ELECTRONIC T-NOTES


CHESSBASE USA'S WEEKLY ON-LINE NEWSLETTER


FOR THE WEEK OF FEBRUARY 25, 2001


LEARNING A NEW OPENING -- PART 4

by Steve Lopez

In the last installment, we looked at means for getting advice from Fritz on how to handle the early middlegame in our chosen opening line. This week, we'll look at another great source of information: ChessBase 8.

I once referred to ChessBase's opening report as "the biggest advance in chess database technology" (or words to that effect -- I'm too lazy to go look up my exact words). It's truly a startling feature the first time you see it. Heck, I'm still blown away by it nearly three years after its introduction. If you want to go back and read my initial impressions of it, they're in the July 5, 1998 issue of ETN.

This stage of learning a new opening is the perfect time for using the Opening Report. Not only will it find all the games in a database that use the opening variation we're examining, but it will also provide ideas on what to play next and how to handle the resulting middlegames.

You'll recall that the variation we're examining is in the Queen's Gambit Accepted and goes like this:

1.d4 d5 2.c4 dxc4 3.Nf3 Nf6 4.e3 e6 5.Bxc4 c5 6.0-0 a6 7.a4 Nc6 8.Qe2 cxd4 9.Rd1 Be7 10.exd4 0-0 11.Nc3 Nd5

The first step is to fire up ChessBase 8 and designate a database as the reference database. Usually, this will be your largest database. However, there are a few schools of thought on this selection. One opinion is that the reference database should be your largest annotated database. This makes sense, as the master and GM commentary definitely aids in your understanding of the games. Another school of thought is that the reference database should be your largest database, period (even if unannotated), because the larger number of games will give you a better picture of what's actually been played in top level chess. A third opinion is that a database of nothing but games on the opening in question should be used. That's typically my choice if I already have a database on a particular opening. In a later installment of this series, we'll describe how to do this. For the time being, we'll stick with the first two options. I'd really rather not weigh in on either side of the debate, but I know that I'll get a passel of e-mails from people putting a virtual gun to my head and asking me for advice. So let me head it all off at the pass and say that, for right now, what you'll want to use is your largest annotated database due to the commentary contained therein. This means using either the Mega Database or the ChessBase Opening Encyclopedia (I'll eventually use both, but we'll get to that in a future installment).

To designate a reference database, just right-click on its icon to get a menu of commands. Select "Properties" to get a dialogue box. You'll see a check box that says "Reference database" -- put a check in it and click "OK". It's as easy as that. Keep in mind that you can only have one database designated as the reference database, so your previous choice will no longer apply. (And this is no big deal. It's a whopping four clicks -- I change reference databases several times a day, depending on what I'm doing with the program). For this article, I've selected Mega Database 2001 as the reference database.

Next open a new game window and either make the opening moves manually or cut and paste them into the window. For those following along at home and using the QGA example from these articles, just highlight the variation given above and hit CTRL-C to cut it into the Windows Clipboard. Then tab over to your "new game" window and hit CTRL-V to paste it in. Make sure that 11...Nd5 is highlighted or this next bit of magic won't work properly.

Go to the Tools menu and select "Opening report". You'll see a ChessBase text screen appear. Superimposed over it is a progress bar showing you that the program is working. After a minute or so, you'll see a dialogue appear asking you if you want to "process moves" and "calculate plans". The answer here is a most emphatic "yes!" Make sure both boxes are marked and click "OK". You'll immediately get a "blank" tree window -- no, your computer didn't lock up. It just takes ChessBase a couple of extra minutes to generate the information (note the blue title bar for this window; it says "Working..."). After a bit, you'll be returned to the text window and see the process complete.

This text window is your opening report. We see the variation's moves given, followed by a chessboard graphic showing the last move played by a colored arrow. We next get a link that will bring up all 246 games from Mega DB 2001 in which this variation was played. Don't click on this link yet -- we'll come back to it.

Next we see a section called 1. History. This gives us a bit of history on the opening, with links to the earliest game (chronologically) in which the variation was played, the most recent grandmaster game, and the most recent game:

Earliest game: Grau,R - Piazzini,L 0-1, Buenos Aires 1935
Latest grandmaster game: Pinter,J - Makarov,M 1-0, Elista ol (Men) 1998
Latest game: Mitenkov,A - Afonin,A -, RUS-Cup03 (Geller mem) 1999

The underlined areas in the Opening Report are actually links to the games in question. So if you want to see the variation's debut, you'd click on the "earliest game" link. (Note that the underlined areas in this article are not Internet links, they're merely illustrative of what you'll see in the Opening Report).

