Uber voting ratios - revisited

[Tangerine]

Tangerine, your argument about close-minded people works the other way around too, you know. As in, there are people who will think a Pokemon deserves to be in OU no matter what the evidence is against it.

Yes.

This is PRECISELY why doing two simple majorities is worthless - you're literally agreeing with me here.

You say that because if something is indeed broken, then after a while, even the stubborn people will break and think it is Uber, well I disagree. If they are close-minded by definition, then they will continue their beliefs that a Pokemon is OU for as long as they play. And even if what you assume were to be true, how long are we all supposed to wait? Garchomp has been "tested" for over a year, and it's still on the fence with its tiering. If we opt for anything more than a simple majority, we risk having a minority of close-minded OU voters decide the metagame for a long time. So long, in fact, that I don't think we will ever reach a balanced metagame before the next generation comes out. You act as if we all have an infinite amount of time to run all of these tests, but that's simply not the case. I, for one, would like to see a stable environment in which I can confidently build a team that I won't have to tweak at the next whim of ignorant voters.

I believe you misunderstood me here - retesting the suspect against in the stage will take a month or two - while anything else will take years and years. People won't change their mind in a month if they know "if they vote this way you can get it out of the system". Rather, force them to stick with it, and then over time, we'll find it broken or not. You're completely misinterpreting my point - I'm saying that people will actually "play the game" once we're finally "done" with this iteration of suspect testing and if Garchomp or something is actually broken then we'll 100% see it.

...this is also true, then what exactly do you think we're going to get out of the Suspect Test process or any tiering process, anyway?

Well, the first comment you quoted was an obvious exaggeration. However, if we DID promote arguments to decide tiers then I presume we'd have more users looking over arguments to the point where they are capable of creating some kind of theory and capable of seeing the big picture - something that isn't promoted *at all*. I think having debate based tier changes would be a lot better than voting, and you know that i've thought this way forever now lol

Secondly, I'd much rather have just a few people just decide tiers and live with it - it would be a lot easier than this entire process and we'd get pretty much identical results anyway since in the end it doesn't matter that much what pokemon is banned or not in the metagame since pretty much every user have no grasp on how the metagame runs and how things are interconnected in the big picture.

 

[Philip7086]

Yes.

This is PRECISELY why doing two simple majorities is worthless - you're literally agreeing with me here.

What I was trying to point out is the fact that there are people on both sides of the issue who will vote for their own stubborn belief, and because of that, increasing the Uber requirement to something more than a simple majority does not fix the problem at all. Instead, our efforts should be focused on increasing the quality of voters.

I believe you misunderstood me here - retesting the suspect against in the stage will take a month or two - while anything else will take years and years. People won't change their mind in a month if they know "if they vote this way you can get it out of the system". Rather, force them to stick with it, and then over time, we'll find it broken or not. You're completely misinterpreting my point - I'm saying that people will actually "play the game" once we're finally "done" with this iteration of suspect testing and if Garchomp or something is actually broken then we'll 100% see it.

I don't see how you can make this claim when Garchomp itself has been used in the OU environment for over a year, and people still can't agree on its tiering. Raising the requirement for something like Garchomp to a 2/3s majority then forcing people who don't agree with one another to play with it until we all can agree does not make any sense whatsoever, especially coming from someone who believes most people are close-minded.

Again, we should stick to making sure the voter pool is comprised of the best possible candidates only, and then putting our faith that if a majority of them think something is Uber, then it should be so. And if that majority is slim, then maybe give people a little more time to think it over again in another test. But if even then the vote goes to Uber, then maybe it's time to accept it.

 

[Tangerine]

What I was trying to point out is the fact that there are people on both sides of the issue who will vote for their own stubborn belief, and because of that, increasing the Uber requirement to something more than a simple majority does not fix the problem at all. Instead, our efforts should be focused on increasing the quality of voters.

I recognized that there were people on both sides. And I'll ask the same question as I did Earthworm - how do you plan on increasing the quality of voters? Secondly, how does this even solve the issue? The people who are stubborn about this are the people who will go out of their way to vote - and the chances are you can't filter them out through any scheme like the ones you presented.

I don't see how you can make this claim when Garchomp itself has been used in the OU environment for over a year, and people still can't agree on its tiering. Raising the requirement for something like Garchomp to a 2/3s majority then forcing people who don't agree with one another to play with it until we all can agree does not make any sense whatsoever, especially coming from someone who believes most people are close-minded.

http://www.smogon.com/forums/showthread.php?t=45619

Your last statement is a definite strawman - While people are closed minded - they are allowed to be closed minded because all they need to do is defend themselves for that time period (a month or so) for the second test, and it's gone. I can 100% guarantee that people will not be open minded on the second test of stage 3 regarding things like Garchomp.

Raising the requirement has a simple basis - if it's broken, then people will abuse it to the point where everyone will agree that Garchomp is broken, or at least a solid majority, like the last garchomp test. The real issue is platinum changes and now the introduction of latias and possibly manaphy.

 

[Philip7086]

I recognized that there were people on both sides. And I'll ask the same question as I did Earthworm - how do you plan on increasing the quality of voters? Secondly, how does this even solve the issue? The people who are stubborn about this are the people who will go out of their way to vote - and the chances are you can't filter them out through any scheme like the ones you presented.

On the contrary, I know of a lot of stubborn Garchomp-to-OU supporters who did not put much effort into making voting requirements, and vice versa. Creating more strict requirements serves to weed these people out and only allow those who actually care to vote. Also, even close-minded people have come to their own conclusion through some method of reasoning or another, so I don't see what's wrong with letting them vote twice the same way. It's the on-the-fence voters who will be given another chance to rethink their choices should option 3 be implemented. I'm sure you don't mean to insinuate that every single voter has already made up their stubborn mind on every suspect to the point that another test won't change anybody, do you?

As to how I plan to increase voter quality, I have already discussed it, so you can review my previous posts. If you feel that my methods are insufficient, then instead of shooting them down, I hope you will instead come up with a different idea for selecting qualified voters, because as I've said before, I believe that is the real issue here.

Raising the requirement has a simple basis - if it's broken, then people will abuse it to the point where everyone will agree that Garchomp is broken, or at least a solid majority, like the last garchomp test. The real issue is platinum changes and now the introduction of latias and possibly manaphy.

The whole point of increasing voter quality is so that we do not have to wait months or years to decide what is OU and not. If we get it right the first time, we can achieve the true competitive metagame faster, and let the real playing begin. This is something Smogon should be pushing for, instead of taking so long that nobody ever plays in the true DPP metagame, because the next generation came out before testing ever finished.

 

[MythTrainerInfinity]

I've stated this before several times, but...Having qualified voters discuss it in public only amongst themselves prior to the vote will help weed out some of those bad voters (I was afraid of saying this outright because I don't want "yes men" abusing the system) and could help on the fence voters make up their minds. Overall it should increase voter quality. It should be fairly obvious when someone is making a bunch of bull up vs. someone who knows what they are talking about. This allows the whole forum to sort of analyze it and if someone sees something that just makes them go "what?" then that can be addressed.

A couple people even if they are amazing and neutral may overlook something a potential voter wrote in their paragraphs.

If a simple majority of what I consider to be a highly qualified group of people vote a Pokemon uber, then that is good enough.

2/3 vote to make a solid decision would be very hard and it would take too long to test in my opinion... A simple majority just doesn't feel like its a firm enough decision either especially if it reeks of uncertainty. A simple majority doesn't scream overwhelmingly Uber to me, which as stated before an Uber should dominate OU and that should be obvious amongst highly qualified experts. So I think that at least a certain percent should be required for at least stage 3 to overturn something.

Secondly, I'd much rather have just a few people just decide tiers and live with it - it would be a lot easier than this entire process and we'd get pretty much identical results anyway since in the end it doesn't matter that much what pokemon is banned or not in the metagame since pretty much every user have no grasp on how the metagame runs and how things are interconnected in the big picture.

Well, with a debate you're going to eventually have to have a vote to ensure accuracy of the way people felt.

These things allow new people to get involved and potentially contribute, which Smogon needs. With a shifting metagame you would need to have the few people who do make the decision be completely up-to-date with the metagame otherwise its not going to be fully accurate. So, why not just go along with what we're doing and just refine the processes? For future gens it'll be easier, by then the processes should be (hopefully) ironed out. So, in the long run its good that we're struggling along here now instead of later.

To those running these tests and such... Thanks for being patient with all of this.

