lichess.org
Donate

Where do you think the average chess player loses in the Opening, Middlegame or the Endgame

"... All the time one should be striving to upset the balance of the position in one's own favour, meanwhile patiently biding one's time. ... 'lead in development', 'strong centre', 'open files', etc. Steinitz taught that one should try to pick up such advantages one by one, no matter how tiny they may be, in preparation for launching the attack. ..." - The Development of Chess Style (1997) by Max Euwe/John Nunn
Symbols like ± have been used to indicate some difficulty for one of the players.
We use "+-" for "white has a decisive advantage".

"... All the time one should be striving to upset the balance of the position in one's own favour, meanwhile patiently biding one's time. ... 'lead in development', 'strong centre', 'open files', etc. Steinitz taught that one should try to pick up such advantages one by one, no matter how tiny they may be, in preparation for launching the attack. ..." - The Development of Chess Style (1997) by Max Euwe/John Nunn Symbols like ± have been used to indicate some difficulty for one of the players. We use "+-" for "white has a decisive advantage".

@rbhoopathi04 said ^

... tpr draws a winning rook endgame

At least he didn't lose it.

@rbhoopathi04 said [^](/forum/redirect/post/4jgfkDVa) >... tpr draws a winning rook endgame At least he didn't lose it.

@Toscani writes: "I'm not even convinced that 50% of the members on lichess has even completed the learn section on lichess."

Nice reminder!

I'd forgotten it was even there. I've done 4 percent of it!

I've probably done it by proxy, through books, over a long period of time. But it wouldn't hurt to finish it here on LIchess, as a refresher.

Nice suggestion, thanks.

I don't cheat, however. Cheating is for the underconfident, and once a man becomes the owner of a spaniel he struts through life!

Yes, even a wicked, turkey-stealing spaniel.

Post Script: there's nothing else to say, however. @tpr nailed it, earlier.

@Toscani writes: "I'm not even convinced that 50% of the members on lichess has even completed the learn section on lichess." Nice reminder! I'd forgotten it was even there. I've done 4 percent of it! I've probably done it by proxy, through books, over a long period of time. But it wouldn't hurt to finish it here on LIchess, as a refresher. Nice suggestion, thanks. I don't cheat, however. Cheating is for the underconfident, and once a man becomes the owner of a spaniel he struts through life! Yes, even a wicked, turkey-stealing spaniel. Post Script: there's nothing else to say, however. @tpr nailed it, earlier.

The average rating on lichess is somewhere around 1400-1600 (for example the average rapid player on lichess is ~1400: https://lichess.org/stat/rating/distribution/rapid ) because 1500 is the default/initial rating.

A bot that simulates 1500s is @maia5 : restricting to opponents of similar strength we see that
https://lichess.org/insights/maia5/cplBucket/phase/opponentStrength:30

Average per game: 36.7 moves; 10.18 opening moves, 11.78 middlegame moves (excluding 6% of games ended in the opening), 21.78 endgame moves (excluding 30% of games ended earlier)

Game phasePerfect≤ 10 CPL≤ 25 CPL≤ 50 CPL≤ 100 CPL≤ 200 CPL≤ 500 CPL> 500 CPLNumber of moves
Opening24.3%29.5%18.5%13.8%8%3.9%1.7%0.2%51,084
Middlegame27%17.8%13.7%11.8%12%9%6.8%1.9%59,278
Endgame50.5%10.7%9.2%8.2%7.7%6.2%5.4%2.1%89,638

I gave the above information to "mathgpt" and it gave:
image.png

(Note: it set the value of each centipawn-loss-bucket to their midpoint except blunders (>500CPL) which was arbitrarily set at 600CPL)

(Incidentally this means pretty much no 1500 is winning against @LeelaQueenOdds which makes sense)

The average rating on lichess is somewhere around 1400-1600 (for example the average rapid player on lichess is ~1400: https://lichess.org/stat/rating/distribution/rapid ) because 1500 is the default/initial rating. A bot that simulates 1500s is @maia5 : restricting to opponents of similar strength we see that https://lichess.org/insights/maia5/cplBucket/phase/opponentStrength:30 > Average per game: 36.7 moves; 10.18 opening moves, 11.78 middlegame moves (excluding 6% of games ended in the opening), 21.78 endgame moves (excluding 30% of games ended earlier) > > Game phasePerfect≤ 10 CPL≤ 25 CPL≤ 50 CPL≤ 100 CPL≤ 200 CPL≤ 500 CPL> 500 CPLNumber of moves > Opening24.3%29.5%18.5%13.8%8%3.9%1.7%0.2%51,084 > Middlegame27%17.8%13.7%11.8%12%9%6.8%1.9%59,278 > Endgame50.5%10.7%9.2%8.2%7.7%6.2%5.4%2.1%89,638 I gave the above information to "mathgpt" and it gave: ![image.png](https://image.lichess1.org/display?op=noop&path=DT1RB9SuZzad.png&sig=8390ba0700cbefa6998676a174149de9283530e9) (Note: it set the value of each centipawn-loss-bucket to their midpoint except blunders (>500CPL) which was arbitrarily set at 600CPL) (Incidentally this means pretty much no 1500 is winning against @LeelaQueenOdds which makes sense)

Oh I thought the question was "how much do people lose in the opening, middlegame, or endgame" not "where do they lose".

Since most people just resign, I think where games actually end is a good indicator of when a game is finally lost.
So only

  • 5.7% of games lose in the opening,
  • 23.6% in the middlegame, and
  • 70.7% in the endgame.

This is against maia5 who basically never resigns, so the actual percentages should move even earlier than that.

Oh I thought the question was "how much do people lose in the opening, middlegame, or endgame" not "where do they lose". Since most people just resign, I think where games actually end is a good indicator of when a game is finally lost. So only - 5.7% of games lose in the opening, - 23.6% in the middlegame, and - 70.7% in the endgame. This is against maia5 who basically never resigns, so the actual percentages should move even earlier than that.

@ArYaNnViNd said ^

Come on tell me how all of us can improve

Why do you need an answer to this question? You don't play chess. In real chess, 15 seconds should be spent on one move, not on one game.

@ArYaNnViNd said [^](/forum/redirect/post/Pycx8S1Y) > Come on tell me how all of us can improve Why do you need an answer to this question? You don't play chess. In real chess, 15 seconds should be spent on one move, not on one game.

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