to beat one very predictable one that is seen everywhere.

Things like Gengar or Salamence or Lucario that are said to "have no counters" do not work in the same way as my Groudon example above. They have no counters because of their diverse movepool, but to defeat every threat in the metagame would require more than four moveslots on any one of these pokemon. Lucario could use HP Ice to take out Gliscor, but then it would lose either Crunch or Stone Edge, opening itself up to other counters.

 

[Jumpman16]

I think Ray's a pretty good example:

Adamant, 438 Attack Rayquaza's Life Orb, once–Swords Danced Extremespeed on a 361HP/196Def (min/min) Mamoswine: 92-108%

Further, remember that "in common battle conditions" Mamo's HP will likely not be above 87.5% HP, so while the analysis doesn't suggest no HP EVs there's a pretty big chance that Rayquaza is OHKOing Mamoswine unless it runs Jolly and max Speed (which isn't in the analysis either, to be fair). While "it doesn't OHKO a 4HP/min Ray guaranteed with 359 Attack Ice Shard without at least an EB" cancels out with "in common battle condition LO Ray won't be at 100% either", it says something that there are pretty much only two pokemon that can safely try to "revenge kill" LO Ray and one of them has to start to overspecialize to do so, straying immediately from "common" itself.

Most important though, we realize that what makes people think Garchomp is uber now pretty much solidifies why Ray would still rape standard in spite of Weavile and Mamo: Yache Berry. With very similar defenses Yache Ray would take pretty much all unSTAB ice attacks just like Garchomp does, and combats Chomp "you have a 80% of even touching me anyway" in Sandstream with a "you have a 0% chance of even touching me" if they do not have at least 290 speed or the bulkiness to take a +2 Dragon Claw/EQ/Outrage (or 302 SpA Overheat) if slower.

This is pretty much limited to HP Ice Gengar because it's immune to ES (but still may die because of Yache) and Bronzong because it may not die to Overheat or +2 Atk Outrage, though 54-64% and 62-74% is done respectively from an Adamant 438Atk/302SpA (max/min) Yache Rayquaza to the "standard" 338HP/268Def/330SpD Bronzong, which is pretty lol if Ray actually has a LO. For completeness LO Ray does 81-95% to the standard Bronzong with a +2 Outrage and 71-84% with Overheat.

So yeah, I think Rayquaza fits the Portrait of an Uber well enough, theorymon sure is fun sometimes. I agree with the rest of your post and encourage more from others.

 

[Syberia]

Yeah, I forgot about Extremespeed. Though the point of mentioning that in the first place was that the kind of logic being used there is bad logic anyways.

 

[X-Act]

The reason why I haven't posted here is because I literally didn't have anything to add... until now.

I think how a Pokemon sweeps is really the only thing that makes a Pokemon uber. Even Giratina would be uber because it would sweep everything in Standard. Its offensive capabilities are only a 'modest' 100 Atk, 100 SpA and 90 Spe, but when combined with 150/120/120 gastronomic defenses, Giratina would sweep in Standard as there's nothing that would be able to inflict enough damage to Giratina to stop its damage dealing. Imagine it like a buffed-up version of GSC Snorlax.

This suggests that a measure of how good a Pokemon is is the ratio of the amount of damage it can deal with the amount of damage it can withstand, and, indeed I used this definition of rating a Pokemon in my Stats Ratings applet. My applet only takes base stats into account, but surely the definition in bold would still apply even if we take everything into account. Unfortunately I haven't been able to do this mathematically (and hence, objectively) yet - and I'll probably never be able to do it completely - but that doesn't remove anything out of the definition.

