But not a cheater!
I got Fritz for analysis purposes. I learned how to use blundercheck, I know what the numbers mean and such, however I can't always make out why Fritz picks a certain variation as the best. I can see obvious mistakes that I missed such as forks, but when Fritz has a long, drawn out variation without any obvious mistakes it is hard for me to see sometimes why one variation is better than the other. Also, there are times where I'm left wondering why my opponent would not have played a different move than the one Fritz suggested. Is there any way to get more details, help from Fritz, or has anyone worked out a method to see exactly why Fritz suggests a certain variation?
Also, is there a way I can load pgn games directly into fritz. The database stuff is quite confusing for me, and the instructions provided aren't very helpful. Thanks.
Originally posted by amolv06I'm applying at the same university you're attending right now.
But not a cheater!
I got Fritz for analysis purposes. I learned how to use blundercheck, I know what the numbers mean and such, however I can't always make out why Fritz picks a certain variation as the best. I can see obvious mistakes that I missed such as forks, but when Fritz has a long, drawn out variation without any obvious mistakes it is hard for e stuff is quite confusing for me, and the instructions provided aren't very helpful. Thanks.
As you know Fritz evaluates positions in 1/100ths of a Pawn. This can be changed by changing the “Threshold” setting. The default value is set at 30 (about 1/3 of a Pawn) because supposedly that’s considered the value of a tempo according to the programmers. If you set the value at 100 (equal to one Pawn) Fritz will not show you any improvements unless they are a Pawn better, meaning it’ll mostly point out tactical errors. If you set it at say, 15, you’ll get a lot more positional evaluation.
While not very scientific, look at the following position with White to move:
After about 2-3 mins to evaluate the position I got the following results:
Shredder
21. Nge4 (+1.54)
21. Nxf7 (+1.39)
Rybka
21. Nge4 (+0.63)
21. Nxf7 (+0.49)
Both engines recommended 21. Nge4 (White actually played 21. Nxf7 and won easily). Practically speaking what’s the difference between the two moves? Who knows how to utilize 1/10th of a P advantage? So while it's easy to spot tactical errors, engines will not give you positional plans to follow. In a quiet position where not much is happening tactically, several moves may be evaluated within a few hundredths of a Pawn which doesn't mean much.