ChessBase Light 2007 - part 5

6/23/2007 – Part Five in our ongoing series about ChessBase Light 2007 continues the discussion of the engine analysis pane. Learn how to display multiple variations as well as lock the engine on a specific position in the new ChessBase Workshop.

ChessBase 14 Download ChessBase 14 Download

Everyone uses ChessBase, from the World Champion to the amateur next door. Start your personal success story with ChessBase 14 and enjoy your chess even more!


Along with the ChessBase 14 program you can access the Live Database of 8 million games, and receive three months of free ChesssBase Account Premium membership and all of our online apps! Have a look today!

More...

In the previous ChessBase Workshop column we began an examination of chess engines, learning some specialized terminology and discovering the basics of the engine analysis pane in CB Light.

We didn't get to the real meat of the engine analysis pane, though: the main view in which variations are displayed. Let's fire up the engine again and have another look at its analysis pane:

 

In the position Fritz10 is analyzing, it's Black's move. The main portion of the engine analysis pane shows the variation which Fritz suggests (and which also assumes best play for both players in Fritz' estimation). The variation begins with 20...a6 and ends with 27.Bd6. The display also provides us with other interesting and useful information.

The first thing we notice after the variation is Fritz' evaluation of the position after 27.Bd6; if both players make the best moves which Fritz has found so far, Fritz evaluates that final position after 27.Bd6 as being 11/100ths of a pawn in White's favor. This is a very, very small edge, so Fritz is calling this an even position (hence the "=" sign).

The next bit of information tells us exactly how deep (in plies -- see the previous ChessBase Workshop column) Fritz analyzed to arrive at its preferred variation. It looked fifteen plies ahead in its brute force search and thirty-three plies ahead in its selective search to arrive at its evaluation.

Next we see how long Fritz took (given in the format HH:MM:SS) to arrive at its presently preferred variation; in this case it took twenty-one seconds.

Finally we see how many individual positions Fritz had evaluated at the point it arrived at its preferred variation. This number is given in kilonodes (i.e. thousand positions). In this example, Fritz had evaluated 21996 kilonodes, or 21,996,000 unique positions.

When you activate a chess engine to analyze a position in ChessBase Light, the engine will run until you stop it; it will literally run for hours if you so choose (barring a power outage or other mishap). In theory, the engine could run forever. So the basic rule of thumb is to let the engine run for as long as you like -- there's no "best" amount of time (and, although in theory the engine will provide better analysis the farther it looks ahead, there does come a "point of diminishing returns" at which the game tree has become so vast that it's really not worth waiting around for the engine to squeeze out that one extra ply depth).

There are two buttons which we didn't consider in the previous column -- the "+" and "-" buttons. Clicking the "+" button causes the chess engine to display one additional variation for each time you click it. So, for example, if we click it once:

 

...the engine will display the best variation it's found so far, along with the second best variation it's discovered. We can see that at the point we stopped the engine, the best variation (from the point of view of Black, the moving side) scored +0.06 for White, while the second best variation scores +0.14 for White. Why does a higher-scoring variation get relegated to "second-best" status? You may recall from the previous column that a positive score means the position is in White's favor while a negative score means that the position favors Black. It's Black to move in the position being analyzed, so higher scores are worse from Black's point of view.

Both variations scroll off the righthand side of the engine analysis pane; you can use the button on the bottom scroll bar to slide the display to the right and see the remaining moves of each variation.

Let's click the "+" button another time:

 

And we see that 20...a6 is still the best bet for Black, but that 20...f5 and 20...h6 aren't far behind (in fact, you can see how Fritz10's evaluation has changed; 20...h6 was its former second-best pick, but further analysis has dropped 20...h6 to third place behind 20...f5. All three of these moves are evaluated within a tenth of a pawn of each other, so for human purposes it'd be fairly safe to play any of the three).

Note, though, that when you choose to display more than one variation, you don't get the same wealth of information (elapsed time, depth, and number of positions analyzed) that you received when you chose to display just a single variation. So let's click the "-" button twice to return to our single-variation display:

 

If you're playing around with this engine analysis feature as you're reading these columns, you might have already noticed another aspect of the engine analysis pane: if you skip to another move in the game you're viewing, the engine stops analyzing the position it was previously working on and begins to analyze the new current board position (i.e. the position that you skipped to). How do you prevent this from happening?

There's an easy way to "lock" an engine into analyzing a particular position -- right-click in the engine analysis pane to get a popup menu, then select "Lock Engine" from this menu:

 

The engine will now continue to analyze the same position while you're free to jump around to other moves in the gamescore. You can unlock the engine at anytime by just repeating the process: right-click and select "Lock Engine" again to unlock it.

We'll have still more information on the engine analysis pane coming up in the next ChessBase Workshop. 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.


© 2007, Steven A. Lopez. All rights reserved.

 


Discussion and Feedback Join the public discussion or submit your feedback to the editors


Discuss

Rules for reader comments

 
 

Not registered yet? Register