New UU - Order of Operations

[Caelum]

Given there was some confusion over this and there was no definitive reference for people to refer to, Doug felt it was appropriate so here it is. Probably won't be as eloquent as Jump's but alright.

The process follows the same "cast" or "mold" that OU is intended to follow so it might be helpful to refer to the "Order of Operations" thread here before reading this. Additionally, it would be helpful to read Obi's PR Thread: "My Suggestion for UU."

Background Information:

Initially, we had the case of 50 (give or take) theorymon BLs and then UU. The goal was to figure out appropriately if any of the BLs would be acceptable in UU play. Testing each and every BL individually in a rigorous manner would be great, but the practicality of that is obviously nonexistent. So, instead, BLs (and NFE's) were removed entirely and essentially everything not in Standard or Ubers was considered UU. From this we needed to collect a group of "potential BLs", suspects, which I'll explain before where it is now.

Completed Stages:
1- BLs & NFEs are dropped into UU

2- Then testing of the metagame began and it actually solidified rather quickly and a fairly settled tier had developed at the end of the first month.

3- Then, a Stark thread was open (go look for yourself) that allowed any users to nominate what they considered to be potentially broken. The arguments were required to be accompanied by sufficient reasoning using the "characteristics of an uber" (with obviously BL supplemented in place of uber). Myself, RB-Golbat, and (now) Tangerine judged the validity of the arguments and we are in the process of revising our edits into the posts but they are overall completed. This led to the identification of them (6 currently).

To Be Completed Stage (feel free to give feedback before any of these are implemented: criticisms, comments, concerns etc):

4 - The identified suspects in stage 3 would be removed from the UU standard ladder and that metagame would unfold for a month.

5- At the end of that month, each suspect identified in stage 3 would be subject to vote to those that met the rating requirements of 1655/55 using their experiences with and without the presence of suspects to judge the validity of them in the UU tier.

6- Then stage 3 would be repeated to determine if there are anymore suspects that exists; the process would be repeated as needed (and may occur at random times due to certain OU Pokemon dropping in usage and becoming BL).

Other Notes:

- For Stage 3, it was determined once the nominations were combed through and poor ones removed they would be tallied. Then, those suspects with only 1 nomination would be thrown out of the pool. Finally, any nomination that satistified approximately 10% (give or take + or - 2%) consensus would be considered a suspect for this round.

- Comment on this especially: All of the suspects were originally planned to be removed at once. That was the original premise because we were working under the assumption we would have a large number of potential BLs and for the sake of time, that would've been most practical. However, we actually have very few potential BLs so this can change entirely to analyze them individually like in OU, thoughts? That would mean we would likely need a suspect UU ladder as well. With only 6 it becomes feasible to use a separate ladder (particularly with the high activity on the current UU ladder). So the point below may not apply. This is mostly being considered because of the necessity of a "rating reset" - do we want that?

- We ultimately decided to just use the UU ladder for this purpose, rather than creating a seperate one due to possible lack of inactivity and conflicting with the suspect process of OU.

- The intention was to reset the ratings on the UU ladder in the next few days, is that appropriate and if not, is there another alternative we can go down?

Anyway, post questions / concerns / criticisms whatever.

 

[Seven Deadly Sins]

Comments.

First off, I think with the level of activity, a one-week test (or maybe just 5 days) can be viable for each suspect. I think that in the case of these suspects, since each suspect has a huge impact on the others, I think the following order would be best.

1. Abomasnow: Has the widest metagame impact, and also has the least impact on all other suspects. A major priority.

2. Crobat: As an incredibly fast revenge killer and fighting-type counter, as well as speedy Taunt or Weather user, its impact on the metagame is huge and, in my opinion, affects the tiering of other suspects (such as Gallade and Froslass).

3. Froslass: If Crobat leaves, Froslass will probably be absolute horseshit in the metagame, and as such will be a priority.

4. Raikou: Being able to outspeed Staraptor is huge in the metagame, and as such its tiering impacts Staraptor's tiering.

5: Staraptor: Since it outspeeds Gallade and is immune to Shadow Sneak, it is a major revenge killer to Gallade, and as such, affects Gallade's tiering immensely.