Next you see a couple of bar graphs. The first is a straight representation of how often the variation was played in each year covered by the database. Keep in mind that any statistics prior to 1985 are suspect. The "chess information explosion" occurred in the mid-1980's with the introduction of computerized game databasing. There are a lot of events prior to this time (even as recently as the 1970's) that are incomplete or missing entirely from the public record. So you should concentrate on the period from 1985 onward. We see that this QGA variation seemed to reach a peak in 1992. Is it simply because the variation was "in fasion" during that time? This is what the next chart tells us. The "fashion index" has nothing to do with Elle McPherson. It's a graph that shows how often the variation was played compared to the number of times that it statistically should have appeared (given an expected flat rate over the years). We get a figure of 247% for the period 1985 to 1990. This means that the opening appeared 2.5 times more often than would be expected statistically. This drops off to 146% from 1990 to 1995 -- still a goodly number (1.5 times more than the statistical expectation). It plummets to 54% after 1995. So we can see that this opening variation was very much "in vogue" from 1985 to 1995.

At the class level, this information isn't terribly useful or enlightening at this point. Most of our opponents aren't going to play the full eleven move line anyway. But this might come in handy for future reference to see if a line's been "busted" after we reach a point in our chess development at which such things actually matter. At the untitled level, most of us can get away with dang near anything as long as we understand the ideas better than our opponents do. (I know I'll take a ton of flak for that statement, but I've played thousands of chess games in my life against untitled players and I find the statement to be true. Knowledge is power and that's what this whole series is about. I'm not recommending that you run out and start playing 1.a3 just to be different, but I am saying that a variation that's "busted" on the GM level is often still perfectly playable down here in the fishpond).

Our next section reads as follows:

2. Players

a) Strong grandmasters who used this line as Black:
Lajos Portisch Result=4.5/7 1971-1986 Elo-: 2626 games: 7
Jaan Ehlvest Result=3/4 1988-1992 Elo-: 2601 games: 4
Robert Huebner Result=2.5/3 1984-1987 Elo-: 2616 games: 3
Tigran V Petrosian Result=2/4 1963-1981 Elo-: 2600 games: 4
Predrag Nikolic Result=1.5/2 1984-1994 Elo-: 2620 games: 2
Vladimir B Tukmakov Result=1/3 1972-1987 Elo-: 2572 games: 3
Mihail Tal Result=1/2 1985-1986 Elo-: 2582 games: 2
Murray G Chandler Result=1/2 1987-1988 Elo-: 2582 games: 2
Vlastimil Hort Result=1/1 1976 Elo-: 2600 games: 1
Boris V Spassky Result=1/2 1970-1971 Elo-: 2690 games: 2

b) Other notable players:
Matthew Sadler Result=7.5/11 1988-1993 Elo-: 2473 games: 11
Yuri Yakovich Result=6/8 1985-1992 Elo-: 2505 games: 8
Ildar Ibragimov Result=5/8 1989-1997 Elo-: 2520 games: 8
Carlos Garcia Palermo Result=5/10 1980-1991 Elo-: 2464 games: 10
Christer Hartman Result=5/5 1984-1995 Elo-: 2271 games: 5
Sergio Slipak Result=4/7 1992-1994 Elo-: 2427 games: 7
Zeljko Djukic Result=3/6 1989-1996 Elo-: 2401 games: 6

You get the player's name, their results with the variation (points/total games), the range of years in he/she played the variation, the average Elo rating of the player's opposition, and a direct link to the games in question. Note that clicking on a player's name brings up that player's ID card (see ETN January 7th, 2001).

I usually come back to this section later (after I've examined the opening's ideas), but it can be quite useful now. You can see at a glance the caliber of the players who use the variation and maybe even run across the name of a personal favorite player (you know me, I'm definitely coming back to those Tal games later!).

Our next section starts giving us some hard data:

3. Statistics

Black scores well (54%).
Black performs Elo 2481 against an opposition of Elo 2452 (+29).
White performs Elo 2413 against an opposition of Elo 2442 (-29).
White wins: 51 (=21%), Draws: 125 (=51%), Black wins: 70 (=28%)
The drawing quote is high. (20% quick draws, < 20 Moves)

White wins are of average length (42).
Black wins are of average length (40).
Draws are short (30).

Here we get some initial impressions about the opening's success (or lack thereof). The news for White is not good, but it's not terminal either. The variation appears to favor Black. But note the high quotion of draws. The games don't last forever (around 40 moves), so it's a safe guess that you get a logical progression from opening to middlegame to endgame (if the games were pretty long, say 60 moves on average, you'd know to start boning up on your endgame techniques). We also see that draws are common and usually happen around move 30. One may infer from this that the opening is not very dynamic, that it's relatively common for static or balanced positions occur in this opening. Again, this is good news for Black.

We'll talk about statistics later. For now, it's sufficient to know that stats often lie like a cheap rug. Don't put 100% faith in these numbers, but they are useful as guideposts for getting a general feel of how the variation typically plays out (at least at the titled level anyway).