 

[makiri]

While I agree with the assessments of "Capable of sweeping standard in common battle conditions with one turn of set up or less" and "How easily a pokemon a can sweep a team in common battle conditions with little to no risk to itself or its team." I really think they need to be slightly amended to include Pokemon like Lugia and Deoxys-D, they have almost 0 sweeping potential after 1 turn of a Calm Mind, but because they are suck fantastic walls, they are able to max out their Calm Minds realitively easy in any enviroment. I have to agree with X-Act here because of the fact Pokemon like Lugia, Deoxys-D, and Giratina exist in the Uber tier currently and show no signs of being removed, these Pokemon can't really be OHKOed by any reasonable attack, can heal themselves easily, and hit back when needed. Or they can go another route and just be the best walls in the game, Lugia can spread status, phaze, set up screens; Deoxys can Spike, spread status; and do this with little to no risk to themselves.

 

[Syberia]

Such a definition would have the unpleasant side-effect of allowing Wobbffet back into OU on a technicality, however, unless we expand it to include the amount of damage that is also caused by whatever it alllows to set up.

I think that the only "real" definition of uber that is required is that the pokemon are too strong for OU. Each pokemon should then be evaluated on a case-by-case basis, as having a hard definition of uber can really do nothing but cause petty debates over technicalities. If "overcentralization" is the only definition, then people will endlessly debate the merits of statistics. If "uncounterable" is the only definition, then people will endlessly debate just what it means to be a counter, and certain things will most likely slip through the cracks of whatever definition is inevitably chosen. By choosing a simple definition of "too strong for OU," without any strings attached, it allows each pokemon to be evaluated individually on all of its merits.

 

[DougJustDoug]

I agree with X-Act's line of thought that the power of a pokemon is defined by "the ratio of the amount of damage it can deal with the amount of damage it can withstand". There are soft issues behind this that cannot be easily expressed objectively, but should be taken into account.

"The amount of damage it can deal" is tied not only to the pokemon's base stats and base move power, but the ability to boost that damage dealing capability. Obviously, a pokemon like Rayquaza, who sports a majestic 150 base attack, becomes a damage dealing machine after a single Dragon Dance. DD also has the advantage of increasing speed, which is commonly considered an offensive trait. The same can be said for Garchomp, who starts with a frightening 130 base Atk and the ability to double it in a single turn with Swords Dance. However, damage dealing boosts are not just limited to buffs like DD and SD; it also applies to Kyogre's access to perma-rain. Boosted water STAB, plus risk-free Thunder is gamebreaking for a pokemon with 150 base SpAtk.

"The amount of damage it can withstand" has more soft issues than the offensive issues mentioned above. Obviously, defense stats and type resistances are the primary factors in a pokemon's ability to withstand damage. But, the impact of Stealth Rock and Sandstream on a pokemon is a huge defensive consideration.

Why is Garchomp so much more deadly than Gyarados? Both sport huge stats, a single common 4x weakness, and access to a big offensive boosting move. Well there are several reasons that Gyar is not mentioned in the same breath with Chomp -- but two significant reasons are Gyar's weakness to SR and lack of SS immunity. If Gyar can be forced from play, is has a massive liability to carry upon returning to the field, considering the likelihood that rocks will be up and it will eat 25% upon switching in. If Sandstream is also in effect, then Gyar is at a huge disadvantage. Chomp is immune to SS, and resists Stealth Rock. These two factors are a big boost to Garchomp's defensiveness in the current metagame.

And, not to belabor Garchomp's many advantages, but Chomp also has access to Sand Veil, which is yet another soft factor that contributes to it's defensive capability or "amount of damage it can withstand". Other pokemon with abilities that allow them to avoid damage, should be factored into their overall defensiveness. Stuff like Hydration and Poison Heal are easily recognized; but Togekiss' Serene Grace combined with Air Slash, should be considered a huge defensive boost as well.

So, I think a comprehensive comparison of "the amount of damage it can deal with the amount of damage it can withstand" should be a factor in classifying a pokemon as uber. Such a comparison should not be made solely on hard stats and move power, but factoring in the reasonable soft issues that also contribute the pokemon's offensive and/or defensive prowess in the current competitive metagame.

 

[Brawley]

I think we should use Syberia's description as if we use X-act's and the one in the OP as they don't apply to wobbufect and deoxys-d, also doug garchomp has 130 base attack.