 

[Hipmonlee]

The reason that two simple majorities is not pointless is it resolves issues dealing with the possibility that this outcome was a statistical anomaly, which is possible with such a close outcome.

So if people refuse to change their votes based on any amount of testing, then having a second vote will still demonstrate that the result wasnt a freak occurrence.

And if people wont change their vote then the whole suspect test has been a waste of time. But, the fact is some people will change their votes. They may be in a minority, but they will change..

One of the issues with this debate that hasnt come up yet, is that pokemon by its nature is a game of diversity. People use very different teams. It may be that some people by the nature of their style of play dont notice the uberness of a certain pokemon as much as others might. Which is more or less why I support a simple majority, but not one that might just be a statistical anomaly.

Also I really think we should be making more effort to explain to voters their responsibilities. If you really make an effort to impress upon people the importance of voting according to our definitions or philosophy then I doubt there are more than a handful of people cynical enough to do what Philip has been describing..

Have a nice day.

 

[Tangerine]

On the contrary, I know of a lot of stubborn Garchomp-to-OU supporters who did not put much effort into making voting requirements, and vice versa.

Okay, but that doesn't solve the issue - adding more and more filters isn't going to stop stubborn closed minded people from voting. Again, how are you going to filter? You have never mentioned it - you continually just repeat "we need to filter out people better" which is great - but "how"? And how will this filter out closed minded people? I definitely agree with you we need better filters but how are we going to make this practical? we are not mind readers. There's literally not much people can do about this. I'm personally not going to make sure everyone knows what they are talking about before I let them vote, and I doubt anyone would.

. It's the on-the-fence voters who will be given another chance to rethink their choices should option 3 be implemented. I'm sure you don't mean to insinuate that every single voter has already made up their stubborn mind on every suspect to the point that another test won't change anybody, do you?

While there are a few users who actually play the suspect test like they are supposed to - I still believe the majority of the players have decided beforehand how the suspect would be. I think suspectexp was a nice thing because it literally forces people to experience the suspects so it may change minds - but I don't think it's done on a long enough interval for it to be effective - which is why I make the claim that a revote in a month (which is what is currently being proposed) - is going to be worthwhile. I think at this point we just literally just say "they're OU", and let people play with it for a long period, before getting another chance to vote.

The whole point of increasing voter quality is so that we do not have to wait months or years to decide what is OU and not. If we get it right the first time, we can achieve the true competitive metagame faster, and let the real playing begin. This is something Smogon should be pushing for, instead of taking so long that nobody ever plays in the true DPP metagame, because the next generation came out before testing ever finished.

Again, does it really matter?

At this point we can just say "Okay, these Pokemon did not meet supermajority, they are now OU until further notice". Is it REALLY going to matter? If it's broken, then we'll know, and tweak the metagame accordingly. I think it's a bad thing to assume that after this we're 'done" and the metagame will be "perfect" (it isn't), especially since it takes a long time for people to mess around with strategies and mechanics before finding ways to abuse it - at which point of it being abused, we'll hear about it and deal with it accordingly.

Let's not enslave ourselves to this process and pretend that this is going to get us to some perfect metagame - while it has successfully filtered out candidates that were considered broken, it's not like things aren't going to pop out in the future. If it is truly broken, then it'll be clearer in the future as people find ways to optimize it and maximize the usage in the game, and everyone has to use it or deal with it. Let's not kid ourselves regarding htis.

Well, with a debate you're going to eventually have to have a vote to ensure accuracy of the way people felt.

Since when did we vote on a debate? let's not do that since it'll turn into pure fluff.

The reason that two simple majorities is not pointless is it resolves issues dealing with the possibility that this outcome was a statistical anomaly, which is possible with such a close outcome.

It does not resolve it - it makes it less unlikely. If the poll was 52% the first time and 51% the second, you can bet I'm still not going to be convinced.

So if people refuse to change their votes based on any amount of testing, then having a second vote will still demonstrate that the result wasnt a freak occurrence.

That's irrelevant to what you're referring to as a statistical anomaly.

And if people wont change their vote then the whole suspect test has been a waste of time. But, the fact is some people will change their votes. They may be in a minority, but they will change..

Or rather, the second round of stage 3 is coming WAYYYY too fast for people to change their minds.

 

[Hipmonlee]

Is the second round of testing going to be one month?

The first was significantly longer, I personally havent heard anything about the length of the second test..

Have a nice day.

 

[MythTrainerInfinity]

I've suggested a method of filtering, yet it keeps on getting ignored...

You really don't seem like you hold the voters to very high esteem, Tangerine... Sure there have been some mistakes in the past, but like if we educate the voters on what their responsibility is wouldn't that help? In a debate you will have to know how many people stand for each side, so I guess in my mind that qualifies as a vote.

Well at least when I vote I try to ensure my votes are as unbiased as possible, which is why I refused to vote on Manaphy this past test. I know there are probably many other users who take this seriously enough to vote without bias. As getting people with their own agendas there's always going to be people like that in a vote or a debate. However, if we can see what and how they think in public amongst other voters it will be easier to tell if they are spouting off a bunch of bull or not.

How much longer are we going to just craft theories about how this will pan out? No offense to people, but I think this whole process is taking way longer than it needs to be. I understand the need to analyze and discuss this, but HG/SS is out in Japan and we aren't even completely done with one cycle of suspect tests. I think a goal of ours should be to like at least finish a cycle of Suspects before a new game comes out just in case it presents something drastically different to the metagame.

 

[Aldaron]

I'm going to ignore all references regarding the intelligence levels / capacity of the voters, as I find most of these to be extremely condescending and just overly cynical spouting.

This boils down to something very simple: the compromise between ideology and practicality.

I think the people stringently demanding a supermajority with no recognition of the very high probability that garchomp will probably always be voted uber by a simple majority in a stage 3 type test are just being unreasonable by demanding we stick to their ideals (remember, while I agree that banning is a more serious action than unbanning, this is an ideal you are imposing on "us")

Remember, even if Garchomp is being voted by a simple majority, it doesn't mean _nothing_ nor is it "worthless." Just because it doesn't mean a lot (enough to ban), doesn't mean it doesn't mean something.

I also believe the people demanding only 1 simple majority are being unreasonable. As the proponents of the supermajority option have stated, there is a lot of uncertainty in 1 stage 3 simple majority vote (especially one that was voted uber by LESS than the number of remaining voters). It doesn't hurt the process by declaring Garchomp "unresolved" (not ou, just unresolved) and testing it again.

That's why I proposed and still advocate the compromise between ideology and practicality and hope that we go with 2 simple majorities in stage 3 testing or just 1 super majority.

This isn't compromise for the sake of convenient compromise or "flowery meadows of compromise" as blame game is implying that some users might be supporting this for; this is compromise for the sake of practicality.

Also, to address a point that was made regarding the integrity of the suspect process if we were to change this in hindsight. I feel like the suspect process wasn't set in stone, and necessary changes will inevitably go through during the first few times we go through it. I noticed that the administration has already taken this view, as it ignored the promise of public voted for ALL the suspects AFTER a designated period of time by deciding to make Deoxys-S uber without a vote and not after that designated period of time. So, we're already changing the noted "rules" of the suspect process to make it go better; is changing again that much of an issue?

EDIT:

I really want to get rapid discussion on this again. I don't want this to die and then we move forward with Latias and Manaphy in stage 3 testing because it lost its momentum later on. I think it's evident that the majority of PR (>75%) doesn't believe 1 simple majority is enough, so at the very least I hope we aren't going to move ahead with only Latias and Manaphy in the next iteration of stage 3.

 

[Tangerine]

To me the purpose of the entire suspect process was the simply test out Pokemon that would never join the OU tier (Manaphy, Latios, Latias, and friends) and give them a shot in the current OU environment. This means the following. The suspect test should only aim to weed out the obvious ubers, or else in the future we’ll get the same issue again. Meaning that it’s good that we filtered out things like Latios—which was voted overwhelmingly uber.

Why should the suspect test only filter out the obvious? The idea behind this is simple—it simply takes too long for users to properly abuse a set. Deoxys-S took many months in the ladder before the dual screen set was decided to be overwhelmingly broken. Garchomp usage only took off in the later days of D/P, and Lucario’s best set wasn’t discovered until much later after the metagame began. Things that are broken tend to be obvious and completely visible (not that I’m implying Lucario to be broken—it’s just an example) and I believe this is definitely why we need to give it the benefit of the doubt when it comes to voting Pokemon Uber. It simply takes a long time before we can truly say “this is broken” or “this is not broken”. We need to give people time to ADJUST—whether it be time to figure out how to deal with the threat like Garchomp or time to figure out how to properly abuse Garchomp. The end result is that it is highly unlikely that it’ll stay in a 50/50 even split after a long stretch of time.

