Provisional correspondence rating should not change as much if still in opening book • page 1/1 • Lichess Feedback • lichess.org

I started using Lichess recently and also started playing correspondence games. I have won all my five games so far and reached a rating of over 2100, which is way higher than my real strength. And I think the algorithm could have used information to know it shouldn't have changed the rating as much. My games:

Rating +241 after opponent resigned after three moves after a slight inaccuracy ->
Rating +228 after opponent misses mate-in-1 after sixteen moves
Rating +6 after opponent resigned before his second move
Rating +18 after opponent times out after four moves
Rating +129 after opponent resigns in a worse position after 12 moves

Except game 2, all games were still well within the opening database (which I use to play correspondence as I like to play positions I usually never get to as I don't know that many openings well).

I understand that the algorithm uses the uncertainty for my rating to change the rating more, which makes sense. However, in four out of the five games I won, I had not actually thought myself yet but just picked sensible moves from the opening database.

So I wonder whether it would make sense to reduce the change in rating for correspondence games if the game was at a position that is still included with more than 5 or 10 games in the opening database.

I started using Lichess recently and also started playing correspondence games. I have won all my five games so far and reached a rating of over 2100, which is way higher than my real strength. And I think the algorithm could have used information to know it shouldn't have changed the rating as much. My games: 1) Rating +241 after opponent resigned after three moves after a slight inaccuracy -> 2) Rating +228 after opponent misses mate-in-1 after sixteen moves 3) Rating +6 after opponent resigned before his second move 4) Rating +18 after opponent times out after four moves 5) Rating +129 after opponent resigns in a worse position after 12 moves Except game 2, all games were still well within the opening database (which I use to play correspondence as I like to play positions I usually never get to as I don't know that many openings well). I understand that the algorithm uses the uncertainty for my rating to change the rating more, which makes sense. However, in four out of the five games I won, I had not actually thought myself yet but just picked sensible moves from the opening database. So I wonder whether it would make sense to reduce the change in rating for correspondence games if the game was at a position that is still included with more than 5 or 10 games in the opening database.

tpr

#2

Correspondence rating is a joke anyway. Some think hours over a move, some only seconds.

MikeDee

#3

That's fine though in my view as the rating will reflect the playstyle. If you generally play quickly in correspondence your rating will simply be a bit lower as your win probability is also lower.

My point is that there is actually very little information about my playing strength as long as I am still in the opening database and then the opponent resigns or times out.

That's fine though in my view as the rating will reflect the playstyle. If you generally play quickly in correspondence your rating will simply be a bit lower as your win probability is also lower. My point is that there is actually very little information about my playing strength as long as I am still in the opening database and then the opponent resigns or times out.

tpr

#4

"1) Rating +241 after opponent resigned after three moves after a slight inaccuracy" * He has a right to resign
"2) Rating +228 after opponent misses mate-in-1 after sixteen moves" * Blunders happen even in correspondence
"3) Rating +6 after opponent resigned before his second move" * You cannot force him to play
"4) Rating +18 after opponent times out after four moves" * You cannot force him to play faster
"5) Rating +129 after opponent resigns in a worse position after 12 moves" * Worse position is deadly in correspondence

Here is a game from the 32th ICCF World Championship Finals
https://www.iccf.com/game?id=1164303
So even there blunders or clerical errors happen and they do count for rating, norms, and the World Championship title.

"1) Rating +241 after opponent resigned after three moves after a slight inaccuracy" * He has a right to resign "2) Rating +228 after opponent misses mate-in-1 after sixteen moves" * Blunders happen even in correspondence "3) Rating +6 after opponent resigned before his second move" * You cannot force him to play "4) Rating +18 after opponent times out after four moves" * You cannot force him to play faster "5) Rating +129 after opponent resigns in a worse position after 12 moves" * Worse position is deadly in correspondence Here is a game from the 32th ICCF World Championship Finals https://www.iccf.com/game?id=1164303 So even there blunders or clerical errors happen and they do count for rating, norms, and the World Championship title.

MikeDee

#5

I'm not suggesting those games shouldn't count at all, but that they should be less overweighted for new players so that you cannot get a swing of 240 rating when the other player resigns after move 3.

oofmagoof

#6

I think it's a fair. If I lose a game because of something like me spilling water, something I may be more prone to, it should affect my rating. If I lose because my opponent saw a better line in the opening explorer, then that's fair enough. It's just another variable.

Now, whether the opening explorer should be allowed or not (or implemented differently) is a better question.

I think it's a fair. If I lose a game because of something like me spilling water, something I may be more prone to, it should affect my rating. If I lose because my opponent saw a better line in the opening explorer, then that's fair enough. It's just another variable. Now, whether the opening explorer should be allowed or not (or implemented differently) is a better question.