6. Gallade: Its usage impacts all the other suspects' usage the least, and as such, is most likely to have the least impact on other suspects' tiering while being most impacted by the other suspects, and as such, should be tested after all the other suspects have been evaluated.

Next, I highly believe that ratings SHOULD be reset on the UU ladder mainly because of the fact that the current ratings have major holdovers from the old UU ladder, such as the broken "trolly" rating and Eo Ut Mortus' #1 place in the ladder.

 

[darkie]

The ratings of UU players should definitely be reset when suspect testing is started. I believe we established that this was a mistake with Shaymin-S - as such, we really shouldn't do that again.

Since there are only six suspects this time, I don't see a problem with a separate ladder. You know as well as I do that UU is immensely popular and I don't see it dwindling down simply because of a second ladder.

On the other hand, it may become time consuming to test each suspect seperately. A single week is really not enough to assess the significance of a suspect to the metagame. I would vouch for the tests to be two weeks at the very least. It takes at least that long - if not more - for a healthy environment to be established.

As far as order (if we do decide to go with separate tests), I don't think it really matters. Every single point that SDS made about a certain suspect impacting others can be reversed. Staraptor affects Raikou just as much as vice versa, and for the same reasons, too! If Staraptor were ousted, then Raikou's usage might decrease since outspeeding Staraptor wouldn't be as much of a priority. (I'm just being a devil's advocate here).

 

[RBG]

Here is what I think:

Ratings Reset is the best method

We only want 1 UU ladder. The problem isn't participation. It is that currently, the OU suspect tests are more important and that suspect ladders already have lower levels of participation than the regular ladders. I don't feel that dividing the users on even more ladders and begging for even more ladder use is what we want.

Also, I am for doing all the suspects at once. There is no "right" or "best" order to test them. This is 100% objective when you get down to it.

 

[Great Sage]

I think we should remove all the suspects at once, and in the second iteration of the UU test put in provision for potentially unbanning something we ban the first time. The ratings should definitely be reset; the ratings so far aren't necessarily representative of how well people really play, since much of the time so far has been experimentation with a completely new metagame.

 

[DougJustDoug]

Philisophically, I don't like the idea of erasing seven months worth of UU playing history and saying that it is irrelevant for rating good battlers. There are many players who have consistently battled for several months to achieve a certain rating, and to achieve a certain deviation. I'm not sure it's fair to throw it away, just because the UU tier has changed.

We didn't reset the ladder when Weezing, Aerodactyl, and other former-BL's were moved down. I admit that change was not as big as the recent change to UU, but it was certainly significant.

The UU ladder is very different than it used to be, but it's also still the same -- it is the "ladder for pokemon below OU". In the sense that it is still the "less-than-OU ladder", you could argue that player ratings should still apply. If the new pokemon invalidate old UU teams, and the old UU players cannot adapt, then their rating will decline. But, to summarily wipe out all that rating history because "it's no longer relevant" -- that doesn't seem right.

As a practical matter, it may be no big deal. Many players use new alts all the time. For all I know, most UU players may be looking forward to a reset. Perhaps I am making a mountain out of a molehill. I'd like to hear from entrenched highly-ranked players from the "old UU" to see what they think. Will they feel slighted if their rating history was completely erased? If not, then I'll stop tilting at windmills...

 

[Caelum]

Well, like Great Sage said, if we went the route he described there would be a reevaluation period down the line.

The only reason I even started to consider a separate ladder was because of Doug's concerns, is that "fair" essentially? I'm not a good person to decide that because I honestly could care less what my rating is, but to some people it does matter so that has to be considered.

 

[Legacy Raider]

Instead of erasing the ratings of all players on the UU ladder, we should consider asking all the players who plan on trying to reach voting eligibility to create new accounts on the UU ladder. This way, we have fresh new ratings for everyone who is going to vote, without deleting the ratings of old time players who might not be wanting to participate in this phase of the test for their own reasons. If we can ask people to create accounts with "___ UUTest" or something, we can determine what these new accounts are quite easily, and then it is simply a matter of checking if these newly made accounts have the right deviation and rating when it comes down to voting time.