Again, the results of a simple majority is not obvious. Two simple majority still doesn’t make it “obvious” a Pokemon is uber or not—50%+1 is simply ambiguous on whether or not something is broken or not. This is why I actually propose either we deal with a supermajority—or at this point, simply say “These are all OU now, break them as you see please” and ban the Pokemon that are actually broken (although we might have to test something quickly for Manaphy or something). But the thing is—if it’s broken, then people will mention it, and our players will see it, and that’s when we can simply vote again. This gets rid of the problem of dealing with another suspect ladder and the voter fatigue. Setting a time period on this is something I rather not see since it puts in an arbitrary limitation on “find out in this time frame or else” which is counterproductive. The suspect test, again, will give you a good first impression, but how to properly abuse a set takes a long period of time, as I’ll say it over and over again, and not to mention how to properly deal with the Pokemon. If the first impressions are overwhelmingly uber even without discovering any sets – then that’s an obvious uber, and we’re happy to ban it. However, if the results are even slightly bit ambiguous, then I believe it warrants the OU status, until people bring it up again for it being broken. I think relying on two simple majorities is simply dodging this issue and may be too rash in this.

I will say it again and again. If something is broken, it should be obvious to everyone it is broken, not a 50%+1, but a majority stronger than such.

"Later" doesn't mean a thing to me when we are already figuratively compromising the integrity of the Stage 3 process by potentially changing what "majority" we go by in the middle of the Stage, as I've pointed out. You may clarify that your "later" would come before the votes are actually made, but that is much more important than you have indicated thus far. How would a compromise on a ratio after knowing Garchomp or Manaphy was voted uber or ou by 58% or 61% in Stage 3-2 even remotely be unbiased, especially when we know which one will mean at least a month more of testing?

While yes, I do see your concerns, but just because we know the results and we will pick a ratio alongside with it, does not necessarily mean we're compromising future votes (in fact i didn't even know manaphy was 58%). I think my post addresses this, that in given time to properly abuse it, the results will end up stronger, not more ambiguous, since people would have time to break it to their heart's content. We'll definitely be biased in a 1 month time frame, but as the time frame grows larger and larger, this issue becomes less and less significant.

 

[Jumpman16]

A simple majority doesn't scream overwhelmingly Uber to me

Nor should any of the Suspects almost by definition. The only overwhelmingly Uber or not Uber (OU UU NU) pokemon we have right now are Suspects.

Anyway, I feel I posed a lot of legitimate concerns with honoring any kind of supermajority in the middle of the process, and none of these have yet been addressed. Aeolus and I will not allow this process to be further delayed by a lack of a consensus on whether or not a lack of consensus really indicates a lack of consensus. If you want to address our concerns, do so as soon as possible, and if not, Doug can finally configure the Stage 3-2 Suspect Ladder.

 

[Theorymon]

I have a question, will this poll apply to UU too? I feel that if so, then Frosslass (and maybe Raikou) need a retest. Just looking at the Frosslass poll, Frosslass was banned from a 10-9 vote! Also, because UU is based off of the usage of OU Pokémon, that means the tier constanlty changes, thus there could potentially be metagame changes that makes a Pokémon less broken in UU. I don't play UU, so please correct me if this shouldn't be the way UU suspect voting is handled. I just feel that a Pokémon being banned because of a 1 vote difference to be silly.

 

[Cathy]

Anyway, I feel I posed a lot of legitimate concerns with honoring any kind of supermajority in the middle of the process, and none of these have yet been addressed. Aeolus and I will not allow this process to be further delayed by a lack of a consensus on whether or not a lack of consensus really indicates a lack of consensus. If you want to address our concerns, do so as soon as possible, and if not, Doug can finally configure the Stage 3-2 Suspect Ladder.

I would argue that any disadvantage to changing the procedure now and requiring a second simply majority to ban the close cases (where close is defined operationally as < 2/3 majority) is outweighed by the disadvantage of making a climactic tiering decision that will affect competitive pokemon perhaps for a long time based on a vote that was decided by three votes out of eighty-five, when nine people hadn't voted. It seems that a large majority in this thread would agree that this was just not convincing.

Is there really a significant amount of harm to requiring a second simple majority to ban Garchomp and Shaymin-s, even though that wasn't the original plan? I think the advantages outweigh the disadvantages, and I'd say that everybody who has agreed with something other than a simple majority in this thread was aware of the practical consequences of changing the procedure part way through when coming to their opinion. So although it may be dubious to require a supermajority for that Garchomp vote that already took place, I think the second simple majority requirement could be a fair compromise for stage 3, including retroactively applied.

 

[Jumpman16]

i would agree but only because less than two dozen people voted. a pool of 19 screams "sample size"—hell, 10-9 is literally 50% + 1 in this scenario, and if just three ou voters vote uber instead you'd have that magical supermajority that everybody seems to love

To me the purpose of the entire suspect process was the simply test out Pokemon that would never join the OU tier (Manaphy, Latios, Latias, and friends) and give them a shot in the current OU environment. This means the following. The suspect test should only aim to weed out the obvious ubers, or else in the future we’ll get the same issue again. Meaning that it’s good that we filtered out things like Latios—which was voted overwhelmingly uber.

Why should the suspect test only filter out the obvious? The idea behind this is simple—it simply takes too long for users to properly abuse a set. Deoxys-S took many months in the ladder before the dual screen set was decided to be overwhelmingly broken. Garchomp usage only took off in the later days of D/P, and Lucario’s best set wasn’t discovered until much later after the metagame began. Things that are broken tend to be obvious and completely visible (not that I’m implying Lucario to be broken—it’s just an example) and I believe this is definitely why we need to give it the benefit of the doubt when it comes to voting Pokemon Uber. It simply takes a long time before we can truly say “this is broken” or “this is not broken”. We need to give people time to ADJUST—whether it be time to figure out how to deal with the threat like Garchomp or time to figure out how to properly abuse Garchomp. The end result is that it is highly unlikely that it’ll stay in a 50/50 even split after a long stretch of time.

I disagree because I feel that the only reason Garchomp is close to 50/50 is because the Platinum changes were more a detriment to Garchomp than a benefit. There is no pokemon that we've had more time to "adjust" to than Garchomp and people are still allowing themselves to be trolled by it after over two years. I would argue that the main reason Garchomp even ever got up to 72% (with voters in a pool I wouldn't exactly call "pure") is because the Yache SD set was discovered, and now it's back close to the 50/50 split it was at in January of 2008 (after it took over Blissey's top spot in Standard usage). I would therefore further argue that anything that would convincingly sway it either way is going to be actual movepool changes, but unlike with Bullet Punch Scizor, HGSS's changes don't point either way—Super Fang doesn't make Walrein a Garchomp check (maybe if it got Ice Shard [watch out for Glalie I guess, Standard]) since Gravity is definitely a considerable plus for a pokemon whose can now use the best physical move in the game and only have four pokemon in OU resist it (Celebi, Breloom, Heracross and Ninjask...great now I want to try a Gravity strategy with SD/AA/EQ Gliscor).

So I don't think new cartridges and their new moves should be the reason Garchomp circumvents a vote in the middle of a Suspect Test, or any pokemon for that matter. The fact that it has been tested across three different cartridge releases is a testament to how long this process has taken, and I feel that only movepool changes would change anything noticeably. Even then, I have seriously no reason to believe that people are going to be able to break Garchomp more than it was already with four other Suspects around it. Which actually raises a point I've only mentioned once or twice if at all—does anyone seriously think that Garchomp will be voted less uber when you remember it doesn't have Latios and Skymin to check it now? I'm not implying that it 100% would be voted more uber than 52% in Stage 3-2 if we tested it, but any support characteristics that Latios and Skymin could have had in relation to Garchomp are surely outweighed by it having two fewer pokemon to deal with that outspeed and OHKO it without overspecializing.

Again, the results of a simple majority is not obvious. Two simple majority still doesn’t make it “obvious” a Pokemon is uber or not—50%+1 is simply ambiguous on whether or not something is broken or not. This is why I actually propose either we deal with a supermajority—or at this point, simply say “These are all OU now, break them as you see please” and ban the Pokemon that are actually broken (although we might have to test something quickly for Manaphy or something). But the thing is—if it’s broken, then people will mention it, and our players will see it, and that’s when we can simply vote again.