 

[Cathy]

Note that it will take several iterations to implement this portrait. Starting from all pokemon being legal (except of course Arceus since it isn't a pokemon), we follow these steps:

1. We ban the pokemon with too high of the ratio described by X-Act.
2. We let this new environment play out for a bit.
3. Return to 1.

Seeing as the initial environment is the same the "ubers metagame" we already know the initial set of bans, but our starting point should be with just those few banned, not the current environment.

Also... I'm experimenting a bit with a way to quantify this ratio in real battle conditions. If this is successful it could be a new and interesting way to objectively decide what is uber.

(Brawley is right that this definition will not make Wobbfufet uber.)

 

[Jumpman16]

Yeah, I forgot about Extremespeed. Though the point of mentioning that in the first place was that the kind of logic being used there is bad logic anyways.

That's fine, because it is our responsibility to both hammer down the proper logic to be using when considering what is uber and toss out any bad logic used when we arrive at our best definition. I'll come back to the notion of "definition", though.

The reason why I haven't posted here is because I literally didn't have anything to add... until now.

I think how a Pokemon sweeps is really the only thing that makes a Pokemon uber. Even Giratina would be uber because it would sweep everything in Standard. Its offensive capabilities are only a 'modest' 100 Atk, 100 SpA and 90 Spe, but when combined with 150/120/120 gastronomic defenses, Giratina would sweep in Standard as there's nothing that would be able to inflict enough damage to Giratina to stop its damage dealing. Imagine it like a buffed-up version of GSC Snorlax.

This suggests that a measure of how good a Pokemon is is the ratio of the amount of damage it can deal with the amount of damage it can withstand, and, indeed I used this definition of rating a Pokemon in my Stats Ratings applet. My applet only takes base stats into account, but surely the definition in bold would still apply even if we take everything into account. Unfortunately I haven't been able to do this mathematically (and hence, objectively) yet - and I'll probably never be able to do it completely - but that doesn't remove anything out of the definition.

I really like this criterion because it actually touches upon the defensive uber characteristics while speaking to the offensive ones as well. If we append the "in common battle conditions" that will eventually be a constant in this discussion to this definition it gets even better. Even just taking base stats into account and using Slaking as the quintessential example of a BL/OU pokemon (actual, accurate tiering) in an uber pokemon's shoes (stats in this case), we realize that "the amount of damage it can withstand" is lower than that of many of its BL/OU counterparts because of its Truant Ability.

And actually, even "the amount of damage it can deal" is not as high as you may be inclined to think even with a (given) Choice Band, because as I realized/discovered months ago when redoing the Threat List, the fact that it can't immediately deal the finishing blow to pokemon like Bronzong and Skarmory the turn after it uses Fire Punch even if it guesses right (because a CBed Fire Punch still doesn't OHKO either, and without a CB "the amount of damage" is automatically 33% less even with a correct guess) is pretty much the reason is isn't OU, let alone uber. As far as this thread's goal—having the biggest majority of people agree with the uberness of a pokemon—is concerned, I think that "Slaking with Inner Focus or Thick Fat" or any other Ability that doesn't necessarily make Slaking that much better outside of the comparison to Truant (or more accurately, to use a definition just suggested that we will probably agree upon, "that doesn't increase the amount of damage it can deal or reduce the amount of damage it can withstand in comparison to Abilities like Pure Power and Wonder Guard"), would be overwhelmingly perceived as uber. That may seem like a long-winded, needless tangent but I think it serves to show how comprehensive the ratio of the amount of damage it can deal with the amount of damage it can withstand really is, even with pokemon that we'd already agree are correctly tiered right now.

Both sides of the ratio, therefore, in Slaking's case are negatively impacted by Truant. While we want to be wary of creating definitions based solely on our experience (a posteriori) because we don't have experience in the standard metagame with most of the current suspects (and the Lati@s tourney and the brief Manaphy stint were not nearly comprehensive enough), I don't see much of any harm in looking at the pokemon who are obviously uber and obviously standard in helping us to define what is uber. This is why I brought up Rayquaza as an obvious uber, and why it will help us to examine why pokemon like Gyarados are not uber. We don't want to mold our definitions just to keep things like Wobbuffet in uber or Deoxys-S in standard, though, as that reeks of subjectivity and unfairness.

