The best way to use Fritz for Analysis

I recently purchased Fritz 10, and like it a lot. I have entered some of my games, and used the analysis feature, etc. but I don't feel like I'm learning anything. How do you analyze your games with Fritz in a productive way?

Originally posted by chesskid001
I recently purchased Fritz 10, and like it a lot. I have entered some of my games, and used the analysis feature, etc. but I don't feel like I'm learning anything. How do you analyze your games with Fritz in a productive way?

I run full analysis and then look at all the improvements fritz suggests. If there is anything tactical that I missed I get my buddy to give me 30 lashes. Missed mate is 50 lashes. Punishment is doubled if I go on to lose the game. I would say this is probably the most motivational way to learn from fritz.

I think it is difficult to use a computer to analyse your games. The temptation is to let Fritz do all the work and see the suggested move as an improvement but not know why. I would suggest using Fritz only to check for tactical oversights AFTER you have sat down and analysed your game by exercising your own brain.
You could try using the Find Position function after a few moves in the opening to find GM games from that position and see what the GM played in such a position. Better still, would be to then find an annotated game from that position to look at.

I agree about using Fritz only after you've tried hard to find improvements yourself. Then my preference is to step through the game and variations in "infinite analysis" mode while displaying the top three moves. Sometimes the top move will be too computer-like to understand but there may be a more human-like move evaluated closely behind in 2nd/3rd place.

Tapestry is right. Engines will only tell you where you went wrong tactically. In a position where planning is involved it won’t be much help in giving you the correct strategy. I was recently looking at a game where Fritz gave White a small advantage. Black surprisingly traded a B for a N with the idea of eliminating a weak P even if it meant giving White a passed P in the process. Fritz changed its evaluation to even more in White’s favor (from about 2/3rds to about 1.25 Pawns). Actually it turned out the game was drawn because White could make no progress. The strange thing is Fritz’ evaluation did not change for nearly 20 moves before it finally realized there was no win. The master playing Black clearly realized he had excellent drawing chances or he would not have initiated a trade that Fritz did not even consider.

The moral of the story is I had to figure out for myself what the reasoning was behind the trade of the B for the N that yielded White a passed P.

I just let fritz 10 analyze a game i lost against a 2000 player. It wasn't much help. As someone wrote, if it's tactics, Fritz will show it, but long term planning...

Originally posted by buddy2
I just let fritz 10 analyze a game i lost against a 2000 player. It wasn't much help. As someone wrote, if it's tactics, Fritz will show it, but long term planning...

But you can be sure that up into the 2000 level, you'll probably miss tactics that fritz will find.

In addition to tactics, I would think it would be really good at endgames too, if you have games that get that far.

Originally posted by buddy2
I just let fritz 10 analyze a game i lost against a 2000 player. It wasn't much help. As someone wrote, if it's tactics, Fritz will show it, but long term planning...

Buddy's statement triggers a question I've had rollin' around in my noggin for awhile: Are computers at all capable of strategizing? Or are they simply tactical wizards?

Originally posted by mcreynolds
In addition to tactics, I would think it would be really good at endgames too, if you have games that get that far.

not so sure about it...I had a game in which after some trades it was going in such of those position where you can have exterior passed pawn against central passed pawn(when exterior wins) and the Fritz was adviceing for those trades(for black)...and after I made its suggestions(3-4 moves)...is suddenly realized that the position is lost for black(exterior pawn)...
maybe if I would had let it to analyze more it would have found it...

Originally posted by basso
Buddy's statement triggers a question I've had rollin' around in my noggin for awhile: Are computers at all capable of strategizing? Or are they simply tactical wizards?

even Fritz is better than many of the gms in strategy. rybka is better than all GMs ever lived in strategy.

so, the time computers were accused being only tactical wizards has passed.

The training/sparring facilities in Fritz are ok! If you have the training database, look at the screenshot I posted a while back. If you don't get a set of GM games, analyse them with training selected and then do the quizes.

The sparring facilities are good too. If you are as strong as Kramnik, play Fritz with it unleashed and see what happens (don't play anti-computer chess, play normal chess). Post the game here with the evals (don't cheat as the evals will tell us).

Originally posted by vipiu
not so sure about it...I had a game in which after some trades it was going in such of those position where you can have exterior passed pawn against central passed pawn(when exterior wins) and the Fritz was adviceing for those trades(for black)...and after I made its suggestions(3-4 moves)...is suddenly realized that the position is lost for black(exterior pawn)...
maybe if I would had let it to analyze more it would have found it...

Hmm... maybe i was wrong, although it is possible that it evaluated both plans as lost, in which case it wouldnt be concerned about which one gives the greater practical chances.

The only thing is, I hear people say "I analyzed the game with fritz, and..." I think that maybe the best way is to use the evaluation profile and find the biggest changes in Fritz's evaluation, and then try to figure out why. or if you have a coach, ask him.