People will literally never stop mentioning that Garchomp is broken for as long as it is in OU. This will never change.

This gets rid of the problem of dealing with another suspect ladder and the voter fatigue. Setting a time period on this is something I rather not see since it puts in an arbitrary limitation on “find out in this time frame or else” which is counterproductive. The suspect test, again, will give you a good first impression, but how to properly abuse a set takes a long period of time, as I’ll say it over and over again, and not to mention how to properly deal with the Pokemon. If the first impressions are overwhelmingly uber even without discovering any sets – then that’s an obvious uber, and we’re happy to ban it. However, if the results are even slightly bit ambiguous, then I believe it warrants the OU status, until people bring it up again for it being broken. I think relying on two simple majorities is simply dodging this issue and may be too rash in this.

I will say it again and again. If something is broken, it should be obvious to everyone it is broken, not a 50%+1, but a majority stronger than such.

And I think it's more counterproductive to have Garchomp in OU where people are bound to say that it is broken, as many have for almost two years, especially when we have yet to decide what kind of "supermajority" is even good enough for us besides just throwing around that word with one person being ok with 55% and another with 60% and another with 67% and yet another with 70%, all the while with potentially 54% or 59% or 66% or 69% of the people who play standard competitive pokemon havign to deal with a pokemon they truly feel is uber.

While yes, I do see your concerns, but just because we know the results and we will pick a ratio alongside with it, does not necessarily mean we're compromising future votes (in fact i didn't even know manaphy was 58%). I think my post addresses this, that in given time to properly abuse it, the results will end up stronger, not more ambiguous, since people would have time to break it to their heart's content. We'll definitely be biased in a 1 month time frame, but as the time frame grows larger and larger, this issue becomes less and less significant.

We had three months to break Garchomp, and we feel it's 52% broken even with two other "overwhelming" ubers in Latios and Skymin with it. Nothing significant will change that doesn't have anything to do with movepool changes, and we haven't even decided what percentage is "significant" anyway.

 

[MythTrainerInfinity]

Yeah a 10-9 is close, but with the small pool of voters it doesn't seem right to want to say the margin wasn't large enough.

Nor should any of the Suspects almost by definition. The only overwhelmingly Uber or not Uber (OU UU NU) pokemon we have right now are Suspects.

True, but some votes were clearly one sided. I still would like to see a discussion of experts of a particular Suspect test discuss things though to shed more light on things. When I saw some people vote a certain way in some Suspect tests I was pretty surprised. However, I was disappointed that some of them really did not supply their reasonings in public. Allowing us to gauge on how the players felt in their own words should make the vote feel more justified instead of voting without having to supply reasoning in public.

Anyway, I feel I posed a lot of legitimate concerns with honoring any kind of supermajority in the middle of the process, and none of these have yet been addressed. Aeolus and I will not allow this process to be further delayed by a lack of a consensus on whether or not a lack of consensus really indicates a lack of consensus. If you want to address our concerns, do so as soon as possible, and if not, Doug can finally configure the Stage 3-2 Suspect Ladder.

Yeah, as much as I'd like to see us coming to an agreement on a certain percent being the cut-off, I'd love to finish this. I still support Tangerine's suggestion of 55%. I may have seem like I was really hasty making that other poll, but I wanted quick and painless data, to aid us in getting stuff figured out ASAP.

Once this is all said and done though I think it'd be best to save for the policy changing of this until after we finish Stage 3 and figure things out before we go through another cycle. So yeah, getting this done is more important. I feel we'll be better to judge all of this better in hindsight anyway.

For a first time going through a cycle of this Suspect stuff I daresay even though we hit a few bumps in the road we still are pretty successful and people should be more thankful of it. I mean its better than just a select few people assigning tiers behind closed doors (ex. Third Gen) and allowing other members of the community voice their opinions. I don't want to sound like a suck up, but people need to be more appreciative of that.

 

[lilyhollow]

This isn't compromise for the sake of convenient compromise or "flowery meadows of compromise" as blame game is implying that some users might be supporting this for; this is compromise for the sake of practicality.

I don't want anybody to misinterpret the post I made, so I'd just like to point out that I agree with this entirely (with the obvious exception of the implication that i disagree with this entirely, which i disagree with entirely). The point of my post was to assert that voting a Pokemon "Uber" by two simple majorities is no more convincing of its "Uber" status than just one simple majority vote; it's convincing of a sort of limbo status, and we may decide from there that for the sake of practicality we want to ban all limbo Pokemon anyway, but it's not more convincing of Uber status. To me, that seemed like something easy to overlook given that I myself somewhat overlooked it when considering Hipmonlee's compromise as he first proposed it. If this was something that everyone else has already considered and finds very clear, particularly those who agree with the compromise, then I obviously have no problems with the "happy meadow" or whatever. That's the only real "implication" to be found anywhere in my post, not that such a compromise would be entirely pointless.

 

[Tangerine]

I feel that the only reason Garchomp is close to 50/50 is because the Platinum changes were more a detriment to Garchomp than a benefit. There is no pokemon that we've had more time to "adjust" to than Garchomp and people are still allowing themselves to be trolled by it after over two years. I would argue that the main reason Garchomp even ever got up to 72% (with voters in a pool I wouldn't exactly call "pure") is because the Yache SD set was discovered, and now it's back close to the 50/50 split it was at in January of 2008 (after it took over Blissey's top spot in Standard usage). I would therefore further argue that anything that would convincingly sway it either way is going to be actual movepool changes, but unlike with Bullet Punch Scizor, HGSS's changes don't point either way—Super Fang doesn't make Walrein a Garchomp check (maybe if it got Ice Shard [watch out for Glalie I guess, Standard]) since Gravity is definitely a considerable plus for a pokemon whose can now use the best physical move in the game and only have four pokemon in OU resist it (Celebi, Breloom, Heracross and Ninjask...great now I want to try a Gravity strategy with SD/AA/EQ Gliscor).

While Movepool changes definitely do have a strong effect - I think it's too much to say that it is the only way it will sway it either way. Just like the Yache SD set has skyrocketted Garchomp to its #1 Spot to the point where it overwhlemed every other Pokemon, I do believe that if it's actually broken, people will be able to use it in such a degree. How can we argue something is "broken" if over 40% of the population are totally capable of dealing with it and the other 50%+1 think it is too much? Who is playing it correctly here? If the over 40% of the population can do it, then there is no reason why other members cannot. If this proportion got narrower and narrower, then we can get a more convincing take on Garchomp - but I do think having over 40% thinking it's "just perfectly fine" is literally giving into people refusing to adjust to metagame changes (the remainder of the user). You can definitely see this trait in our wobbuffet test or even the PR posts. People do NOT like to change, and do not want to adjust to the metagame, but rather break the metagame so it's better for them.

Secondly, your assertion implicitly claims that all the possible way of using Garchomp is already there and there is no other way of using him, and that people have already discovered every single way of dealing with him and then decided so. If this were true, then a simple majority would definitely be enough. But are you really confident that this is true? Are you really that confident with the competitive community to the point where you believe people have found every possible use of garchomp and every way to deal with it in the span of 3 months while dealing with every other suspect? I think this would definitely be difficult to argue for it - and I think because of this, it is wise for us to use a much stronger result to deal with Garchomp.

So I don't think new cartridges and their new moves should be the reason Garchomp circumvents a vote in the middle of a Suspect Test, or any pokemon for that matter. The fact that it has been tested across three different cartridge releases is a testament to how long this process has taken, and I feel that only movepool changes would change anything noticeably. Even then, I have seriously no reason to believe that people are going to be able to break Garchomp more than it was already with four other Suspects around it.

I do too, and I have not used this as a reasoning in my post what so ever.

Which actually raises a point I've only mentioned once or twice if at all—does anyone seriously think that Garchomp will be voted less uber when you remember it doesn't have Latios and Skymin to check it now? I'm not implying that it 100% would be voted more uber than 52% in Stage 3-2 if we tested it, but any support characteristics that Latios and Skymin could have had in relation to Garchomp are surely outweighed by it having two fewer pokemon to deal with that outspeed and OHKO it without overspecializing.

Is it really that harmful to "make sure"? I'm definitely not confident enough to say so. If what you said is true then Garchomp would have a MUCH stronger majority voting for uber on the second test, and there would be nothing to argue about.

Secondly, I could definitely argue it - as Latios and Skymin are both primarily special threats. Without them in play, it allows players to put more focus on physical threats such as Garchomp. This would definitely let Garchomp be significantly less uber. I'm definitely sure someone who is actually familiar with the suspect metagame could make a stronger argument, but there definitely is a case for it.