I have to agree with X-Act here because of the fact Pokemon like Lugia, Deoxys-D, and Giratina exist in the Uber tier currently and show no signs of being removed, these Pokemon can't really be OHKOed by any reasonable attack, can heal themselves easily, and hit back when needed. Or they can go another route and just be the best walls in the game, Lugia can spread status, phaze, set up screens; Deoxys can Spike, spread status; and do this with little to no risk to themselves.

I think the ratio X-Act brought up addresses this enough, and besides that we can hammer down the primarily defensive characteristics in due time.

Such a definition would have the unpleasant side-effect of allowing Wobbffet back into OU on a technicality, however, unless we expand it to include the amount of damage that is also caused by whatever it alllows to set up.

I think that the only "real" definition of uber that is required is that the pokemon are too strong for OU. Each pokemon should then be evaluated on a case-by-case basis, as having a hard definition of uber can really do nothing but cause petty debates over technicalities. If "overcentralization" is the only definition, then people will endlessly debate the merits of statistics. If "uncounterable" is the only definition, then people will endlessly debate just what it means to be a counter, and certain things will most likely slip through the cracks of whatever definition is inevitably chosen. By choosing a simple definition of "too strong for OU," without any strings attached, it allows each pokemon to be evaluated individually on all of its merits.

The reason we can't do this was originally raised in #is: "too strong" is really every bit as subjective as "uncounterable" and "overcentralization". I wish it could be that simple, but Garchomp has proven that it really isn't or else it would have been obviously uber months ago. Yache Berry may have been the cherry on the top of the uber cake, but if all it takes is a damage-reducing berry to make something uber then we aren't thinking about this the right way. Again, this topic serves to arrive at the best, most accurate definition of uber, which is why we are trying to hammer down the characteristics that make a pokemon obviously uber. It's plural for a reason—there are many such characteristics and we should do our best to hammer down the best ones so we can all go forward with them for months and years to come. Unfortuntely, "too strong" isn't accurate or objective enough to reach our goal.

So, I think a comprehensive comparison of "the amount of damage it can deal with the amount of damage it can withstand" should be a factor in classifying a pokemon as uber. Such a comparison should not be made solely on hard stats and move power, but factoring in the reasonable soft issues that also contribute the pokemon's offensive and/or defensive prowess in the current competitive metagame.

Perfect. There are quite a few "soft issues" that must be considered, which isn't too frightening a notion when we realize that the "boosts" of rain and sun and the impact of field hazards like SR and Sand Stream are still pretty objective. The latter issue is somewhat addressed by what I feel is an objective "in common battle conditions", and I am as inclined to believe that Gyarados would more closely skirt the line between OU and Uber if it were affected by SR and sand the same way Garchomp is (consider Wacan Berry Gyarados and that Garchomp takes Thunderbolts better than Garchomp takes Ice Beams if it's difficult for you to see why a case for "Gyara = uber?" could be made).

I think we should use Syberia's description as if we use X-act's and the one in the OP as they don't apply to wobbufect and deoxys-d, also doug garchomp has 130 base attack.

Again, while we don't really want to mold our characteristics with specific pokemon in mind, we are attempting to hammer down characteristics in the plural rather than trying to make a catch-all definition right off the bat. I would ask everyone to please understand this when thinking of posting in this topic (and besides, "too strong" leaves us in no-mans land with regard to Garchomp, as I've stated).

Note that it will take several iterations to implement this portrait. Starting from all pokemon being legal (except of course Arceus since it isn't a pokemon), we follow these steps:

1. We ban the pokemon with too high of the ratio described by X-Act.
2. We let this new environment play out for a bit.
3. Return to 1.

Seeing as the initial environment is the same the "ubers metagame" we already know the initial set of bans, but our starting point should be with just those few banned, not the current environment.

