You gotta keep this in mind--Whimsicott only went down to UU halfway through the month. Possibly why its usage is in the crapper.
Well, Hopefully Whimsicott will rise, if not, I have serious doubts about the UU players @_@
You gotta keep this in mind--Whimsicott only went down to UU halfway through the month. Possibly why its usage is in the crapper.
But the UU statistics look like they were compiled by a group of chimpanzees.
But the UU statistics look like they were compiled by watching a group of chimpanzees.
Chandelure has no business being #3 lol. Roserade being #1 I would gather is mostly for Spikes. Spikestacking is ridiculously effective in UU, especially with Stoutland bringing in the rear, and literally everything physical having trouble beating Hippowdon without at least a couple of layers. Cobalion is way too low, easily the best offensive Fighting type in UU right now. All of its sets pretty much win games. I'm just a little saddened Weavile is falling so low, but again, Spikestacking...
Antar, how are you going to collect rated stats for this month? Will every battle show the rating of both players, or will you still use end-of-the-month ratings?
So the way things are now, the server creates TWO log files. The first is the battle log, which is all we've had to work with up to this point. The second contains full team stats, including movesets, items, etc. It also contains player ratings.
I work with battle logs. Innocent Criminal works with the other stuff. So come November, I'm probably going to turn the compilation of "1337" stats over to him, using the player rating at the time of the battle, while I focus on metrics.
Oh that's nice. Now just to be sure, is there a reason we have 1337 stats (where only teams of players with a rating above a certain threshold) instead of weighted stats (every time a pokémon is using by a player with X rating, X is added to its usage total)?
Tornadus +25
Laziness? I don't want to have to fine-tune a weighting function. Besides, higher-rated trainers weight themselves because they battle more. The idea behind "13
Then again, I assume that a consistent group of noobs have started to use him because it was nominated as a suspect (lmao).
Can someone explain to me how the hell can Scizor work as a lead in OU and why it's the 6th most used lead ?
| 2 | Electivire | 57 | 0.391% | 1.24±1.08 | 3.76±1.91 |
Oh no, Swampert's in trouble! I'll have to remember to play a bunch of games with him in December.
A large part of it is that the metrics are crap. When the standard deviation is on the order of the thing you're measuring, you're doing it wrong.1337 Stats
OU
Sorted by average number of KOs per battle:
Code:| 2 | Electivire | 57 | 0.391% | 1.24±1.08 | 3.76±1.91 |
Ummmmm.....!??!?!?!
Also, I've been noticing Chansey has been slowly climbing up OU's stats every month, whilst Blissey has been falling.