lichess.org
Donate

New Rating Categories For

The problem I can see is that the longer time controls would mostly attract weaker players and the ratings there would not be meaningful. In fact I think this is already the case now in Classical to some extent. After a few games there my percentile is 88%, while it's much lower in faster categories. Maybe I'm better at longer time controls, but I also think the chess level may be lower there.
Thanks everyone for replying. Most of you are agreeing with this model below, which adds just one new category. I think it’s a good start:

0--Bullet--3--Blitz--8--Rapid--30--Classical--300
|========|======|========|===========|

I agree that moving blitz upper limit from 8 to 10 min would be nice too, as some suggested, but that can be done latter.

themanp15 (#3), indeed, abrupt changes are confusing but new categories can be added gradually. The excess of rating types may also be confusing, but we already have 14 different types. Ratings of regular chess can be shown in a small chart/diagram, as in the link below. I think that was what FilipiSilvano suggested in #11.

i.imgsafe.org/1cef62c28c.png

johnbartholomew2 (#6), you have played just 3 classical games and no blitz games, so maybe that’s why you haven’t yet felt any need for more categories.

The number of active players this week for Bullet/Blitz/Classical categories is 17k, 44k and 44k, respectively. So I don’t think regular chess categories will get too diluted at all. For variants, it ranges between 800 and 1400 players per week (racing kings variant is still in 500 per week because it’s too new), so I think 3 categories would work fine, but we can leave that for latter. A diluted would be something like 50 players per week. That would make the ratings of that category to be too spaced and inaccurate, but would not make it any harder for us find opponents for our seek ads, or players for our tournaments.

SoulPeace (#7), the pool idea is nice (I have tried it in ICC), but we already have the recurrent torments. Having an additional rating for specific time controls seems nice, but the fact that it overlaps with regular categories is a bit confusing. Best option would be that games affect pool rating plus regular category rating. Thus 3+1 would affect both your 3+1 pool rating and your blitz rating too. But I don’t think we need so much complexity yet.

JacquesD (#12), do you think the average player strength on classical rating category is weaker? So why do top blitz ratings go to 2900+ and top bullet ratings go 3100+, while top classical goes only 2700+? In my opinion, bullet/blitz pools are actually easier, because most players (like me and you) have trouble with time pressure and/or can't think accurately under a fast pace.

Others, however, have much higher bullet/blitz ratings. For instance, look at the ratings of all top bullet players. Also, check all bullet seek ads and the ratings of those players. Their classical ratings are usually equal or lower than their bullet ratings, because most of them do fine under time pressure.

So, on the contrary of what you though, I think that for longer time control categories the average player strength will be even higher and closer to FIDE ratings.
Looking at other sites, chess.com is the same as ICC ie

0--Bullet--3--Blitz--15--Standard-->
|=======|=======|==========>

However, I think a new rapid category as proposed would be ideal:
0--Bullet--3--Blitz--8--Rapid--30--Classical--300
|========|======|========|===========|

And I voted :)
http://www.strawpoll.me/11093155
I have voted too for:
0--Bullet--3--Blitz--8--Rapid--30--Classical--300.
The great majority seems to like the four categories system.
I'll try to ask thibault what he thinks about it when I see him somewhere in chat. You other people feel free to do the same in case you see him. ;)

This topic has been archived and can no longer be replied to.