 

[Chou Toshio]

On the Subject of Individual Testing: I really have mixed feelings about this for a number of reasons. Of course there's all the reasoning that goes into the individual testing for OU. For UU though, my anti-individual testing thoughts would be:

1- In my opinion, 6 suspects really is a pretty good number that we have to go through, and my guess is that we are likely to be adding to that list in the future. Time should be a concern in the "suspect process" even though we tend to ignore any need for haste in OU.

2-As much as we like to say UU is the "second balanced tier," the fact is that it will always be kind of at the mercy, so to speak, of what happens in OU. Yes, pokemon falling from OU to UU is a separate process (if Alakazam drops, it would be UU, not suspect until labled one as per stage 3), but the result is we are always going to have some inate instability as to what is UU. For that reason, there's a bit of incentive to test things more quickly.

However, as mentioned I would also agree with the merits of having individual testing. Dropping everything in was good for labelling suspects, but there is certainly reason to test them individually.

On the subject of Ratings:

I feel that "The right to play the ladder" and "the right to take part in the test" are two separate issues. I suggested this once in another thread, but if you are concerned with "fairness," why not just have everyone who wants to participate in the test register new accounts at the start of it? You could have them all confirm these new accounts in a thread in Stark Mountain.

edit: Damn LR beat me to it. xD

 

[DougJustDoug]

If people want to play under their familiar Shoddy alt, we could collect a "sign-up sheet" for UU testing, and I can reset just those players that sign up and request it. If I did that, it would be a one-time thing though. I would not want to get requests every day for ratings resets...

 

[QibingZero]

Actually, I think participation is a concern. A large part of the reason we were even able to make quality UU suspect nominations after such a short amount of time was the fact there was so much activity on the UU ladder. On the other hand, I had serious issues attempting to decipher anything substantial while playing the Latias suspect ladder, due to the lack of participation. When there are more people playing, and people are playing consistantly throughout the entire test, the results are significantly more telling. In the past month on the UU ladder I have been able to play enough games - against enough teams - in order to feel that I have been able to mold a quality team myself, and, along with that, a very good picture of the state of the metagame. This is largely because I've been able to make good adaptation after seeing how my team fares after a group of battles - something the lack of participation on the Latias suspect ladder prevented me from doing. We've seen how long it can take for certain pokemon to be used to their full potential in the most active metagame we have available, and in that respect it seems that the single month of suspect testing (which had something like 1/16th of the battles that standard did in the same time, at best) isn't always going to have the desired effect.

I'm not saying this is a direct result of having a separate ladder, but I do think it comes into play significantly. It has to be said: the average user is more likely to play the metagame they enjoy the most when given the choice. While this is obviously not a popularity contest, enjoyment affects even the most dedicated Smogonite. For me, personally, it's very difficult to be motivated to play when I can't get games after sitting in the queue for over a half hour and mentioning things in the chat, and when I can't make adjustments to my team due to the fact I'm not facing enough variety.

It's mostly for this reason that I prefer not having either a separate UU suspect ladder or to test suspects one at a time. Further splitting the smaller pool of UU players (1/20th of the OU battles before January, 1/5th of the OU battles last month) just seems like it would create a situation wherein we have the same problems I mentioned in the paragraphs above. Furthermore, regarding individual testing, I don't see how it would even be done unless we had a separate ladder. There is also the issue that testing the suspects individually will only draw out a process, that many believe, will still be going on after these suspects are decided upon. The biggest point is that of Clefable, whom no one (myself included) was able to make an acceptable argument for, but of whom also a large amount of the UU community actually believes is suspect. Shaymin is much the same.



As far as ratings go, I have absolutely no problem with the ratings being reset. Ratings are regularly reset on many ladders as 'seasons' go by, as it is, and a substantial change to the ladder itself should certainly be considered as a 'change of season'. Now, I personally have been able to maintain a high rating on the UU ladder, and my concern is that it would give me a huge advantage in the suspect test. Now an advantage isn't a huge problem, but the fact that I would literally have to play only a handful of battles to make the required rating is. I would then have absolutely no knowledge of the metagame without the suspects, and yet would still be eligible to vote. That becomes even more of a problem when you realize that many players with high ratings have been (and this is very subjective) 'abusing hail teams' to achieve such high ratings in the first place. My intention isn't to say many hail-team players are not, at the same time, good players who are capable of making coherent suspect arguments. It's just that my personal judgment when playing the ladder was that hail-team ratings did not seem to usually match up with how good I considered the individual players to be.