Also... I'm experimenting a bit with a way to quantify this ratio in real battle conditions. If this is successful it could be a new and interesting way to objectively decide what is uber.

(Brawley is right that this definition will not make Wobbfufet uber.)

Thanks for posting in this thread first off, as for obvious reasons you are not obligated to. However, Smogon's plan with regard to handling "suspects" (pokemon currently in the uber tier that may not be accurately tiered ther or vice versa) is outlined in my Order of Operations thread. Either method will indeed take several iterations, but we would be closer to the "true metagame" by handling the pokemon one suspect at a time in the metagame we feel is most close to the "true" one, which is a metagame without any current suspects. As stated in that thread it will difficult if at all possible to determine the culprit(s) in a metagame with all the current suspects (Latios, Latias, Mew, Manaphy, Darkrai and Garchomp) that would result after such a banishment of pokemon with "too high a ratio", simply because, as I stated earlier in this post, we have not adequately tested many of the current suspects in the standard metagame so we literally don't know their capabilities in standard.

Anyway, we've made some progress here, and later I'll attempt to sum up the best characteristics offered so far in the interests of us chiseling them down to make them as accurate as possible.

 

[X-Act]

I'd tweak my own definition slightly, to make it clearer:

A Pokemon's power is the ratio of the amount of damage it deals on average by the amount of damage it is dealt in common battling conditions.

'Amount of damage it can withstand' was basically the opposite of 'Amount of damage it is dealt'. I actually meant the latter. The following is the reason why:

Arceus is an exaggerated example of a Pokemon that deals a lot of damage and takes little damage - so the ratio for it would be very high (high number divided by low number).

Magikarp would be an exaggerated example of a Pokemon that deals very little damage and takes a very high amount - so the ratio for it would be very low (low number divided by high number).

Then there are Pokemon whose ratio is near 1 (i.e. they deal roughly the same amount of damage they are dealt). For example, Alakazam deals very high damage but is dealt a lot of damage as well, whereas Bronzong deals little damage but is dealt a low amount of damage too. Both of these Pokemon's ratios would be similar, even though they are obviously very different Pokemon.

 

[jrrrrrrr]

A Pokemon's power is the ratio of the amount of damage it deals on average by the amount of damage it is dealt in common battling conditions.

I was going to make a really long post in this thread but stupid X-Act decided that he had to sum up everything I was going to say in one sentence. I really support this idea, although quantifying it could be kind of tricky. With things like Lugia, who is "obviously uber" but I will argue against it just to raise a point, we would need to quantify how high a ratio of damage dealt:damage taken is "too high". Lugia has those insane 106/130/154 defenses, but it has bad STAB and only has 90/90 Atk stats (lower than Deoxys-Speed). In OU it may not actually be able to damage that much even though it could take a beating. This quote from X-Act outlines the only potential flaw I could see with that system:

Then there are Pokemon whose ratio is near 1 (i.e. they deal roughly the same amount of damage they are dealt). For example, Alakazam deals very high damage but is dealt a lot of damage as well, whereas Bronzong deals little damage but is dealt a low amount of damage too. Both of these Pokemon's ratios would be similar, even though they are obviously very different Pokemon.

Lugia is in a similar boat. Sure, it would take absolutely no damage from nearly everything in OU, but it really can't hurt that much either. How much do Lugia's (or uber suspect pokemon X's) defenses make up for its lack of offensive prowess? How weak would a pokemon X's defenses have to be to warrant its sky-high atk stats meaningless? This is a really solid framework, but as I mentioned earlier we are going to have difficulty quanitfying it (at least on the offensive side of the spectrum).

*I am also taking Speed into account as both an offensive and defensive stat. I would say that Deoxys-Speed has amazing defenses simply because it can't be hit most of the time and that nullifies any problems with its base HP.

 

[Hipmonlee]

Make sure you consider the possibility of a pokemon that doesnt do damage itself, but allows other pokemon to do a lot of damage.

Like a level 5 Kyogre, a BP Mew, those reflect/lightscreen deoxys or Wobbuffet.

