Separate names with a comma.
Discussion in 'Smogon Metagames' started by Antar, Jan 5, 2017.
I think 5-10 days is probably a reasonable expectation for the time being.
Stats for April and May are now up.
Hey so I'm gonna put this here but I can put it in PR if needed, Antar how achieveable would it be to do the sampling of stats from (for example) the 25th of the previous month to the 25th of the next so that stats could come out on the first? Perhaps this is something only I care about but I feel like having the shifts being cleaner from month to month would be good for future organization. As it sits it feels like the boundary is really soft so it makes it hard to refer to a few months back easily.
Stats for the month are now up, however, these stats only cover up to the 23rd, since a glitch prevented PS from actually saving battle logs from the 23rd of June to the 3rd of July. Thus, battle-counts are going to be low, and pre-tier-shift usages are going to be more dominant than they usually are.
Just to clarify, which OU ladder mark is used for rankings? Also, just to double check, 3.41 is the benchmark to be in OU, correct?
Reviloja753, 1695 stats are used for the OU-UU cutoff, but in the first month after a tier update, nothing would be allowed to fall that got more than 1.72% usage or allowed to drop that got less than 6.7% usage. In any case, we're not doing tier shifts this month.
Stats for July are now up.
Antar are the battle counts accurate for march-april-may? did the kerfuffle with the stats for those months have any effect on that? i'm just wondering because the battle counts for those months are lower than i'd expect for some metagames
MAMP, for some metas, you might have to combine across several variants (alpha-beta-suspect...), and there was a spot where a bunch of logs for May never got saved. But otherwise I haven't heard of any issues affecting counts.
Stats for August are now up.
what does average weight/team (the one on there at the top http://www.smogon.com/stats/2017-08/gen7pubeta-1500.txt) mean and how is it calculated?
galbia, see #5 in the FAQ or the Weighted Stats FAQ for more information on how teams are weighted.
Stats for September are up.
I don't know if this has been discussed before but shouldn't the rise/drop cutoff be lower for mega-evolutions just because it's not possible to play several of them in the same team ?
The link in in the FAQ that fixed it is from 2008 when the 1 mega per team limitation didn't exist and I couldn't find any other thread about it.
Nailec, in short, no. Tiers are primarily threat lists. If you don't see a mon or a forme 1 battle in 20 [being used by a competitive player], then it's not a threat worth preparing for. Doesn't matter why.
Out of curiosity, would there be a way to see how much less usage pokemon A would have with pokemon B? For example, in high-ladder AG, Tapu Lele sees increased usage with pokemon such as Deoxys-A. However, it isn't used as much with extremespeed users such as mega rayquaza or arceus, because those pokemon are significantly worse without their priority moves.
The chaos file has a stat called 'checks and counters' which shows how often pokemon A is played with pokemon B which should help you see this.
For example, here you can see that tapu lele's most common teammate is deoxys-attack by far.
However if what you're looking for is seeing how much extreme speed usage on a mega rayquaza drops when paired with tapu lele, I don't think there's a way you can see it from the published data.
according to that website maractus only has 1 counter in OU
Yveltal doesn't see any usage in OU, UU, RU, or NU.
That means that it should be in PU. Yveltal for PU was right.
By the way, what is the counter?
It will be absolutely necessary to use it in OU as soon as people start using Maractus teams.
Nailec thanks! I'm not looking for a rise or drop of usage for extremespeed mrayquaza, I was just giving it as an example of a pokemon that might see decreased usage.
it's Heatran. by the way what he meant by that is that the site is very innacurate.
No, the site is just usage based. Something has to be used for it to beaten.
Yes and since its usage based its really fucking innacurate thanks for providing nothing.
Just because the data is meaningless doesn't mean it's innacurate.
@Summonerrisa4o1, yes. It's called a suspect test.
Stats for October are now up.