1. New to the forums? Check out our Mentorship Program!
    Our mentors will answer your questions and help you become a part of the community!
  2. Welcome to Smogon Forums! Please take a minute to read the rules.

August's stats, and the new tier cutoffs

Discussion in 'The Policy Review' started by Antar, Sep 14, 2011.

  1. eric the espeon

    eric the espeon maybe I just misunderstood
    is a Forum Moderator Alumnusis a CAP Contributor Alumnusis a Researcher Alumnusis a Tiering Contributor Alumnusis a Contributor Alumnus

    Joined:
    Aug 7, 2007
    Messages:
    3,694
    Raising in tiers due to usage shows that the Pokemon's power in the higher tier is sufficient for it to be raised, which not the same as being broken in the lower tier at all. It's the opposite of the classic "<poke> sucks in Ubers, move it down" suggestion brought up by newer players. Pokemon move up into banlists if they are too strong for the tier they are in, and they move up into usage tiers if they are strong enough to fend for themselves in the higher tier. Applying this to your Gastrodon example, Gastrodon has now proved itself powerful enough to be viable in OU (as measured by OU usage statistics), and therefore cannot be classed in a lower tier than OU.
    I should have been more specific. Power is not uniform across tiers and can only be measured relative to the rest of the threats available (the most stark example I can think of is Sheddy in gen 4. For some time, sheddy was quite deadly in Ubers despite being utterly useless in all the tiers below.), and because of this these there will always be Pokémon in a higher usage tier which would not break the metagame below them if they were to drop. However, usage in a particular tier is a fairly good measure of power within that tier, and since one of the primary aims of the tiering system is to ensure diversity of choice it makes perfect sense to construct lower tiers around Pokémon which are less effective in existing tiers.
    Viable, perhaps, but less viable than those used more in OU (on the whole, because of course players don't always use the best available for their teams and new sets will continue to arise). Usage in a tier is not a perfect measure of power within that tier, but so far no one has been able to come up with another sensible non-arbitrary system.
  2. JabbaTheGriffin

    JabbaTheGriffin Stormblessed
    is a Tutor Alumnusis a Super Moderator Alumnusis a Smogon Media Contributor Alumnusis a Tiering Contributor Alumnusis a Contributor Alumnusis a Battle Server Moderator Alumnus

    Joined:
    Jan 21, 2007
    Messages:
    3,024
    When talking to jc104, he did mention that since the 2nd and 3rd methods resulted in similar results, which adds to their validity. I definitely see the merit in that. The entire system is pretty arbitrary, I guess, so if a couple different methods produce similar results that are "close enough" to what the normal method would be I guess that's fine as a last ditch effort. But as I don't think we're in any rush right now, statistics are nice to have, but we don't exactly have to rush out and try to pump out tier lists.

    Just to clarify: I'm not waiting until after testing to update the tiers. I'm waiting until a tiering process is decided upon before UU is updated, that way any drops/changes don't ruin the tier without any way to correct the tier. While this may be over-precautionary, I just don't like the idea of the inability to make tiering decisions on a shifting tier. Therefore, while the fact that UU is static may be boring, I definitely think it's in the best interests of players in the longrun.
  3. Antar

    Antar That's Dr. Antar to you
    is a Battle Server Administratoris a Programmeris a Super Moderatoris a Community Contributor
    Official Data Miner

    Joined:
    Feb 17, 2010
    Messages:
    3,083
    Would you be willing to put this to a vote, or is this an executive decision? Because most of the people I've talked to seem quite eager for the new tiers and would be fine with a bit of chaos and shifting in the short-term.
  4. Honko

    Honko
    is a Programmeris a Tiering Contributoris a Contributor to Smogon

    Joined:
    Dec 6, 2009
    Messages:
    1,159
    Thanks for clarifying, that's good to hear. I wish we had some more information on the timeframe for those decisions though; an estimate from Arti or BK would sure help everyone feel a little less in the dark.
  5. Blue_Tornado

    Blue_Tornado

    Joined:
    Apr 14, 2009
    Messages:
    3,309
    Yes, I'm pretty sure people actually want the new tiers for the chaos. I wouldn't advise a vote though as that would just take a lot of time. Can't we agree on a certain new tier list, then agree on redoing the process once we vote on what the process should be? (Like, after a few weeks, maybe less.)
  6. Antar

    Antar That's Dr. Antar to you
    is a Battle Server Administratoris a Programmeris a Super Moderatoris a Community Contributor
    Official Data Miner

