The computer is fine to tell you about tactical blunders you have made, and utterly useless to teach you how to play.
The best learning method is doing your own (flawed) analysis of your games in as much depth as possible, and only after doing that consulting an engine or, even better, a good player, who can explain you the faults in your findings.
The computer is fine to tell you about tactical blunders you have made, and utterly useless to teach you how to play.
The best learning method is doing your own (flawed) analysis of your games in as much depth as possible, and only after doing that consulting an engine or, even better, a good player, who can explain you the faults in your findings.
+1
Yes - thanks - I think this is a great idea. I thought maybe I could try to find where things started to go wrong in lost games - maybe turning points where my opponent started to build attacks or where a poor move of mine opened a possible good line for them. I could then look at what I might have done differently.
I usually ask for computer analysis after games - but I rarely see why the computer has made some judgement eg:
11. d4 Nf4?? ( 11... Bg4 12. h3Bxf3 13. Qxf3 Rfe8 14. Qb3
Qb4 15. Nc3 Qxb316. cxb3 Nb4 17. Bf4 Bd6
18. Bxd6 cxd619. Rae1 ) ( 11... Nf4 12. Bxf4 Qxf4
13. Qxe7 Bf514. Re1 Bxc2 15. Nc3 Qd6
16. Qxd6 cxd617. Re7 Rfe8 18. Rae1 Rxe7
19. Rxe7 Rb8 )
This is going so deep that I can get nothing useful from it to learn from. In a game I was just looking at the suggested 'best' move would have involved going a rook down for no gain that I could see. It it possible that the 'best' move is only best when it is followed by very precise play - the sort of thing a good player could handle? As a mediocre (at best) player I think that I need lines with a good margin of error built in and which avoid risky sacrifices or needlessly exposing the king etc. I'd love to get a computer analysis by a computer with a rating of say 1600 - I think it might mean more to me!