Fritz calculation training - part three

...entered the following variation (details of which were provided in the last column):

...and, after starting the analysis/scoring mode, received the following feedback from Fritz:

I mentioned last week that Fritz provides us with tools with which we might discern where we went wrong in our calculation. Looking again at the variation I entered, I'm thinking that the move Bf4 may have been a big mistake -- I'm not happy with the space Black seems to be gaining on the g-file. So how do we check this move specifically (as opposed to checking the entire calculated variation)?

Note the toolbar button which looks like a computer chip. This button activates the Fritz engine in the (hopefully by now) familiar Infinite analysis mode (in which the engine will run until we manually stop it). Clicking on the move ...Bb6 (which immediately precedes Bf4, the move we wish to check) to highlight it, we click the button I just described:

...and immediately see the Fritz engine go to work, just as we normally see it perform in Infinite analysis mode in the regular game window:

At this point Fritz is indicating that its preferred variation from this point (after ...Bb6) starts with an advance of the e-pawn from e4 to e5. If Fritz's suggested variation is played out to ...Rg7, the position will be dead even.

Now that we know what Fritz thinks we should have played, let's see how much worse things would be after our proposed move Bb4. Click once on Bb4 in the Notation pane to jump to that position, then restart Fritz as described above:

And we see that Fritz thinks our proposed move leads to a half-pawn deficit for White. No wonder we scored so poorly on our calculated variation! Does it get still worse? Let's have Fritz analyze the position at the end of our calculated variation and see its verdict:

Stopping the engine again at an eighteen-ply search depth, we see that things get a bit better for White but that he's still behind. So, to sum up, we didn't lose any points for entering any illegal moves, but we didn't score terrifically well with the legal moves which we did enter.

By the way, what happens when/if we enter an illegal move in our calculated variation? Let's find out. We'll begin with the following position from a database game:

It's White to move. Let's start with Nce3 followed by the illegal ...Be4 and then the legal Raa1:

Then we'll start Fritz' analysis of our work by clicking the toolbar button which looks like a computer terminal behind a paper and pencil. After a short wait, we receive this feedback from Fritz:

This is just plain dismal on our part. Fritz eliminates the illegal move and everything after it, simply scoring what remains. Since our entire "calculated variation" consists of just one move, Fritz gives us a score of zero. Zip. Bupkis. Squadoo.

There is a way to check our calculations for illegal moves beforehand, though. Let's input that same illegal variation as before, but this time we'll have the software double-check our work for legality without grading our overall calculated line of play. There's a button on the toolbar for this:

Clicking this button causes Fritz to point out any illegal moves in the calculated variation we've entered so far. Let's click it and see what we get:

Fritz clearly marks any illegal moves by highlighting them in red and placing the word "Illegal" after them.

It's a short column this week, but we've actually covered a lot of ground. Next time around we'll cover the rest of the toolbar and kick around some additional ideas for using the Calculation training feature. Until then, have fun!

You can e-mail me with your comments on ChessBase Workshop. All responses will be read, and sending an e-mail to this address grants us permission to use it in a future column. No tech support questions, please.

© 2008, Steven A. Lopez. All rights reserved.