    Joined:
    Feb 17, 2010
    Messages:
    3,083
    I guess I'm not completely sure what's up-in-the-air about the process. Isn't it the same as always (update usage-based tier cutoffs every three months, then have suspect voting to get the BL(s))?
  7. Mario With Lasers

    Mario With Lasers Self-proclaimed DEAD king
    is a Forum Moderator Alumnusis a CAP Contributor Alumnus

    Joined:
    Nov 15, 2006
    Messages:
    7,165
    This may be minor, but I vote for using Maestro's method for the next tier update just because I don't want Hippowdon in UU, at least not yet.
  8. NatGeo

    NatGeo sun's coming up
    is a Contributor Alumnus

    Joined:
    Sep 4, 2010
    Messages:
    2,428
    Actually, if Hippowdon does go UU that would have a boatload of complications for UU suspect testing and just cause mayhem in that tier, and I don't think UU needs another powerful weather like hail already is. Besides I don't think we can make Hippowdon automatically BL, so I guess we'll just have to wait until later when Hippowdon does officially fall into UU using Maestro's method.

    tl;dr We really should use Maestro's method to collect stats so UU can have a bit to clean all the dust (frost?)
  9. Antar

    Antar That's Dr. Antar to you
    is a Battle Server Administratoris a Programmeris a Super Moderatoris a Community Contributor
    Official Data Miner

    Joined:
    Feb 17, 2010
    Messages:
    3,083
    The word seems to be that we're waiting for some decisions to be made as to the tiering process. At this point, even if all those decisions were made TOMORROW, we might as well wait until October so we can go based on September's stats.

    And yeah, the general consensus seems to be, Maestro's Method or none at all.
  10. Blue_Tornado

    Blue_Tornado

    Joined:
    Apr 14, 2009
    Messages:
    3,309
    Uh, I thought we were choosing a compilation method that would be used for ALL future tier lists, not just this one. The aftermath it causes in regards to this round, while not something to be left aside, isn't supposed to decide the method, as we ARE thinking long-term here. While I don't have an opinion on this myself, I do know that deciding on this method just because of Hippo, and then realizing it's a mistake after a few rounds, is a possible scenario, and more thought should be put here.

    Not to mention, I'm of the opinion that Hippo is a positive effect on the metagame, but this thread is about a different topic to begin with.
  11. Oglemi

    Oglemi <&Oglemi> i am a gay badger buddy :3
    is a Tournament Directoris a member of the Site Staffis a Community Contributoris a Pokemon Researcheris a Smogon Media Contributoris a Contributor to Smogonis an Administratoris a Tiering Contributor Alumnus
    C&C Leader

    Joined:
    Oct 13, 2009
    Messages:
    8,398
    Let's get something straight here. What Pokemon does or does not fall should have no basis on which method we pick (if we do pick one, I'll get to that). We should be picking the most mathematically correct one if anything (which is somewhat arbitrary but w/e). Our tiers are based on usage, not on what we do or don't want to fall.

    Now, with the usage plug-in implemented (which I still guess it is, nobody has actually confirmed it yet), we will probably go back to our old method and just skip over August as well and just use September's stats (whether or not we combine them with May and June's is questionable).
  12. Antar

    Antar That's Dr. Antar to you
    is a Battle Server Administratoris a Programmeris a Super Moderatoris a Community Contributor
    Official Data Miner

    Joined:
    Feb 17, 2010
    Messages:
    3,083
    First off, I've heard nothing about this either, and I'm more inclined to assume that no news ISN'T good news.

    Secondly, as mentioned above, I think it would be silly to discard August's stats (and at least the first half of September's--since, at the earliest, the plugin went into effect on the 16th)--the fact that Maestro's Method and the Full Team method yield such similar results really suggests that they'd be very similar to what the usage plugin would have generated. May and June were EXTREMELY long ago, in terms of relevance to the metagames. If we're not going to use Aug/Sept, we should just use October's stats alone.
  13. Heysup

    Heysup Monsters are dangerous and kings are dying like flies.
    is a Tiering Contributoris a Contributor to Smogonis a Forum Moderator Alumnus

    Joined:
    Jun 28, 2008
    Messages:
    5,571
    If I understand what Antar is saying (I'm not stat-gathering savvy), I agree with his second point. I feel that keeping the status relevant to the bigger picture is more important than it is to be identical (rather than just really similar) - we need more current stats. Additionally, we want a more stable list then just one month of statistics. Basically what I'm saying as that we should just use the August/Sept/Oct stats for UU (if we've had those tiering decisions by then).

