-
Welcome to Smogon! Take a moment to read the Introduction to Smogon for a run-down on everything Smogon, and make sure you take some time to read the global rules.
Gen 2 GSC Viability Ranking (OU)
- Thread starter Jorgen
- Start date
- Status
Wait a minute, under current Smogon rules isn't Smeargle allowed to run Spider Web as its fourth move? I'd hardly call SleepTrapPass "just a gimmick", since it's considered so deadly that some ban it.
PO automatically bans Sleep Trapping as a whole, which while it's primarily meant to stop Perish Song BS also nips Nightmare in the bud and catches Smeargle in the crossfire. It's pretty difficult to even test it when you basically have to force people to play on PS just to even be allowed to use it, even if by Smogon rules it's technically allowed, since most people are playing old gens on PO by default. :/
Even without the ban on it, I probably wouldn't use Smeargle on, like, every team, despite my previously expressed disdain for its legality. In fact I'd probably be too chicken to use it more than every once in a while, I mean Sleep Trap is strong but if it goes even slightly wrong you're basically playing 5-on-6. I'm just not too huge on Pokemon that do one gimmick without contributing to conventional offensive or defensive roles on top of it.
Well, you don't use Suicune on every team either, but it's still A-rank. You've called for a ban on SleepTrapPass Smeargle before; this seems to contradict your claim that Smeargle is C-rank under rulesets permitting it to use SleepTrapPass.
And, well, the thing about Smeargle and powerful BPs in general is that it doesn't matter how "weak" your team is if you win anyway. Because you've won, which is by definition all that matters. This is what I was getting at by posting Last Chance; it's actually considered a fairly powerful card despite literally causing you to lose if it doesn't work because that one turn is often all you need to win (and, well, read the explanatory text).
Yeah I said it pretty confusingly, but I've taken a step back from that older "barn smeargle" stance. Was definitely an over-reaction.
Smeargle isn't quite auto-win the way that card can be, and it's not like you're sacrificing as much by running that card in your deck as opposed to running Smeargle over another Pokemon. How close Smeargle is to auto-win is yet to be determined, but until I have enough chances to see all of its sets in action, it's hard for me to imagine it being on par with (and therefore worth using on as many teams as) more flexible mons (Suicune included).
Smeargle isn't quite auto-win the way that card can be, and it's not like you're sacrificing as much by running that card in your deck as opposed to running Smeargle over another Pokemon. How close Smeargle is to auto-win is yet to be determined, but until I have enough chances to see all of its sets in action, it's hard for me to imagine it being on par with (and therefore worth using on as many teams as) more flexible mons (Suicune included).
Hypothetical time: If we somehow calculated Smeargle's (or some other BPer's) win rate, what rates of winning would you consider to be at which ranks?
I mean, clearly a 100% victory rate is "ban" rank, and a 0% victory rate is "not on the list" rank. But what about in-between?
It's hard. I'd probably call for the ban of a 30% across-the-board instawin mon, even if a team's odds of winning should it not get the instawin be zero. It'd easily reduce the game to a weighted coinflip and take the players out of the equation. That said, I would not rank it high if such a mon had a 0% win rate conditional upon the instawin failing, even while simultaneously calling for its ban. Weird, but not if you use a criterion other than power for banning (in this case, it's the "players need to be part of the equaltion" criterion).
Anyway, as for your actual question, I probably wouldn't base an objective ranking on an insta-win rate. Probably not even on an overall win rate. I'd probably base it more on frequency of Pokemon X being used among the set of optimal teams that average a 50% win rate against each other. Because anything other than 50% implies that you're using a bad prior distribution of opposing teams. Or something like that. Does that make sense? I don't know if it does. I tried to flesh out the idea but I ended up confusing myself and it seemed a pointless exercise as such a computation would most certainly be intractable.
Anyway, as for your actual question, I probably wouldn't base an objective ranking on an insta-win rate. Probably not even on an overall win rate. I'd probably base it more on frequency of Pokemon X being used among the set of optimal teams that average a 50% win rate against each other. Because anything other than 50% implies that you're using a bad prior distribution of opposing teams. Or something like that. Does that make sense? I don't know if it does. I tried to flesh out the idea but I ended up confusing myself and it seemed a pointless exercise as such a computation would most certainly be intractable.