Have a nice day.

 

[Brawley]

Jrrrrrrrrrr, I'm pretty sure Lugia with 252 Hp/252 Defense/4 special defense with calm mind, psychic, Hidden Power Fighting/Ice Beam, Roost would be able to take shit damage and stat up very easily basically like a way better suicune, who is annoying to deal with as it is.

 

[jrrrrrrr]

Yes but it would still lose to Toxic, Leech Seed, take damage upon switching in from SR, etc. Not to mention, if you are running that set then you lose a significant amount of your defensive potential because now you aren't using Reflect, Whirlwind, Toxic or Pressure to your advantage. And you STILL have to deal with bad STAB and bad atk stats. 90 base SpAtk isnt going to be doing much until you get a couple of boosts.

Obviously Lugia was an extreme example, I was just using it to illustrate the point of "how much defense is too much for a lack of offense?" and "how much offense is too much for a lack of defense?". Even if Lugia proved to be offensively too powerful for OU as well, my main point/question still stands.

 

[Hipmonlee]

What concerns me about such a system for deciding uberness, is that what we need is a solution that is able to deal with future generations. Given that we dont know what circumstances may exist in future generations a statistical method might prove itself utterly inneffective. In that case we have to go through this rigmarole all over again.

What I want to see is a solution that is appropriate for any possible metagame. This current suggestion already appears to require a special case for Wobbuffet and Baton Passers.

This is why I think we should come back to a vote. Now, we obviously need to sort out our method of voting. Whatever you feel about the outcome, our Deoxys vote was an utter disaster.

Secondly this system wont actually tell us what is and isnt uber, it will only tell us how similar a pokemon is to certain other pokemon. We still will have to determine exactly how similar a pokemon has to be in order for it to be called uber. It is obvious that Garchomp is not as deadly as Rayquaza, but does that mean it shouldn't be uber?

Regardless of this a statistical analysis of the relative strengths of things is a useful tool and I would love to see the results.

Have a nice day.

 

[Tangerine]

Here's a few things I noticed about this thread.

Attempting to quantify whatever "uber" might be seems like a disguised attempt to simplify the game of Pokemon into one specific (or many, depending on how many characteristics Smogon decides on).

Ignoring the flaw inherent in the above situation, it is quite clear that all of the conceptutal "definitions" we have so far of what an uber is (either from this thread or the other thread) does not solve any of the big issues at hand - that even if we somehow manage to quantify power or whatever into numbers so that they are objective, any decision we make because of it is still not objective - people will find a more offensive metagame "fine" while some people may prefer stall.

How easily a pokemon a can sweep a team in common battle conditions with little to no risk to itself or its team.

"How easily" is a pretty subjective term - and it would inherently depend on battling conditions.

I also see a potential flaw to this definition - if a Pokemon X was so powerful that it sweeps through teams easily, then eventually teams will adapt to X and find ways to "deal with it". Of course, this likely makes other Pokemon Y and Z easier to sweep.

Capable of sweeping standard in common battle conditions with one turn of set up or less.

Define "sweeping" - I mean, Lucario isn't sweeping in common battle conditions even with a Sword Dance if Gliscor or Gyarados or Bulky Mence etc are still around. It is of course the goal of Pokemon to make this condition come true - then how will you measure the effect of ONE individual Pokemon if Pokemon is indeed a team game?

Think of it this way - Garchomp would be a lot less devastating to the metagame if we had less physical attackers. Many people say "Okay you left two counters with < 50% health so you can take down Garchomp" (but then fail to complete the argument which goes) and thus it leaves you more vulnerable to other physical threats.

A Pokemon's power is the ratio of the amount of damage it deals on average by the amount of damage it is dealt in common battling conditions.

Maybe - but you mentioned a few of the flaws within this system. How much power is setting up support (SR, TWave, Toxic Spikes) to set up a team? How about "walls" who just recover off damage and bring the so called pokemon with high power to a screeching halt?

