i don't use computer. it is smarter to evaluate yourself
I'm not analyzing games with the computer first anymore. I'm still using it but only after I've gone over the game. The longer the time control or more critical the key position was in the game, the more time I spend on review.
You will never see a GM discuss his average centipawn loss rate. Focus on your own evaluations.
I hear you.
When analyzing your games, don't use an engine, ask yourself what were the key moments and how you could have improved.
Work on improving your analysis and evaluation skills, not your average centipawn loss rate.
The best advice that I've seen in these forums in the past week.
When analyzing your games, don't use an engine, ask yourself what were the key moments and how you could have improved.
Work on improving your analysis and evaluation skills, not your average centipawn loss rate.
The best advice that I've seen in these forums in the past week.
Yeah, except there's nothing worse than doing self-evaluation incorrectly, and concluding you've made a good, solid sequence of moves as white, when the centipawn evaluation goes from +.5 to -2.3. Then you're just reinforcing erroneous thought processes that the CPU could have spared you from.
When analyzing your games, don't use an engine, ask yourself what were the key moments and how you could have improved.
Work on improving your analysis and evaluation skills, not your average centipawn loss rate.
The best advice that I've seen in these forums in the past week.
Yeah, except there's nothing worse than doing self-evaluation incorrectly, and concluding you've made a good, solid sequence of moves as white, when the centipawn evaluation goes from +.5 to -2.3. Then you're just reinforcing erroneous thought processes that the CPU could have spared you from.
Do the same for master games that have been analyzed by masters, and then check your analysis. Of course, you will make errors in your analysis (even masters do), but your skills will improve over time with lots of practice.
When analyzing your games, don't use an engine, ask yourself what were the key moments and how you could have improved.
Work on improving your analysis and evaluation skills, not your average centipawn loss rate.
The best advice that I've seen in these forums in the past week.
Yeah, except there's nothing worse than doing self-evaluation incorrectly, and concluding you've made a good, solid sequence of moves as white, when the centipawn evaluation goes from +.5 to -2.3. Then you're just reinforcing erroneous thought processes that the CPU could have spared you from.
This is one reason it's useful to go back over you games after you've gotten better. Analyze them a year later for example. You'll find new lessons.
Engines can be a useful tool though. I'm not sure whether it's best to ignore them.
I hear everyone about the concerns about 'missing the analysis and just seeing +1 CPU analysis' and thinking you understand the position when you really don't, but I don't find that the case to be true at all.
I think master or near-master level players can def analye without the engine - in fact, a lot of those positional situations where a master might make a particularly good postiional move, are borderline erroneously evaluated by the CPU, or outright played wrong as the CPU dives into a crazy super risky tactical line that could have been avoided with good solid positional moves.
I think lower rated players like me though, are really well served with the CPU analysis as an aid. Most of our errors are pretty clear-cut, even if they're not gross tactical blunders, and the CPU points out the refutation in 3-5 moves pretty clearly. I learn wayyyyy more from looking at my games with a CPU that shows me where I went wrong, than doing self-analysis and reinforcing all the (many) incorrect thought processes I have without the CPU setting me straight.
You will never see a GM discuss his average centipawn loss rate. Focus on your own evaluations.
Incorrect. I have several books by GMs, and they ALL analyze with the computer, and definitely mention when the CPU favors certain unclear positions, and even give the centipawn amount when needed.
That said, as above, I do notice that at the level of GM play, many subtle but crucial positional decisions they have to make, are not aided by the CPU - Stockfish will give 0.3 advantage to their move, versus 0.12 advantage to another move, even with a long calculation time, but to humans, the GM recommendation is much easier to play and more logical despite the near-equal CPU.
GMs definitely don't just surrender to the will of the CPU by any means though. In my Kings Gambit book by Shaw, he actually routinely recommends lines where the CPU analysis is -1.0 or a whole pawn down for white (he even mentions this in his analysis), but white's setup is so much easier to play that even for a GM that can defend extremely well, they'd prefer that -1.0 position as white. (Rosenreter gambit is an example)
hhnngg, you probably misread either what I wrote, or what the GMs wrote. None of them discuss their average centipawn loss rate even tho almost all use engines to help analyse.
The way to use an engine is after you've done all of your analysis and before you publish it! A computer should be used as a very fancy calculator to check your analysis. It should not be used as a replacement for analysis. And, you shouldn't blindly trust the computer's analysis even then. Work out whether you missed something or whether the engine misevaluates the position.