People will literally never stop mentioning that Garchomp is broken for as long as it is in OU. This will never change.

Sure, but remember - what we care about is intelligent reasons on why we should test it, not randoms whining about it on shoddy or even stark mountain. I'm pretty sure we're more than capable of finding intelligent reasons and holding little topics like you have done for wobbuffet.

And I think it's more counterproductive to have Garchomp in OU where people are bound to say that it is broken, as many have for almost two years, especially when we have yet to decide what kind of "supermajority" is even good enough for us besides just throwing around that word with one person being ok with 55% and another with 60% and another with 67% and yet another with 70%, all the while with potentially 54% or 59% or 66% or 69% of the people who play standard competitive pokemon havign to deal with a pokemon they truly feel is uber.

If that's really a problem then we can literally just vote for one right now. I will talk to Colin about this - does this really mean that if we did decide, this objection wouldn't matter? I don't think this is a valid objection for that reason - it just talks about something we haven't done yet.

We had three months to break Garchomp, and we feel it's 52% broken even with two other "overwhelming" ubers in Latios and Skymin with it. Nothing significant will change that doesn't have anything to do with movepool changes, and we haven't even decided what percentage is "significant" anyway.

52% with... quite a number of people not voting/abstaining. One reason why I prefer a supermajority is that it totally gets around this problem. 52% with the population we had and the people who didn't vote is reason enough that we should be looking for a stronger result, if anything, and not a reason to say "it's enough".

 

[Jumpman16]

I would argue that any disadvantage to changing the procedure now and requiring a second simply majority to ban the close cases (where close is defined operationally as < 2/3 majority) is outweighed by the disadvantage of making a climactic tiering decision that will affect competitive pokemon perhaps for a long time based on a vote that was decided by three votes out of eighty-five, when nine people hadn't voted. It seems that a large majority in this thread would agree that this was just not convincing.

First of all, do you have any reason to believe that more than three of these nine people would definitely have voted Garchomp OU? Anything else would not have had Garchomp voted OU by any kind of majority, and we'd be having the same argument. And second of all, an even bigger handful of accounts were not claimed in the original thread with the accounts that qualified in the first place (164 accounts). Do we care about those too?

I don't. 85 is a very large number of voters. The vote may not be convincing due to how close it was, but if you or anyone else is not convinced because of the number of voters we had I don't think there's any pleasing you. I'm positive that if the vote was 65-62 some people still wouldn't be "convinced".

Is there really a significant amount of harm to requiring a second simple majority to ban Garchomp and Shaymin-s, even though that wasn't the original plan? I think the advantages outweigh the disadvantages, and I'd say that everybody who has agreed with something other than a simple majority in this thread was aware of the practical consequences of changing the procedure part way through when coming to their opinion. So although it may be dubious to require a supermajority for that Garchomp vote that already took place, I think the second simple majority requirement could be a fair compromise for stage 3, including retroactively applied.

Again, everyone has yet to address the very reasonable possibility that Manaphy get any sort of majority for uber in Stage 3-2. "Then what?" And what if I'm wrong about the remaining Suspects standing to be deemed more uber in Stage 3-2, or those evidently crucial nine voters were all on the "free garchomp" bandwagon from the beginning and Garchomp gets an OU stage 3-2 tag? This is what I'm talking about when I state repeatedly that my concerns are not being addressed and we cannot move on until they are.

True, but some votes were clearly one sided. I still would like to see a discussion of experts of a particular Suspect test discuss things though to shed more light on things. When I saw some people vote a certain way in some Suspect tests I was pretty surprised. However, I was disappointed that some of them really did not supply their reasonings in public. Allowing us to gauge on how the players felt in their own words should make the vote feel more justified instead of voting without having to supply reasoning in public.

Even if I agreed with this, how would it actually change a 52% vote or a 58% vote or a 65% vote? It certainly doesn't imply that even after only the acceptable "public bold votes" are kept we couldn't still have a 52% or some other random not-so-supermajority.

While Movepool changes definitely do have a strong effect - I think it's too much to say that it is the only way it will sway it either way. Just like the Yache SD set has skyrocketted Garchomp to its #1 Spot to the point where it overwhlemed every other Pokemon, I do believe that if it's actually broken, people will be able to use it in such a degree. How can we argue something is "broken" if over 40% of the population are totally capable of dealing with it and the other 50%+1 think it is too much? Who is playing it correctly here? If the over 40% of the population can do it, then there is no reason why other members cannot. If this proportion got narrower and narrower, then we can get a more convincing take on Garchomp - but I do think having over 40% thinking it's "just perfectly fine" is literally giving into people refusing to adjust to metagame changes (the remainder of the user). You can definitely see this trait in our wobbuffet test or even the PR posts. People do NOT like to change, and do not want to adjust to the metagame, but rather break the metagame so it's better for them.

I believe at its most basic and fundamental level that there should be no pokémon in OU (or UU) that 50% or more of the qualified, experienced population feel is too strong for that tier. That's my definition of "broken", and everything else is not. Also, by the same token, over 30% of the population can deal with Skymin according to the Stage 3-1 vote—why wouldn't the same apply here?

Secondly, your assertion implicitly claims that all the possible way of using Garchomp is already there and there is no other way of using him, and that people have already discovered every single way of dealing with him and then decided so. If this were true, then a simple majority would definitely be enough. But are you really confident that this is true? Are you really that confident with the competitive community to the point where you believe people have found every possible use of garchomp and every way to deal with it in the span of 3 months while dealing with every other suspect? I think this would definitely be difficult to argue for it - and I think because of this, it is wise for us to use a much stronger result to deal with Garchomp.

I'm confident in the unpredictability of the metagame, and how *this* implies that, without movepool changes, there is no saying how much more or less uber a pokemon will be half a year down the road. But are we really, though, supposed to subject the majority of our community to an uber pokemon for five, six, seven months, waiting for another YacheChomp to be discovered? And at what point would we even Suspect test it again? Would there also have to be some supermajority of qualified voters that you're willing to listen to when they say it might be time to make Garchomp a Suspect again, as though such a statement wouldn't come with immediate merit (as it would, forever)?

Is it really that harmful to "make sure"? I'm definitely not confident enough to say so. If what you said is true then Garchomp would have a MUCH stronger majority voting for uber on the second test, and there would be nothing to argue about.

I think it'd be a waste of time precisely because I do think I'm right. And if I'm not, "then what"? Stage 3-3? And what if Manaphy keeps getting alternating votes?

Secondly, I could definitely argue it - as Latios and Skymin are both primarily special threats. Without them in play, it allows players to put more focus on physical threats such as Garchomp. This would definitely let Garchomp be significantly less uber. I'm definitely sure someone who is actually familiar with the suspect metagame could make a stronger argument, but there definitely is a case for it.

Not a very good one, since by the very same token, Garchomp would be drawing attention away from special threats if someone actually does want to focus more on it, and this would open up the path for pokemon like Latias and Jirachi. Which "coincidentally" compliment each other about as well as any two physical or special attacking threats can, and would embody a phenomenon that these same astute players would have to notice is Garchomp's doing unless they're biased or something.

Sure, but remember - what we care about is intelligent reasons on why we should test it, not randoms whining about it on shoddy or even stark mountain. I'm pretty sure we're more than capable of finding intelligent reasons and holding little topics like you have done for wobbuffet.

I was only referring to intelligent voters. If you really want to go down the list of 44 Uber Garchomp voters, question their reasoning, and liken any number of those reasons to random whining, while also questioning the reasoning of the 41 OU votes and not being able to find any problem with the reasoning for OU, you can do so. You should remember more than anyone else that shitty reasoning is not necessarily contained to one side.

If that's really a problem then we can literally just vote for one right now. I will talk to Colin about this - does this really mean that if we did decide, this objection wouldn't matter? I don't think this is a valid objection for that reason - it just talks about something we haven't done yet.

I keep bringing it up because we're definitely not going to move forward with any of your proposals until you can decide what you all really mean by "supermajority". We're not having a repeat of the entire reason this very thread was made months after the fact. I realize that you're not "bluffing" and would actually try to have us come to some consensus on this, but I'm sure you wouldn't just be flattered to hear me say that you are one of the only people in Smogon who actually gets things like this done and why my reiterations are evidently actually necessary. "This is a good idea", "I'm all for this", "I'll help" are some of the most empty, rolleyes-inducing words anyone can ever post, and this is only trumped by people who vote in polls in Policy Review and then don't post, waiting for other people to do stuff for them.

