Smogon Forums

Not a problem. When did it start?

Or do you just want the partial stats for June?

I can always combine...

Actually, the DOU stats for May are borked, since I didn't have doublesoususpecttest listed as a non-singles tier, meaning that all DOU battles shorter than 6 turns got thrown out.

If the May stats are messed up as you say, partial stats for June will work fine. (Unless May stats from the 22nd on are an easy fix, but afaik it's not)

Yeah, I'm gonna (1) fix May's stats and re-upload them, (2) run June's stats and (3) post a combined usage report in the thread you tell me. Might take a few days.

That's fine, take as long as you need!

Wow! It was a lot faster than I expected. I have the stats now. You want I should post them somewhere? Or wait until the ladder is closed and then run it again?

You could post them in the stickied NP: thread or as their own thread

commenting here so I don't clutter up your VM wall: I've been thinking about the Policy Review Stratos made a lot, and while I disagree with his proposal that doesn't mean I haven't been thinking about other ways to solve the "problem." (cont)

Somebody mentioned that they heard that a huge percentage of players on the Doubles OU ladder had a rating of exactly 1000, and it made me curious what the actual numbers were. Is this a data set you'd be able to produce (without too much hassle)?

Keep in mind: 1000 Elo is the lowest possible rating. So players who lose a lot just stay at 1000 Elo. They don't contribute to tiering at the 1695 level. I *can* look into it, it's not a *huge* hassle, but it will take me a while to set aside the time. Ping me about this if you haven't heard from me in, say, a week. I actually need to get the distribution of ratings to verify that the Std.dev is ~130, since that might be "breaking" the stats.