Computers still misevaluate unusual piece balances, and don't appreciate the difficulty humans have in playing complex positions. A computer's recommendation may be a terrible idea for humans. There are also positions where engines don't realize the position is drawn despite a material deficit. Engines are extremely powerful, but they are still just tools that need to be used correctly.
I hear everyone about the concerns about 'missing the analysis and just seeing +1 CPU analysis' and thinking you understand the position when you really don't, but I don't find that the case to be true at all . . . I think lower rated players like me though, are really well served with the CPU analysis as an aid.
I tend to agree. I think there should be some sort of happy marriage of the two (beginner and engine) available.
One of the most frustrating activities for me as a very new player was reading Silman, and his suggestion of filling a notebook with your own analysis of master games. So I sit down to try this, but I didn't know how to pick good or bad moves at all. Not only was it not useful, it was demoralizing. There was the same sort of "lost" feeling when trying to analyze my tournament games by myself.
Meanwhile I was putting blitz games into Fritz. NOT for the final word. I would explore lines like I couldn't on my own with master games. If the computer liked a move I'd play the position against it to find out why. If it didn't like my move I'd play that position to find out why. I'd explore dozens of sidelines for hours (or until I got tired).
I am not a coach, and I don't know the best methods. This did help me though.
Now that I'm better than a beginner (1900) I think Silman's advice is golden. Fill a notebook with my own analysis? Excellent.
15 minute habbits compared to blitz have a good carry over to classical Imho.
At 15 minute you already need to play decent positional chess. Figure out long strategical plans and you need to calculate decent depth variations in your head, sometimes doublechecking them before moving. (Yes in blitz you can practise more openings but it's the chess after the opening that 99% of chessplayers need to improve at.)
It is important though to use your time accordingly. I meet a lot of sub 1800 players (in the before mentioned ICC 15 minute pool) that treat it like a 5 minute game and end up in a much worse position with 10+ minutes left on the clock.
Ofcourse anyone can blunder but those guys usually just don't take the time to figure out what's going on and that's really shooting themselves in the foot because even if not for the current game figuring out what's going on in a 15 minute game trains your calculation ability and board vision quite well.
That said online I play 80%+ 3 minute blitz or so.. 
15 minute chess is draining and hard work when one takes it seriously.
I still think it's overall terrible advice to tell class players to ignore engines and just do self analysis, UNLESS they have a stronger player available to review their analysis and errors.
In the absence of a coach or a stronger player to correct your erroneous ways, you're just reinforcing bad habits, and in the lower ratings levels, the CPU analysis is usually unequivocally correct and easy to understand even by a lower rated player like myself.
At 2000 range, I def think this equation changes, but I don't think it's fair for a 2000 level player to assume a 1200 level player can break down and analyze positions as accurately as they can and expect to somehow learn from it.
I'm not saying that the engines are the be all and end all of chess analysis, but for the vast majority of lower level uncoached players, they're excellent tools for pointing out your erroneous ways and I think those same players can go really wrong on their own.
I agree.
Solo analysis < analysis with help of an engine < analysis with a trainer.
I wound up taking 3-4 days off from chess around Thanksgiving. Not because of the holiday but because of girlfriend dysfunction. The good news is that the girlfriend is gone and there's more time for chess now, yeah! lol
I'm playing well. The lack of stress around me now and the extra sleep I got the last couple of days did me well. I'm just warming back up with some blitz tactics on CT and blitz games on Lichess although I'm flipping back to ICC in a few games. It's really soft on Lichess compared to here and ICC.
Nice game. Nice pressure on his e3 weakness. Your double pawns are benificial as it limits his knight and help control d4.Then you turn attention to kingside(principle of two weakness).I like that the bishop is on Bf6, preventing your opponent's pawn break.
Thank you. I questioned whether or not to 12. Ne5 or not, resulting in me opening up the d-file after the exchange on 13... and in the end it worked out by restricting his pawn break via the bishop on f5 and battery on the file, as well.
30+0 is the perfect time control. I agree with Greg Sahade, anything more than 30mins leads to boring games.
Anything less leads to lower quality of play. Games should not last more than an hour. They should not last <15mins...
game 30 is a terrible time control.
Games should last at least four hours or be over in a few minutes.
i don't use computer. it is smarter to evaluate yourself