@The-Squirrel: How fortuitous! That's exactly how lichess works now :)
it is good that it is gone
my judgement of last week in this thread was premature and wrong; sorry for that; even with the new ratings, its still a pleasure to play here and the interface is superb; the moment we will be able to resize the board (what is in the pipeline as I have heard..), it will be even better;
glad to play here again
As we Austrians say (copyright: Schwarzenegger)
I'll be back!
glad to play here again
As we Austrians say (copyright: Schwarzenegger)
I'll be back!
@AlDra
I have seen that too and I suspect Lichess ratings to now be inflated when compared to chess.com.
Part of the reason could be that Lichess gave back all the points that were lost to cheaters. Chess.com doesn't give points back lost for cheating. However online pools are typically inflated anyway due to more casual players playing online, and only people serious about chess playing IRL tournaments so I think if they are to give back the points from cheating (which I think they should) they should also deflate the ratings again to compensate and make Lichess ratings closer to what our USCF or FIDE ratings would be.
I have seen that too and I suspect Lichess ratings to now be inflated when compared to chess.com.
Part of the reason could be that Lichess gave back all the points that were lost to cheaters. Chess.com doesn't give points back lost for cheating. However online pools are typically inflated anyway due to more casual players playing online, and only people serious about chess playing IRL tournaments so I think if they are to give back the points from cheating (which I think they should) they should also deflate the ratings again to compensate and make Lichess ratings closer to what our USCF or FIDE ratings would be.
Actually, no points were given back, per se.
All ratings were recalculated excluding any games played by anyone who was ever caught cheating.
Since the vast majority of cheaters only started cheating after playing a lot of games on their own, and were usually weaker players who lost more points than they gained, the net effect of the recalculation was to reduce ratings, on average.
All ratings were recalculated excluding any games played by anyone who was ever caught cheating.
Since the vast majority of cheaters only started cheating after playing a lot of games on their own, and were usually weaker players who lost more points than they gained, the net effect of the recalculation was to reduce ratings, on average.
My rating on chess.com (playing 30 minute per side games which would be considered "classical" here) is currently 1661. My same rating at Lichess is 1924.
I once started a thread here asking how people found their chess.com ratings compared to their lichess ratings. This thread was almost instantly moved to the FAQ section for some reason, so pretty much no one responded to it.
I once started a thread here asking how people found their chess.com ratings compared to their lichess ratings. This thread was almost instantly moved to the FAQ section for some reason, so pretty much no one responded to it.
"Part of the reason could be that Lichess gave back all the points that were lost to cheaters."
This is completely inaccurate.
As OneOfTheQ states, points are not 'given back'. When the ratings were recalculated, games against cheaters were not included in the calculations. It's a misnomer to say that points were given back.
"[...] they should also deflate the ratings again to compensate and make Lichess ratings closer to what our USCF or FIDE ratings would be."
Ah, the good old 'Your rating on site A should match your rating on site B'-argument.
That's not how ratings work at all. The distribution and average ability of chess players within all pools is different. It's a tireless waste of time to attempt to get them to correspond with each other and would require a fair bit of mathematical trickery to do right. It's not just a matter of "lets take X amount of points off everyone's rating."
For lack of a better word, it's naive to think that would work as it does nothing to compensate for distribution.
But the real reason it's a waste of time is, "why should we try to match our ratings to FIDE anyway?" FIDE uses the well dated ELO system - a modified version at that. And FIDE ratings only really make sense within the pool of FIDE players. We, and most other modern organisations use the Glicko-2 system, they simply aren't compatible with each other.
This is completely inaccurate.
As OneOfTheQ states, points are not 'given back'. When the ratings were recalculated, games against cheaters were not included in the calculations. It's a misnomer to say that points were given back.
"[...] they should also deflate the ratings again to compensate and make Lichess ratings closer to what our USCF or FIDE ratings would be."
Ah, the good old 'Your rating on site A should match your rating on site B'-argument.
That's not how ratings work at all. The distribution and average ability of chess players within all pools is different. It's a tireless waste of time to attempt to get them to correspond with each other and would require a fair bit of mathematical trickery to do right. It's not just a matter of "lets take X amount of points off everyone's rating."
For lack of a better word, it's naive to think that would work as it does nothing to compensate for distribution.
But the real reason it's a waste of time is, "why should we try to match our ratings to FIDE anyway?" FIDE uses the well dated ELO system - a modified version at that. And FIDE ratings only really make sense within the pool of FIDE players. We, and most other modern organisations use the Glicko-2 system, they simply aren't compatible with each other.
Ok so it's completely inaccurate and that's not the reason.
Still Lichess ratings are now ridiculously inflated against Chess.com ratings.
This is from my own experience of course. If there was perhaps a thread where everyone could compare their ratings across various sites we could see if this is true for most other people. I started such a thread once, but because of some kind of misunderstanding it was buried away in the FAQ section.
Still Lichess ratings are now ridiculously inflated against Chess.com ratings.
This is from my own experience of course. If there was perhaps a thread where everyone could compare their ratings across various sites we could see if this is true for most other people. I started such a thread once, but because of some kind of misunderstanding it was buried away in the FAQ section.
@Seaturtle
1)this is your impresson (as your prefered time control is very rare); many people will have the opposite impression: my lichess bullet rating, for example, is below my official FIDE rating
2) the current lichess rating change made the majority lose 100 - 250 points as the majority play faster time controls then you do
3)chess.com ratings are very different, it matters what time control you look at: their corres. chess ratings are inflated and their bullet ratings are seriously deflated while their blitz ratings are not so much deflated
1)this is your impresson (as your prefered time control is very rare); many people will have the opposite impression: my lichess bullet rating, for example, is below my official FIDE rating
2) the current lichess rating change made the majority lose 100 - 250 points as the majority play faster time controls then you do
3)chess.com ratings are very different, it matters what time control you look at: their corres. chess ratings are inflated and their bullet ratings are seriously deflated while their blitz ratings are not so much deflated
This topic has been archived and can no longer be replied to.