    Speaking of tiering decisions, we are probably at the point in which we need to know when these decisions are going to happen. This would heavily influence are decision whether we should wait or not, wouldn't it?
  14. Antar

    Antar That's Dr. Antar to you
    is a Battle Server Administratoris a Programmeris a Super Moderatoris a Community Contributor
    Official Data Miner

    Joined:
    Feb 17, 2010
    Messages:
    3,083
    Could you inform those of us who haven't been paying attention, what tiering decisions need to be made?
  15. Heysup

    Heysup Monsters are dangerous and kings are dying like flies.
    is a Tiering Contributoris a Contributor to Smogonis a Forum Moderator Alumnus

    Joined:
    Jun 28, 2008
    Messages:
    5,571
    The ones Jabba's referring to regarding the tiering process. We've been putting our suspect tests on hold for a while. I'm sure someone else can give you more information.
  16. Antar

    Antar That's Dr. Antar to you
    is a Battle Server Administratoris a Programmeris a Super Moderatoris a Community Contributor
    Official Data Miner

    Joined:
    Feb 17, 2010
    Messages:
    3,083
    September ends today, folks. So as soon as Blue Kirby sends me the raw data, I can get this month's stats up.

    When I do, I'll go ahead and post a potential usage-based tier list, using Maestro's method and the formula 20*Sept+3*Aug+1*June [edit: and a cutoff of 3.41%].

    I'm hoping that this is the month that we can finally update the tiers. Any updates on the policy discussions would be greatly appreciated.
  17. JabbaTheGriffin

    JabbaTheGriffin Stormblessed
    is a Tutor Alumnusis a Super Moderator Alumnusis a Smogon Media Contributor Alumnusis a Tiering Contributor Alumnusis a Contributor Alumnusis a Battle Server Moderator Alumnus

    Joined:
    Jan 21, 2007
    Messages:
    3,024
    I just want to note that the cutoff is not 3.41% (it ends up around that mark all the time, but that's not what it is). The cutoff is a formula that determines if you're more likely than not to see that pokemon over the course of 20 battles. It's somewhere on site I'll edit it into this post if I find it.

    Also there are problems with UU right now that I want fixed before the UU tier list is updated. I'm planning to speak with the administration soon about the direction of UU, but changing up UU right now does nothing but introduce more chaos to a tier that already is a bit messed up.
  18. Antar

    Antar That's Dr. Antar to you
    is a Battle Server Administratoris a Programmeris a Super Moderatoris a Community Contributor
    Official Data Miner

    Joined:
    Feb 17, 2010
    Messages:
    3,083
    Sorry, what I should have said is "effective cutoff." We're using Maestro's method to infer pokemon not seen (recall the comparative analysis between the six-team and Maestro methods).

    Can you explain a bit more what problems are going on with UU (if there's already a discussion thread about it, a link would suffice)? As far as I've played, UU seems pretty balanced.
  19. Heysup

    Heysup Monsters are dangerous and kings are dying like flies.
    is a Tiering Contributoris a Contributor to Smogonis a Forum Moderator Alumnus

    Joined:
    Jun 28, 2008
    Messages:
    5,571
    It's basically been on hold for an incredibly long time (~2 months) in which we still have suspects involved and cannot do anything about it until the administration has their input. All I know is that we are waiting on a meeting supposedly, but I'm not quite sure as to what the full story is (2 months is more than enough time to schedule a meeting if you ask me).


    EDIT: @ below

    People are already frustrated.
  20. Antar

    Antar That's Dr. Antar to you
    is a Battle Server Administratoris a Programmeris a Super Moderatoris a Community Contributor
    Official Data Miner

    Joined:
    Feb 17, 2010
    Messages:
    3,083
    Just did a bit of reading in Inside Scoop.

    I think I now agree with Jabba. Until a new suspect process is agreed upon and put into place, all that updating the tiers would do would be to throw everything into (deeper?) chaos.

    That being said, if the tiers don't get updated soon, people are going to become frustrated. I could even imagine people turning to other sources for official tiers, and that would pretty much be the death of Smogon (yaay for being overdramatic!).

    Seeing as how I have no control over what the Smogon administration does, all I can do is have stats and potential tiers at the ready, so we can move quickly once everything is decided.

Users Viewing Thread (Users: 0, Guests: 0)