The problem I have with just creating alts to test on the UU ladder is that it still gives off the impression that each player's rating in UU with the suspects is equivalent to theirs without the suspects. Personally, I feel that I might actually be a much different UU player with these suspects removed (and, as I've callously noted, I feel the same about other players too).

 

[reachzero]

This post will have two sections: first, my views on the overall testing process. Secondly, my suggested order to test the Suspects if we decide to test them separately.

Time and participation are both major concerns in the UU Suspect (henceforth sUUspect) testing process. We don't want to take too long in sorting out the sUUspects; several of them are already fairly clear-cut, even "obvious". Since I believe a few of these are genuinely harmful to the UU metagame, getting them out ASAP is really important, in addition to the simple practical concern of wanting to reach the second round of testing sooner. If the testing will last a month, it is essential that we test all six sUUspects at once. Taking a month on each sUUspect would be terrible overkill, and would kill most of the buzz that nUU has generated. I favor testing on the normal UU ladder, since a Suspect UU ladder would also split the UU battles into annoying levels. If the OU Suspect test ladder was quiet, how much quieter would a UU Suspect ladder be (the obvious alternative is that the regular UU ladder would run dry, but then what is the point of having a Suspect ladder?)? More to the point, there is no sense in starting a separate sUUspect ladder because the nUU metagame as we have been playing it is not a "standard" metagame. It was already an "experiment", a Suspect ladder of a sort, designed to allow us to identify potential BLs. It has already at least partially accomplished that mission, and now there isn't really any reason to iconize it as some sort of baseline standard. In terms of dealing with eligibility and ratings, I see no reason not to ask people to create special test alts, as Legacy Raider as already suggested. If we do choose to test our sUUspects one at a time, I think that a shorter time span would be advisable (perhaps two weeks), so that the really bad sUUspects can be ushered out ASAP and the process moved along.

Now, if we do decide to test one by one, this is the order that I would prefer to test in. This list generally is in the order that I view each sUUspect as harming the metagame.

1. Abomasnow: Hail teams are both annoying and overpowered. The sooner we can get Auto-Hail out of the picture, the better off the metagame will be.

2. Staraptor: Probably the most thought-of threat in the current UU metagame, it practically defines whether a UU Pokemon is thought of as "fast", "slow", or "average". Staraptor leaving the UU metagame will probably change quite a bit.

3. Froslass: Even without Auto-Hail, Froslass makes Spikes so easy to get out that she can drastically alter a match within the first 4-5 turns of the game. However, there is a big gap in my mind between Staraptor and Froslass; Froslass, Raikou and Gallade could be scrambled without it making as big a difference as it would to put one before Abomasnow or Staraptor.

4. Gallade: Does not so much change the metagame as play a particular role much better than any other Pokemon able to play the same role. As such, I see it as low priority, even though it is a powerful threat.

5. Raikou: Raikou is neither centralizing nor a unique threat to the metagame, but is more of a special sweeper par excellence, something of a counterpart to Gallade. However, the UU metagame is much more prepared to deal with special sweepers, so Raikou is not as great a priority as the other sUUspects.

6. Crobat: In my opinion, the most "borderline" of the sUUspects, the one that has the lowest likelihood of actually becoming BL. While certainly very good, I doubt the absence of Crobat will even be missed in a Crobat-less metagame, since many other Pokemon can do what it does, albeit not nearly as well. Crobat doesn't really harm the metagame the way that the others do, particularly the first four.

Edit: Those are excellent points made by Imran, and I completely agree with all of them.

 

[Caelum]

I'm fine with a single ladder (that seems to be the general feeling?) since most of these have already been played with together. The 2nd round involves a provisional reconsideration if the voting was "close" (to be determined to be honest) where they would be reintroduced / reconsidered etc. so I don't see it a lot of problems coming from it.

I suppose I could make an announcement for people that want there accounts reset / want to participate post their username and I compile a list a ship it off to Doug. If everyone is cool with that, I suppose that would work well ?