52% with... quite a number of people not voting/abstaining. One reason why I prefer a supermajority is that it totally gets around this problem. 52% with the population we had and the people who didn't vote is reason enough that we should be looking for a stronger result, if anything, and not a reason to say "it's enough".

A supermajority would literally have made no difference with Garchomp, you could not be more wrong about this. Even if all nine of the remaining voters voted OU, it would have been 53% OU. This isn't a supermajority. If they had all voted uber, it would have been 56% uber. Still nowhere close to a supermajority. And we have virtually no reason to believe all nine of these people aren't more likely to vote 6-3 or 5-4 anyway, given the entire reason you guys are still letting "+2 Garchomp" troll you. (is this a reference to its 102 Base Speed? its two percentage points over an actual stalemate? a single swords dance?)

This is why I say "i dont care" when asked if I've come to a conclusion on what impact the abstainers and non-voters (accounts that weren't originally claimed) have on the final vote. In conjunction with the fact that 85 qualified voters is a very large number, I am not going to wait around for a few people to straggle in and finally post in a thread posted by an administrator, stickied by an administrator, referenced in an Announcement by another administrator, placed into the Topic of several channels on IRC, and /walled on the Shoddy Server they spent three months demonstrating they cared about the tiering of competitive pokemon on an alternate ladder.

 

[Tangerine]

I believe at its most basic and fundamental level that there should be no pokémon in OU (or UU) that 50% or more of the qualified, experienced population feel is too strong for that tier. That's my definition of "broken", and everything else is not. Also, by the same token, over 30% of the population can deal with Skymin according to the Stage 3-1 vote—why wouldn't the same apply here?

Let's put it this way.

Given a metagame with a set of Pokemon, there may be a Pokemon that may be broken. We don't know quite yet how to explain broken in a precise definition, so we look at the signs of something being broken. We can't tell you why something is broken, although we can give you symptoms (Portraits of Uber) that may tell you why something is considered broken.

One of the ways we get around this problem of inherent brokenness was to use this voting scheme. In fact, that's precisely the only reason why we have this system currently, because we acknowledge the fact that we don't have a completely objective definition of broken. One of the signs that we look at, of course, is to let "experienced" and skilled players vote, and let them think for themselves if it is broken depending on whatever definition of Uber they have.

Do you see the problem yet? The problem is that your definition of "broken" is something that is CORRELATED with Broken, rather than actually broken. Correlation in no way implies causation, although very, very, strong correlation may give us a pretty good picture of what is going on.

Your definition of broken is similar to going around, getting a bunch of people who read newspapers and keep up with the news (news without any information about economy, just for this example), asking them how they feel about the economy, and declaring a recession if more than 50% of the people believe that the economy is doing badly, despite the fact that they have little idea on what is really going on. 50%, in this case, is far too weak. When the percentage goes higher and higher, that's when we can use this popular polling method to ensure that we have a recession on our hands, just like we need a much stronger correlation to determine if something is broken or not.

In essence, I believe your definition of broken is far too weak to accurately depict the metagame. The differences between our viewpoints is that your definition of uber comes directly only from opinions, and mine comes from the Pokemon actually being broken. I have highlighted other flaws of using purely opinions of other users as the DEFINITION of broken, and these flaws make it more likely that we need a much stronger correlation before we can completely decide in one way or another. I will not repeat the arguments here, as you're most likely familiar with it by now.

But are we really, though, supposed to subject the majority of our community to an uber pokemon for five, six, seven months, waiting for another YacheChomp to be discovered?

Again, it's not actually uber. Just because 50%+1 people think something is uber does not make the pokemon inherently uber, so let's not use these terms to try and bolster your argument.

At this point, it "may be" uber, and just needs a nail in the coffin either way. If the Pokemon is ACTUALLY uber, then we won't need five, six, seven months, as more people will find that you can't deal with it, or someone finds a more devastating way to utilize it. If the Pokemon isn't actually uber, then what happens? then people find ways to deal with it, and less and less people will think it's uber and we'll be all happy.

I'm confident in the unpredictability of the metagame, and how *this* implies that, without movepool changes, there is no saying how much more or less uber a pokemon will be half a year down the road.

Yes, but what is more likely to be less uber - a Pokemon voted 70% uber, or a Pokemon voted 50%+1 uber, half a year down the road? You can bet that the 50%+1 uber Pokemon will more likely be "less uber", as the 70% uber pokemon has strong evidence letting us believe it is broken even under that short timespan.

And at what point would we even Suspect test it again? Would there also have to be some supermajority of qualified voters that you're willing to listen to when they say it might be time to make Garchomp a Suspect again, as though such a statement wouldn't come with immediate merit (as it would, forever)?

We have a lot of active, qualified players who are more than capable of letting people know - much like what happened when people discovered Deoxys S's dual screen set - people will come, and bring it up, and if this seems to be significant with no major disagreements, then we can decide it.

And if I'm not, "then what"? Stage 3-3? And what if Manaphy keeps getting alternating votes?

I don't think we should be using "what ifs" as an argument - although I would argue if anything continues to get alternative votes then it's a pretty good sign that the decision on manaphy is "ambiguous" and we should let it be OU.

Not a very good one, since by the very same token, Garchomp would be drawing attention away from special threats if someone actually does want to focus more on it, and this would open up the path for pokemon like Latias and Jirachi. Which "coincidentally" compliment each other about as well as any two physical or special attacking threats can, and would embody a phenomenon that these same astute players would have to notice is Garchomp's doing unless they're biased or something.

Which would mean it's worth testing again, at the very least, to test hypothesis such as these.

I was only referring to intelligent voters. If you really want to go down the list of 44 Uber Garchomp voters, question their reasoning, and liken any number of those reasons to random whining, while also questioning the reasoning of the 41 OU votes and not being able to find any problem with the reasoning for OU, you can do so. You should remember more than anyone else that shitty reasoning is not necessarily contained to one side.

sure thing. my PM box is open.

A supermajority would literally have made no difference with Garchomp, you could not be more wrong about this. Even if all nine of the remaining voters voted OU, it would have been 53% OU. This isn't a supermajority. If they had all voted uber, it would have been 56% uber. Still nowhere close to a supermajority. And we have virtually no reason to believe all nine of these people aren't more likely to vote 6-3 or 5-4 anyway, given the entire reason you guys are still letting "+2 Garchomp" troll you. (is this a reference to its 102 Base Speed? its two percentage points over an actual stalemate? a single swords dance?)

Remember - when we talk supermajority, we only talk supermajority in ONE DIRECTION - the direction for it being uber. So it would have 100% made a difference. Garchomp would still be OU under supermajority.

also very well done.

This is why I say "i dont care" when asked if I've come to a conclusion on what impact the abstainers and non-voters (accounts that weren't originally claimed) have on the final vote. In conjunction with the fact that 85 qualified voters is a very large number, I am not going to wait around for a few people to straggle in and finally post in a thread posted by an administrator, stickied by an administrator, referenced in an Announcement by another administrator, placed into the Topic of several channels on IRC, and /walled on the Shoddy Server they spent three months demonstrating they cared about the tiering of competitive pokemon on an alternate ladder.

Sure - but you still have to have to admit that the probability is quite significant that the 9 missing votes would have definitely made a difference in the tiering.

 

[Cathy]

Again, everyone has yet to address the very reasonable possibility that Manaphy get any sort of majority for uber in Stage 3-2. "Then what?" And what if I'm wrong about the remaining Suspects standing to be deemed more uber in Stage 3-2, or those evidently crucial nine voters were all on the "free garchomp" bandwagon from the beginning and Garchomp gets an OU stage 3-2 tag? This is what I'm talking about when I state repeatedly that my concerns are not being addressed and we cannot move on until they are.

Manaphy was voted OU. The idea with these proposals is that a supermajority is required to declare a pokemon uber, not OU. If the supermajority to declare the pokemon uber is not reached, then it isn't considered uber under these proposals. No majority is required for the pokemon to be considered "not uber" under these proposals. You might disagree with this proposal, but it's the proposal we've been discussing, and Manaphy is not a special case under it.

The relevant part of the opening post is:

As a corollary of supporting [a supermajority] condition, if fewer people than the chosen ratio vote the suspect uber, it is not considered to be uber yet.

This element of these proposals is essential, and it's also why a supermajority would have made a difference for Garchomp.

I keep bringing it up because we're definitely not going to move forward with any of your proposals until you can decide what you all really mean by "supermajority".