The problem with all of these definitions is that "it depends" on the environment of play. It almost feels as if everyone assumes that Pokemon is a game of 6 1 vs 1 critters when making up these defintiions. Of course we do have a good idea of what kind of environment we play in currently but for the sake of...

What concerns me about such a system for deciding uberness, is that what we need is a solution that is able to deal with future generations. Given that we dont know what circumstances may exist in future generations a statistical method might prove itself utterly inneffective. In that case we have to go through this rigmarole all over again.

If we are deciding on a characteristic of "uber" Pokemon then it should be something that can be applied anywhere.

Of course, in order to do this we must have the purpose behind "uber". Uber is simply something that is not wanted in the standard metagame for whatever reason - although the biggest reason would be that it is "game breaking". This leads to a rather nice discourse on what "power" is but a power of a pokemon depends on the entire team and assuming that every "uber" will be an one man army capable of entire sweeps by themselves is a ridiculous assumption.

I still believe that the only working measure of what "power" is can only be decided by usage. Of course, the fact that there are many subpar and not very competitive players playing (aka, refusing to use certain Pokemon just so they can be "unique") ruins the idea behind this, but the fact that if one is assumed to be competitive, then one will use the "best" pokemon - which gives you an indirect measure of what power is.

I still stand by my conceptual definition I gave in the "definition of ubers thread" which I will paraphrase as

"A Pokemon is Uber if its usage significantly rises X consecutive months" (You can run like a 5% significance test or something to see if it rose by sheer chance or because it really did rise or something)

Again, in order to properly understand this concept, one must know that

1) Usage indirectly measures power (the paragraph above)
2) That if something is a threat then people will go out of their way to deal with a threat - since they want to maximize their chances of winning.(Don't even mention wobbuffet considering "no one can prepare for Wobbuffet")

Point two can be measured of course by the usage stats each month - we can observe and analyze the Pokemon chosen and the most common movesets and analyzed

Then the definition implies that if a Pokemon's usage rises significantly for X consecutive months, we see that this Pokemon is still very, very, good even though people are going out of their ways to deal with it. Hence, "too powerful".

Of coruse usages come and go. I believe the healthy metagame would be one where the usages come and go depending on of course the Pokemon choice in the metagame. If one Pokemon, from month to month, continually dominates the charts then it probably tells us a bit too much about that specific Pokemon.

 

[Hipmonlee]

One thing about your method is that useage of pretty much everything is increasing. And I am worried about other anomalies, like, say Gyarados is the only Garchomp counter, and Gyarados becomes popular because of the threat of Garchomp, so Gyarados useage increases, and ends up increasing more than Garchomp, because it is useful also for stopping Infernape. Then Gyarados gets banned when its overuse was entirely down to the strength of Garchomp, and if Garchomp was banned, useage of Gyarados would level off.

Also even top level players dont adapt immediately to metagame shifts. I'm still using my december team, basically out of sheer laziness.

Have a nice day.

 

[Tangerine]

One thing about your method is that useage of pretty much everything is increasing. And I am worried about other anomalies, like, say Gyarados is the only Garchomp counter, and Gyarados becomes popular because of the threat of Garchomp, so Gyarados useage increases, and ends up increasing more than Garchomp, because it is useful also for stopping Infernape. Then Gyarados gets banned when its overuse was entirely down to the strength of Garchomp, and if Garchomp was banned, useage of Gyarados would level off.

Also even top level players dont adapt immediately to metagame shifts. I'm still using my december team, basically out of sheer laziness.

Have a nice day.

Not in terms of %s

Also, yes, "lazy players" hurt this conceptual definition.

I also thought of that anomaly - but in the end I believe that analyzing it enough to see which are the truly "centralizing" forces that needs to be banned shouldn't be very difficult to do - and the thought of its counters outdoing the pokemon it is supposed to be countered would just mean that the pokemon that it is supposed to counter would decrease in usage, and more players would find ways to deal with the counter in question. This means that it shouldnt be increasing for consecutive months, month after month. It's an anomaly that yes, in theory it might be possible but I'm doubting that its possibility in practice.