 

[darkie]

The sign-ups thing works for me, though that does make it harder for people to join in the test during the middle of it. The new alts idea makes more practical sense, I think (if we discount a complete reset, which makes the most sense to me overall).

 

[Seven Deadly Sins]

My problem with the UU ladder right now is specifically that people that played old UU have their ratings for the new metagame changed by how well they did in the old one which, let's face it, is completely different. I know there are some people who may be bugged by the fact that, say, Eo Ut Mortus is still on the leaderboard despite the fact that he doesn't use that name for UU any more. (at least I'm pretty sure he doesn't)

Since the ratings are going to be decaying anyway, I really think that a blanket ladder reset would be better, especially for this new metagame.

 

[Imran]

I am also supportive of a complete ratings reset. I think "registering an alt beforehand" is a good idea, but I would like people to be able to register a number of accounts if nessecary. We all know the case of Veedrock, who got beaten at the end, although he put all the work and effort into trying to achieve the rating beforehand. Even reachzero had to make a new alt for the Latias test because he went on a losing streak. If you get consistantly beaten on the ladder because of a bad team matchup, it does become hard to get your rating back up again. I know that UU has a larger pool of players than suspect on average, but I still don't think that these players who may not make it with one alt should be left out.

In all honesty though, the players with the good ratings should play well consistantly. Seeing as we are all using Eo Ut Mortus as an example, I have no doubt he can re-climb the ladder, and it doesn't take "that" long. I don't think people will care that much about a ladder ranking that doesn't take too long to achieve, and that they can easily regain after the reset, and if they do, they probably care too much.

 

[X-Act]

Hopefully GLIXARE should alleviate the problem of taking too long to climb the ladder. I hear Colin is going to implement it for Shoddy.

 

[Seven Deadly Sins]

In all honesty though, the players with the good ratings should play well consistantly. Seeing as we are all using Eo Ut Mortus as an example, I have no doubt he can re-climb the ladder, and it doesn't take "that" long.

Eo Ut Mortus climbed the new UU ladder and seized the number 1 spot within a week of the new UU ladder opening (under the name Sitzmark). So yeah, good players that want their ladder rankings preserved can earn it again in this new metagame just like before.

 

[Aldaron]

http://www.smogon.com/forums/showpost.php?p=1791438&postcount=4

That post is why I am confused regarding any confusion for what we have to do.

I'll state exactly what I think the process IS (not what it might be). The most significant point of the suspect process is that it tests every suspect in a suspectless environment.

The first thing to do is Stage 0 (sorry for the cheesiness), and find out exactly what this suspectless metagame is. That means banning these initial 6 first...and then having another period where we play and determine if there are any additional suspects. Only when there are no more additional suspects can we decide whether to continue on to the "stage" process, and determine whether we should ban all 6 or each individually to determine tier status.

 

[Seven Deadly Sins]

When are the suspects going to be removed? It's been... close to 2 weeks since the suspects were decided on, and they're still on the ladder.

 

[Jumpman16]

I'll state exactly what I think the process IS (not what it might be). The most significant point of the suspect process is that it tests every suspect in a suspectless environment.

The first thing to do is Stage 0 (sorry for the cheesiness), and find out exactly what this suspectless metagame is. That means banning these initial 6 first...and then having another period where we play and determine if there are any additional suspects. Only when there are no more additional suspects can we decide whether to continue on to the "stage" process, and determine whether we should ban all 6 or each individually to determine tier status.

This needs to be emphasized. When we implemented the Suspect Ladder on August 8, 2008, it was identical to the then-standard metagame minus Garchomp. Two days later we removed Deoxys-S from the Suspect Ladder as well. We didn't do this just to appease the masses of people who were tired of Garchomp, as partially evidenced by the decision to take DX-S off the ladder. You personally should remember this post about what a suspect-free metagame is: I implore everyone to read that and the post I made following it.