It's not as though we all mean the same ratio, and that's not a flaw. A given person might support a particular interval of "critical ratios" to use. For example, I might support the use of anything in between 60% and 80%. Another person might support the use of anything in between 55% and 65%. So obviously anything in (60%, 65%) would satisfy both of us. It won't be difficult to discern which interval is supported by the most people, and I'm posting two threads to do so now. (My original plan was to decide on proposal #2 or #3 first and then decide on a ratio, but since we haven't decided on either of them yet, it will require two separate discussions.)

 

[Jumpman16]

Let's put it this way.

Given a metagame with a set of Pokemon, there may be a Pokemon that may be broken. We don't know quite yet how to explain broken in a precise definition, so we look at the signs of something being broken. We can't tell you why something is broken, although we can give you symptoms (Portraits of Uber) that may tell you why something is considered broken.

One of the ways we get around this problem of inherent brokenness was to use this voting scheme. In fact, that's precisely the only reason why we have this system currently, because we acknowledge the fact that we don't have a completely objective definition of broken. One of the signs that we look at, of course, is to let "experienced" and skilled players vote, and let them think for themselves if it is broken depending on whatever definition of Uber they have.

Do you see the problem yet? The problem is that your definition of "broken" is something that is CORRELATED with Broken, rather than actually broken. Correlation in no way implies causation, although very, very, strong correlation may give us a pretty good picture of what is going on.

Your definition of broken is similar to going around, getting a bunch of people who read newspapers and keep up with the news (news without any information about economy, just for this example), asking them how they feel about the economy, and declaring a recession if more than 50% of the people believe that the economy is doing badly, despite the fact that they have little idea on what is really going on. 50%, in this case, is far too weak. When the percentage goes higher and higher, that's when we can use this popular polling method to ensure that we have a recession on our hands, just like we need a much stronger correlation to determine if something is broken or not.

In essence, I believe your definition of broken is far too weak to accurately depict the metagame. The differences between our viewpoints is that your definition of uber comes directly only from opinions, and mine comes from the Pokemon actually being broken. I have highlighted other flaws of using purely opinions of other users as the DEFINITION of broken, and these flaws make it more likely that we need a much stronger correlation before we can completely decide in one way or another. I will not repeat the arguments here, as you're most likely familiar with it by now.

It's actually a lot more simple than this. I said "broken" in quotes because I know the only real word we should be using is uber. This in turn should only be thought of in terms of the Characteristics we have all decided on, and in turn all the voters (who are all qualified and experienced) understand that they are only to vote with those Characteristics in mind.

So when I say "broken", I actually meant uber, sorry if I made you think I wanted to discuss the merits/usage of the word. Again, I strongly feel that if more people feel something is uber than those who do not, then it does not belong in standard, and vice versa. The vote could be 301-300 in favor of uber and I would feel the same way...I'd actually feel more confident in making it uber because "statistical anomaly" could be cited as the difference between uber and ou before anything else.

Again, it's not actually uber. Just because 50%+1 people think something is uber does not make the pokemon inherently uber, so let's not use these terms to try and bolster your argument.

That's actually exactly what my argument is, though. I'm obviously in the 50% + 1 camp. It is black and white—you're either in OU or you're not. More testing doesn't make sense to me if I feel our voting pool was both large enough and very pure.

At this point, it "may be" uber, and just needs a nail in the coffin either way. If the Pokemon is ACTUALLY uber, then we won't need five, six, seven months, as more people will find that you can't deal with it, or someone finds a more devastating way to utilize it. If the Pokemon isn't actually uber, then what happens? then people find ways to deal with it, and less and less people will think it's uber and we'll be all happy.

I really don't think so. First, a nail in the coffin being 55% instead of 52% isn't really a much sharper nail. Second, I already stated that I think people will always feel Garchomp is a borderline pokemon, I don't think it will every go more than 2/3 either way...nor do I think it has to in order for us to say that it should either stay or go.

Yes, but what is more likely to be less uber - a Pokemon voted 70% uber, or a Pokemon voted 50%+1 uber, half a year down the road? You can bet that the 50%+1 uber Pokemon will more likely be "less uber", as the 70% uber pokemon has strong evidence letting us believe it is broken even under that short timespan.

I'm not worried about that as much as I am having an uber in standard play affecting the competitive metagame for those same six months while some as-yet-undefined number of people determine it's finally ok to redeclare it a Suspect (which will always be reasonable for reasons I've already stated).

We have a lot of active, qualified players who are more than capable of letting people know - much like what happened when people discovered Deoxys S's dual screen set - people will come, and bring it up, and if this seems to be significant with no major disagreements, then we can decide it.

Yeah except for where this took over four months in the case of both Garchomp and DXS, and standard play was greatly affected by their presence because of this. And I don't think either pokemon remained under "49%" OU the entire time they were allowed in Standard up until their respective gamebreaking sets were discovered.

I don't think we should be using "what ifs" as an argument - although I would argue if anything continues to get alternative votes then it's a pretty good sign that the decision on manaphy is "ambiguous" and we should let it be OU.

The what ifs aren't an argument or me playing devil's advocate. For about the fourth time now, we are not going to have a repeat of the reason this thread was made, which is why percentages and rules needed to be decided on now so that "if" Manaphy gets alternating votes we will take action that isn't just convenient 20/20 hindsight action.

Which would mean it's worth testing again, at the very least, to test hypothesis such as these.

Maybe the reason Latios was considered much more uber than he was in Stage 1 (by 18% points) is because in Stage 3 everyone had to worry about Garchomp and its physical threat too. Why not retest Latios by the same logic to make sure this isn't true?

sure thing. my PM box is open.

We talked about this on IRC but I'm not really bluffing...if you really think you can get 85 voters to all voluntarily PM you their reasoning on Garchomp (and Manaphy too as it also did not reach a canonical supermajority) and then judge them all by yourself and stay confident that this would strengthen your argument, good luck.

Remember - when we talk supermajority, we only talk supermajority in ONE DIRECTION - the direction for it being uber. So it would have 100% made a difference. Garchomp would still be OU under supermajority.

No, it makes 0% difference because of what you just said. The nine votes would not have had it reach a 67% supermajority either way which means you and Colin should really stop citing those nine voters as a difference maker in the Garchomp vote. The person they could have made a difference for would be someone who only wants a 50% + 1 majority, and that's me...but as stated we can't assume that more than six would have voted OU and not made it reach a 50% majority in the direction for it being uber. And as stated, "i dont care" about why nine voters didn't vote in an 85-voter pool. I certanly don't expect this to definitely have no chance of happening in future votes, which is something no one else seems to be afraid of—I'd love to see everyone's reaction if the next Garchomp vote is 41-43 or something with six voters who didn't vote. Because how utterly more convincing this would be than the Stage 3-1 vote.

Sure - but you still have to have to admit that the probability is quite significant that the 9 missing votes would have definitely made a difference in the tiering.

I most certainly do not, why would I admit this? I've said a few times now that by the very nature of the close vote gives me or anyone else virtually no reason that seven or more of the nine voters would vote OU instead of 5-4 or 6-3 either way.

Manaphy was voted OU. The idea with these proposals is that a supermajority is required to declare a pokemon uber, not OU. If the supermajority to declare the pokemon uber is not reached, then it isn't considered uber under these proposals. No majority is required for the pokemon to be considered "not uber" under these proposals. You might disagree with this proposal, but it's the proposal we've been discussing, and Manaphy is not a special case under it.

I think this is fundamentally why I disagree with the concept of supermajorities, or am at least trying to lay the groundrules now for our sake. If we decide on 2/3 + 1 and Manaphy is voted 62% uber in Stage 3-2, "then what?"

 

[Tangerine]

It's actually a lot more simple than this. I said "broken" in quotes because I know the only real word we should be using is uber. This in turn should only be thought of in terms of the Characteristics we have all decided on, and in turn all the voters (who are all qualified and experienced) understand that they are only to vote with those Characteristics in mind.

So when I say "broken", I actually meant uber, sorry if I made you think I wanted to discuss the merits/usage of the word. Again, I strongly feel that if more people feel something is uber than those who do not, then it does not belong in standard, and vice versa. The vote could be 301-300 in favor of uber and I would feel the same way...I'd actually feel more confident in making it uber because "statistical anomaly" could be cited as the difference between uber and ou before anything else.

So basically this suspect test isn't about what's actually broken but what people want done on their metagame for "whatever reason".

I can't say I agree with this - but if this is how you have ran the test, then I can't really criticize it anymore.

