However, average centipawn loss does get lowered during the endgame, where you take a long time to find a forced checkmate, but shorter routes where there at all times. Still, the computer counts that as 0 centipawn loss (as long as you dont lose the checkmate sequence, that is. I have no clue what happens if you lose checkmate sequence but you are still clearly winning).
My avarage centipawn loss on blitz is 63.3 after 2590 games. And my rating is around 1750s.
It's good to check yourself. For example, if I check for my last 2 months blitz games. I see significant drop on my cps to 50.7 from my general 63.3 score. Also my general average opponent rat is 1481.61 but in the last 2 months avarage opponent rat is 1674.5. So I can clearly see a nice improvement for myself.
It's good to check yourself. For example, if I check for my last 2 months blitz games. I see significant drop on my cps to 50.7 from my general 63.3 score. Also my general average opponent rat is 1481.61 but in the last 2 months avarage opponent rat is 1674.5. So I can clearly see a nice improvement for myself.
If you want to compare yourself to other players: use rating or the percentile. The ACPL is scattered somewhat randomly. In some games I got one-digit values, in some 3 digits (>100 ACPL).
Insights gives something like 43 for me.
PS: you noticed that ACPL also depends on color, strength of opponents etc.?
Insights gives something like 43 for me.
PS: you noticed that ACPL also depends on color, strength of opponents etc.?
I don't mind giving mine. 48.6 This is as blitz according to the stat in chess insights. Rapid is 43.8 and classical is 40.7.
I am currently 2005 USCF. That is the only established rating I have OTB that has enough respect.
I am currently 2005 USCF. That is the only established rating I have OTB that has enough respect.
My average centipawn loss in blitz 52. 7 (2200+ blitz).
in bullet 63 (2300 + bullet)
in bullet 63 (2300 + bullet)
Andrew Tang ACPL stats. bullet 3100 ( 54.6), blitz 2800 (40).
Me: bullet 55.4, blitz 46.4, rapid 29.3.
I do think these are dangerous stats though, e.g in bullet one needs to play out K+Q vs K which depresses this number although the blunder rate is higher.
Likewise I am curious about which rating the frequency of 0/0/0 games is maximum. Naively one would say the highest levels but at lower ratings there are probably people hanging queens in the opening allowing easy wins without 'inaccuracies'.
I do think these are dangerous stats though, e.g in bullet one needs to play out K+Q vs K which depresses this number although the blunder rate is higher.
Likewise I am curious about which rating the frequency of 0/0/0 games is maximum. Naively one would say the highest levels but at lower ratings there are probably people hanging queens in the opening allowing easy wins without 'inaccuracies'.
Thanx guys for replying to my question.
I quess there is a lot more in this then i first thought. My rating seems to be stable at 2000-2050 compared to like 150 ELO points lower 8 months or so ago. (It must be more about my health then being a better chess player. I`` ve suffered from a burnout and other psychological problems but recovered a good bit). But it must be logical my centipawn ratio will not drop significantly as i am faced with consistenly higher rated opponents then before. If i would still be playing constantly against 1800-1900 rated players my centipawn stats must have been much different due to easyer to play positions. Less danger to play a bad move when having a better control over the board.
My centipawn stats must be on the low site during opening play and endgames as this is more about memorisation and theory and moves can be played relatively fast and accurately. While losing pawns and pieces left and right regularly during middle game positions at blitz when things are less clear and more complicated.
I guess this makes the centipawn stats less important then ELO. Unless the competition on matchup did not change. Then it should have dropped with more practice. I``ll guess the drop would still be noticable do.
I quess there is a lot more in this then i first thought. My rating seems to be stable at 2000-2050 compared to like 150 ELO points lower 8 months or so ago. (It must be more about my health then being a better chess player. I`` ve suffered from a burnout and other psychological problems but recovered a good bit). But it must be logical my centipawn ratio will not drop significantly as i am faced with consistenly higher rated opponents then before. If i would still be playing constantly against 1800-1900 rated players my centipawn stats must have been much different due to easyer to play positions. Less danger to play a bad move when having a better control over the board.
My centipawn stats must be on the low site during opening play and endgames as this is more about memorisation and theory and moves can be played relatively fast and accurately. While losing pawns and pieces left and right regularly during middle game positions at blitz when things are less clear and more complicated.
I guess this makes the centipawn stats less important then ELO. Unless the competition on matchup did not change. Then it should have dropped with more practice. I``ll guess the drop would still be noticable do.
This topic has been archived and can no longer be replied to.