The main reason I am kind of confused as to why quite a few people think the New UU Suspect Test process is more successful than the OU one is because there is very literally no reason to believe that the "Six Deadly Suspects" are the only Suspects in the UU metagame. This is expressly due to the same caveat in testing all those BL pokemon at the same time that I and others voiced months ago—it will be hard to tell who the real Suspects are. I commend everyone for taking the immense effort in the bold voting process and allowing us to arrive at our six UU Suspects, but there is no way anyone can rationally argue that pokemon like Shaymin, Clefable, Typhlosion, and others that got nominations (if there are any, I'm not reading that entire thread) will not prove themselves to be Suspects in "Stage 6":

6- Then stage 3 would be repeated to determine if there are anymore suspects that exists; the process would be repeated as needed (and may occur at random times due to certain OU Pokemon dropping in usage and becoming BL)

This is why I think those of you who are convinced that the New UU Suspect test will finish faster than the OU Suspect Test or that the UU one is "better" than the OU one should probably reserve judgment until at least a month from now, when the UU metagame without the Six Deadly Suspects has had time to unfold. A time period which, of course, does not guarantee any conclusive facts about Suspects, if Garchomp and Deoxys-S have anything to say about it.

 

[QibingZero]

The main reason I am kind of confused as to why quite a few people think the New UU Suspect Test process is more successful than the OU one is because there is very literally no reason to believe that the "Six Deadly Suspects" are the only Suspects in the UU metagame. This is expressly due to the same caveat in testing all those BL pokemon at the same time that I and others voiced months ago—it will be hard to tell who the real Suspects are. I commend everyone for taking the immense effort in the bold voting process and allowing us to arrive at our six UU Suspects, but there is no way anyone can rationally argue that pokemon like Shaymin, Clefable, Typhlosion, and others that got nominations (if there are any, I'm not reading that entire thread) will not prove themselves to be Suspects in "Stage 6":

This is why I think those of you who are convinced that the New UU Suspect test will finish faster than the OU Suspect Test or that the UU one is "better" than the OU one should probably reserve judgment until at least a month from now, when the UU metagame without the Six Deadly Suspects has had time to unfold. A time period which, of course, does not guarantee any conclusive facts about Suspects, if Garchomp and Deoxys-S have anything to say about it.

I think I should clarify my position here.

I'm not of the opinion that the UU test has been more successful, or is a 'better' method overall. In fact, I don't even think that 'starting with no bans and going from there' is necessarily a better idea (though it was certainly the most reasonable one concerning the BL/UU issues). My arguments are strictly based on the 'popularity' of the ladders and why that is a significant factor in being able to make good judgments on suspects. I think I explain the situation extremely well in my first post in this topic. It's not that the UU test will finish first; it's that quality conclusions can be reached faster when more people are participating and a more dynamic metagame is in place.

As far as the suspects go - I agree 100%. I think it is very likely that Shaymin and Clefable will be suspects in the near future, and that there are even a few more after that as well. I can't say I see the problem with this, though, and I feel that the same would happen if we accurately consider suspects for OU after finishing the current ones.

 

[RBG]

OK, so the next round of Suspect testing for UU is about to start, and I want to gage opinions based on what happened last time.

The idea that has been tossed around the most is a second ladder. The advantage is that people can still test the suspects after they have been named. This was not really possible last round and, in retrospect, was not the best solution.

The main reason i want to do this is because this UU is very similar to the last one, and I don't want to reset ratings, and I don't want the mistake we made with the Shaymin-s test to happen.

Any thoughts, comments concerns? Do you all think a second ladder for UU would work right now or not?

 

[LonelyNess]

From a philosophical standpoint, I think that a second ladder would be best, but from a practical standpoint, do we really have the activity necessary to have a decent test outside of the regular UU ladder?

If you take a look at the ratio of battles from OU to UU, you'll see that UU has about 1/10 of Standard's activity... and Suspect has about 1/20 that of UU...

I'm not saying that the numbers will correlate, but you're talking about a UU suspect ladder that could possibly average somewhere in the range of 200 - 300 battles for the month... That's just simply not enough to test. That's probably how many battles I play just by myself on the UU ladder.

I'd just as soon have the ratings reset on the normal ladder, so that we can actually have a decent pool of battles to go off of.

 

[reachzero]

I agree strongly with LonelyNess. Having more activity is the right way to go with this, and a rating reset to ensure that we have meaningful results. Philosophical purity is great, but it won't make much of a difference is there are only ten or so players actually using the UU Suspect ladder, and I think it very likely that that is what we would see.