I'm not worried about that as much as I am having an uber in standard play affecting the competitive metagame for those same six months while some as-yet-undefined number of people determine it's finally ok to redeclare it a Suspect (which will always be reasonable for reasons I've already stated).

Why should you worry about having an "uber" pokemon in the standard metagame, when uber != broken? If its sheerly because of publicity, then sure. If it's for the actual true metagame, then your worries are wrong.

Yeah except for where this took over four months in the case of both Garchomp and DXS, and standard play was greatly affected by their presence because of this.

Can you really say it was affected for the worse or for the better? Sure, there was an effect - but that is irrelevant.

The what ifs aren't an argument or me playing devil's advocate. For about the fourth time now, we are not going to have a repeat of the reason this thread was made, which is why percentages and rules needed to be decided on now so that "if" Manaphy gets alternating votes we will take action that isn't just convenient 20/20 hindsight action.

Basically, your idea is that we must plan everything out to the end right now, and not allowed to make exception for any cases on the risk of being biased. It has its merits - but it means that we have to get it right the first time, which is hardly the case. I don't see why we can't "improve" upon the process like we have done many times in this entire ordeal.

Maybe the reason Latios was considered much more uber than he was in Stage 1 (by 18% points) is because in Stage 3 everyone had to worry about Garchomp and its physical threat too. Why not retest Latios by the same logic to make sure this isn't true?

Because none of Latio's lines are borderline - this is completely irrelevant.

We talked about this on IRC but I'm not really bluffing...if you really think you can get 85 voters to all voluntarily PM you their reasoning on Garchomp (and Manaphy too as it also did not reach a canonical supermajority) and then judge them all by yourself and stay confident that this would strengthen your argument, good luck.

I thought we agreed that it wouldn't be fair to the voters (and really who is going to voluntarily do shit in this community anyway?)

No, it makes 0% difference because of what you just said. The nine votes would not have had it reach a 67% supermajority either way which means you and Colin should really stop citing those nine voters as a difference maker in the Garchomp vote. The person they could have made a difference for would be someone who only wants a 50% + 1 majority, and that's me...but as stated we can't assume that more than six would have voted OU and not made it reach a 50% majority in the direction for it being uber. And as stated, "i dont care" about why nine voters didn't vote in an 85-voter pool. I certanly don't expect this to definitely have no chance of happening in future votes, which is something no one else seems to be afraid of—I'd love to see everyone's reaction if the next Garchomp vote is 41-43 or something with six voters who didn't vote. Because how utterly more convincing this would be than the Stage 3-1 vote.

That's irrelevant because the supermajority only applies for it to be uber... so I don't know why you keep bringing it up.

I most certainly do not, why would I admit this? I've said a few times now that by the very nature of the close vote gives me or anyone else virtually no reason that seven or more of the nine voters would vote OU instead of 5-4 or 6-3 either way.

Of course, but the point is, there exists a significant enough chance for us to be worried about - at least, those who think of this process as something other than a process where "competitive" players can just remove whatever they don't like in the metagame for whatever reason despite the fact that it's not actually broken.

 

[Jumpman16]

So basically this suspect test isn't about what's actually broken but what people want done on their metagame for "whatever reason".

I can't say I agree with this - but if this is how you have ran the test, then I can't really criticize it anymore.

This Suspect Test isn't about what's broken, you should never have started using that word in the first place. And yes, it is about what people want done on their metagame. In every iteration of the test except Garchomp and Skymin, we have required those people to post that reason so I and Aeolus and even yourself can critique it. The pool is four times its average size, and the purest we've ever had given this test has ensured all the voters actually experienced the Suspects. These, again, mean that Aeolus and I do not think it's necessary to poll the reasoning of the voters.

Why should you worry about having an "uber" pokemon in the standard metagame, when uber != broken? If its sheerly because of publicity, then sure. If it's for the actual true metagame, then your worries are wrong.

Stop using the word "broken" and my meaning won't be misconstrued. The true metagame does not include any pokemon that the majority—any majority—of qualified players feels is uber.

Can you really say it was affected for the worse or for the better? Sure, there was an effect - but that is irrelevant.

Yes I can, as the true metagame does not include any pokemon that the majority—any majority—of qualified players feels is uber.

Basically, your idea is that we must plan everything out to the end right now, and not allowed to make exception for any cases on the risk of being biased. It has its merits - but it means that we have to get it right the first time, which is hardly the case. I don't see why we can't "improve" upon the process like we have done many times in this entire ordeal.

Wait, so I can take shit both for allegedly exhibiting bias in previous Suspect Tests and for definitely attempting to prevent it in future ones? Seems like a lose-lose to me!

And this is not an improvement on the process if we aren't willing to state crucial things right now, for "whatever reason", like what happens if manaphy is voted uber in Stage 3-2 or if Garchomp is voted OU.

Because none of Latio's lines are borderline - this is completely irrelevant.

It's only as irrelevant as the argument for Garchomp being less uber with less Suspects is (the main one being that 50% + 1 is good enough by my own argument).

I thought we agreed that it wouldn't be fair to the voters (and really who is going to voluntarily do shit in this community anyway?)

If you're going to maintain that the fact that we don't know our voters' reasoning for this round is the reason that Garchomp was not voted OU then I don't have a choice but to call your "bluff". And I doubt you actually care what's fair for our voters with you feeling you needed to remind all of us that your PM box isn't closed.

That's irrelevant because the supermajority only applies for it to be uber... so I don't know why you keep bringing it up.

Do you honestly mean to say that if Skymin had had just three of its votes go for OU instead of uber then you all would be clamoring for it to be included in Stage 3-2 as well? And six for Latios?

Let's not kid ourselves here—the only reason we even have this thread and the corresponding ratio poll threads is because 44-41 isn't enough for some of you guys (after the fact, no less). Which is exactly why deciding on some supermajority anywhere between 55% and 95% after the fact is really kind of ridiculous and not the way I want policy run.

Of course, but the point is, there exists a significant enough chance for us to be worried about - at least, those who think of this process as something other than a process where "competitive" players can just remove whatever they don't like in the metagame for whatever reason despite the fact that it's not actually broken.

No, there really is not a significant chance at all. At all. Unless you and the rest of you on the supermajority bandwagon are willing to admit that these same nine votes has as much of an impact on whether both Latios and Manaphy should be tested in Stage 3-2 as well, which I for some reason am the only person to realize.

 

[Hipmonlee]

his Suspect Test isn't about what's broken, you should never have started using that word in the first place. And yes, it is about what people want done on their metagame. In every iteration of the test except Garchomp and Skymin, we have required those people to post that reason so I and Aeolus and even yourself can critique it. The pool is four times its average size, and the purest we've ever had given this test has ensured all the voters actually experienced the Suspects. These, again, mean that Aeolus and I do not think it's necessary to poll the reasoning of the voters.

This test isnt about what people want, or else you would never have had to have polled anyones reasoning. In this context people are using the word broken to mean "far too powerful to be in line with a balanced metagame" which is exactly what this suspect test is about.

Yes I can, as the true metagame does not include any pokemon that the majority—any majority—of qualified players feels is uber.

I am a qualified player, and if you were to poll only me we would have an extremely biased majority that nobody should accept. The point here is that any poll is imperfect (some moreso than others). But that it is better to err on the side of more testing when there is a reasonable likelihood that the vote gave a false result (based on our definition of uber).

And this is not an improvement on the process if we aren't willing to state crucial things right now, for "whatever reason", like what happens if manaphy is voted uber in Stage 3-2 or if Garchomp is voted OU.

We keep testing until no suspects are voted uber.

If a pokemon cant get two simple majorities in a row on this test then it clearly isnt "very apparent" enough that it is "far too powerful to be in line with a balanced metagame".

The argument for a supermajority is essentially the same except it is slightly less clear that it isnt "very apparent" enough, but obviously still clear enough for a lot of people (this isnt the course of action I support here).

No, there really is not a significant chance at all. At all. Unless you and the rest of you on the supermajority bandwagon are willing to admit that these same nine votes has as much of an impact on whether both Latios and Manaphy should be tested in Stage 3-2 as well, which I for some reason am the only person to realize.

But manaphy will be retested and latios was voted uber by more than 9 votes. The point of the supermajority is to protect from this sort of thing, to increase the likelihood (to a point where there are reasonable doubts as to its legitimacy) that, even though the poll (like all polls) is imperfect, the pokemon will only be considered uber if it meets our definition of uber.

Have a nice day.