 

[Hipmonlee]

It certainly to me appears to be more geared toward dealing with sweepers than to dealing with a supporting pokemon or a wall.. Though I guess maybe that is due to their being more potential for a sweeper to be overcentralising than a supporting pokemon..

Have a nice day.

 

[Jumpman16]

What concerns me about such a system for deciding uberness, is that what we need is a solution that is able to deal with future generations. Given that we dont know what circumstances may exist in future generations a statistical method might prove itself utterly inneffective. In that case we have to go through this rigmarole all over again.

I think that, even if this were true, it would still be necessary to come up with as accurate a definition as soon as possible. We shouldn't fail to act now simply because the conditions may change slightly in the future. In actuality, acting now will give us a solid framework upon which we can build in future generations. We wouldn't necessarily have to reinvent the wheel

What I want to see is a solution that is appropriate for any possible metagame. This current suggestion already appears to require a special case for Wobbuffet and Baton Passers.

I'm glad you have realized that this is the goal, but I encourage you to think of ways to alter our existing suggestions so that they fit better. I have already drawn attention to the danger of tailoring our characteristics in order to have them match specific pokemon. So yes, I want to see a solution that is appropriate for any possible metagame as well, but that can only be arriving at by actively trying to arrive at one, not merely pointing out aforementioned pitfalls in a particular method.

This is why I think we should come back to a vote. Now, we obviously need to sort out our method of voting. Whatever you feel about the outcome, our Deoxys vote was an utter disaster.

What kind of vote?

Secondly this system wont actually tell us what is and isnt uber, it will only tell us how similar a pokemon is to certain other pokemon. We still will have to determine exactly how similar a pokemon has to be in order for it to be called uber. It is obvious that Garchomp is not as deadly as Rayquaza, but does that mean it shouldn't be uber?

It is my opinion that with enough characteristics, the line between uber and not uber will become more and more defnied because we will have enough characteristics to apply or not apply to any given pokemon.

Regardless of this a statistical analysis of the relative strengths of things is a useful tool and I would love to see the results.

Have a nice day.

Can you think of any additional characteristics or ways we should alter the ones suggested thus far?

 

[Ancien Régime]

Maybe a good benchmark would be "how would I conceivably deal with this pokemon using common strategies, once it gets set up, for example, how would i deal with DD Yache Rayquaza once it gets set up?"

Also what kind of sacrifice would be required to deal with said pokemon, i.e does it require that I sacrifice 3 members of my team to take it down?

The number of ways one could reply to that question could help in determining "how uber" a pokemon is.

 

[Jumpman16]

Bumping this thread for kind of obvious reasons. For those of us who have not been keeping up with PR, I've realized that those of us who meet the Suspect Test Rating/Deviation requirements still have to depend on our own personal perception of what is uber and what is not to vote. Therefore require as accurate a definition as we would have if we were still relying on the bold voting process that decided the tiering of Suspects.

So...we kind of left off with a few good characteristics, and the only reason I let this thread die was because I, along with everyone else, was under the impression that a definition of uber did not matter for the Ratire/Deviation vote process, not because I think our efforts were futile. We ahve some good stuff here as far as what to consider:


How easily a pokemon a can sweep a team in common battle conditions with little to no risk to itself or its team.

How capable a pokemon is of sweeping standard in common battle conditions with one turn of set up or less.

The ratio of the amount of damage it can deal with the amount of damage it can withstand.


Going to open it up now to you guys to see if we can think of anything else. I've also reread about a third of this thread so I may double post if I think of something "first" when when I'm done rereading (which I encourage you all to do). I must stress, though, that I'm confident we can hammer something that that's really concrete, and that even if we can only think of one of two more things, everybody will then be on the same page regading future suspect and the concept of uber in general.

 

[Syberia]

How easily a pokemon a can sweep a team in common battle conditions with little to no risk to itself or its team.

How easily it, being a single pokemon, allows another to do the same thing. A la Mew and Wobbuffet.

Question for X-Act, but with regards to "ratio of damage dealt to damage taken," is there any way to compute that?