So basically what we're measuring is "in the Nash equilibrium, which things get used to what degrees?", or "what would this page look like if all players had infinite skill?" Nice premise, but it has some issues. Snorlax is safe in the top spot with its equilibrium usage rate of 100%, but Raikou and Zapdos typically don't fit on the same team, so their usage rates will sum to slightly over 100% at most. In turn, this means something like Cloyster, which is utile on almost all teams and doesn't really compete with other standards (only with Forretress and - for spin Cloy - Starmie), is probably going to fall between them (and indeed, for the past 12 months Raikou's usage has never exceeded that of Cloyster on PO, and Cloyster's usually higher on PS as well). Are you prepared to give Cloyster S-rank? Probably not. And I wouldn't be, either.
Well, Lax team vs. non-Lax team isn't 50-50, so given a real meta where there is some (small) percentage of non-Lax teams, you can do that sort of empirical analysis (you can use various formulae to compare the win rates of Pokemon with different usage rates). Lax's corrected win rate in the ideal meta is strictly not "50%", but "indeterminate", because those formulae end up dividing 0 by 0 in the "non-Snorlax wins divided by non-Snorlax teams" term.
I think that, given some sort of (computationally intractable, but who cares) mixed-strategy Nash equilibrium of teams and plays, those corrected win rates are exactly what we're trying to measure.
Last edited:
Trying to define what it is that we're measuring isn't particularly silly IMO. We are taking it elsewhere, though.
Are you preaching to the choir or simply confused and arguing with nobody? I don't think anyone in here was saying that "the general ban of PO on sleeptrapping" wasn't retarded.
Sleep trapping on its own is not particularly useful, it's just the core that enables Nightmare, Perish Song, and Smeargle trap-passing. I disagree with PO banning it, both in the fact that I don't think it's reliably powerful enough to deserve being banned at all and also their poor implementation (it should not be enabled in custom, private, unrated challenges), but I understand why. It's simpler to say "ban this + that" than "ban this + that + [other1 OR other2 OR other3]," especially since there's no reason to use this + that anyway without that third move. It's a bit more reliable than Umbreon trap-passing with Confuse Ray though, although I did own Jorgen with much RNG this past SPL with it (before choking away the lead it gave me).
Eh, Sleep Trapping on its own isn't BAD (Jynx gets to actually use Dream Eater, and it's not terrible on Gengar either). It just has non-luck-based counters, which HypnoMissy flat-out doesn't.
And how many people are there?
I meant "are there enough people there to get a read on what the meta looks like with just a ML/Hyp/PS ban".
Has Showdown fixed GSC and RBY yet or do they just not give a shit?
what issues are left? afaik only trap passing is broken in GSC (and powder snow because I haven't fixed my commit yet)
Low Kick, phazing moves and the SleepPerishTrap ban are the ones that were broken in GSC last I checked. Low Kick because the damage isn't variable with weight before ADV (but it was variable on PS's GSC OU), phazing because somebody had hacked it to "slower negative-priority moves go first" instead of actually putting in the proper fix ("phazing moves fail if they go first"), and the SleepPerishTrap ban only banned Hypnosis/Mean Look/Perish Song instead of <sleep move>/<permanent trap move>/Perish Song (Jynx and Smeargle get access to non-Hypnosis sleep moves, and Smeargle gets Spider Web). Have those been dealt with yet?
RBY last I heard Reflect and Counter weren't set up right (crits are supposed to ignore screens in RBY; Counter's supposed to carry over turns). Have those been fixed?
What do you mean by "trap passing is broken"? You mean it wears off on PS as it would in Gen 5/6, instead of being passed as it should?
Last edited:
Mean Look doesn't pass in GSC on PS yet is what I mean. A fix was attempted but didn't solve it. Also phazing does fail if they go first in PS I believe. And iirc The Immortal asked what the proper thing to do about the SleepPerishTrap issue (when it was discovered Jynx could get through teambuilder legality checks) and I'm not sure if someone answered (it was another thread).
Reflect isn't correct for RBY on PS, not sure about counter.
Reflect isn't correct for RBY on PS, not sure about counter.
Yes, Immortal asked what was correct after I said what was correct. Since Immortal was quoting me, I assumed my word alone wasn't considered reliable enough to act on, and others' confirmation was being sought. So I didn't reply, and for some unknowable reason nobody else did either.
Okay, any ideas when it will be correct?
Leech Seed isn't working correctly as well. Currently if the seeded Pokémon doesn't attack, for example if he gets fully paralyzed or is fast asleep, then Leech seed doesn't work. This needs to be fixed, it should do damage after the affected Pokémon moves, doesn't matter if it manages to attack or not.
- Status