Now we come to the real guts of the Opening Report:

4. Moves and Plans

a) 12.Qe4

85 games, 1964-1998, =1988
White scores below average (50%).
Elo-: 2476, 78 games. Performance = Elo 2451
played by: Polugaevsky, 2635, 0/1; Spassky, 2630, 0.5/1; Petrosian, 2625, 1.5/2; Ljubojevic, 2620, 0/1; Tal, 2610, 0.5/1;

You should play: 12...Nf6 Click for games

[Note: a chessboard graphic appears here in the opening report, but is omitted in this article]

Polugaevsky,L - Hort,V 0-1; Marin,M - Ibragimov,I -; Malaniuk,V - Yakovich,Y 0-1; Petrosian,T - Portisch,L -; Kempinski,R - Ruzele,D 1-0; Ribli,Z - Portisch,L -

Main line:
13.Qe2 Nd5 14.Bd3 Ncb4 15.Bb1 b6 39%, 2364 57 games

Critical line:
13.Qh4 Nd5 14.Qg4 Nf6 15.Qh4 Nd5 48 % Black. Click for games

Plans White:
..(Nd5-f6)/Qe4-e2/Bc4-d3/ ..(Nc6-b4)/Bd3-b1/Nf3-e5 (9) Click for games
..(Nd5-f6)/Qe4-e2/Qe2-e4 (21) Click for games
..(Nd5-f6)/Qe4-h4/ ..(Nf6-d5)/Qh4-g4/ ..(Nd5-f6)/Qg4-g3/ ..(Nf6-h5)/Qg3-g4/Qg4-f4/ ..(Nf6-d5)/Qf4-e4 (2) Click for games
Nf3-e5/Ra1-a3/Nc3-e4/Ra3-h3 (2) Click for games
..(Nd5-f6)/Qe4-h4/ ..(Nf6-d5)/Qh4-e4/ ..(Nd5-f6)/Qe4-e2 (2) Click for games
..(Nd5-f6)/Qe4-e2/Bc1-d2/Nc3-e4 (2) Click for games

Plans Black:
Nf6-d5/Nc6-b4/b7-b6/Bc8-b7 (10) Click for games
Nf6-d5/Nd5-f6/Nf6-h5/Nh5-f6 (3) Click for games
Nf6-d5/f7-f5/Nd5-f4/Nf4xh3 (2) Click for games
Nf6-d5/Nd5-f6/Nc6-b4 (3) Click for games
Nf6-d5/Bc8-d7/Ra8-c8 (2) Click for games
b7-b6/f7-f5/Ra8-c8 (3) Click for games

Alternative: 12...Ncb4 Ljubojevic,L - Huebner,R 0-1; Milov,V - Spangenberg,H 0-1; Portisch,L - Kortschnoj,V -; Petrosian,T - Spassky,B 1-0; Zoler,D - Raetsky,A 0-1; Vera,R - Garcia Palermo,C 0-1

Ooooooooo, baby! I'm starting to feel all tingly in some very interesting areas. This is the part that totally freaked me out back when I first saw the opening report in action. This, my friend, is the true beauty of the opening report. Let's break this down subsection by subsection.

First we get the recommended move for White: 12.Qe4. Note that I've provided only one recommended move for White in this article. There are similar sections for eight other moves for White in the opening report! Hold on a second, I'm getting tingly again...

Next we get the number of games, the span of years in which the move was played, and the median year of this span. We see how White did statistically, derived from the games in the database. We get the average Elo of the players who used this move, the number of database games in which a rating was available for White, and the average performance rating for those players based on their results. Then we see some player names and results; note that each name is a link to the games of that player in which he played the move in question.

Now we get a recommendation for Black (based on the statistics) and a link to those games. This is followed by a chessboard graphic and links to "key " games -- important games in which that move was played by Black.

We next see the main line -- the variation that was played most often, along with an Elo average and a link to those games. After that, we get the critical line. This is the variation that is most noteworthy, based on a combination the number of times it's played, the Elo ratings of the players involved, and the results of those games. This is ChessBase's way of telling you that this is the variation at which you should be looking and concentrating. And, of course, you get a link to those games.

And now we come to the extra-tingly part. ChessBase 8 provides you with the middlegame plans for each side. This isn't done verbally -- it's done via piece placement. The squares where the pieces go will tell you a lot about what's happening in the middlegame. Look at the provided piece placements and think about why the pieces went to those squares. And I'll bet that you don't need to be higher than USCF Class D to figure it out in most cases. ChessBase has done the real grunt work here; it's gone into a database of over 1.6 million games and quickly determined what games apply and where the pieces go in those games. This alone would take weeks of work to do manually. It's consequently not too much to ask of us to take a half-hour to look at the info and do some thinking for ourselves. And this "thinking for yourself" business is what learning a new opening is all about -- by thinking about the variation, you're increasing your understanding and knowledge, hopefully beyond the understanding and knowledge possessed by your opponents. And that's ultimately what ChessBase is all about: knowledge and understanding. Now you know why I'm all tingly.

And we finally come to the final part of this section: a brief subsection of links to an alternate line for Black.

But wait! There's more!

5. Opening keys

Key: 12.--- [followed by the key symbol]

This is a quick link to the opening key for that variation -- another way to get games in which this variation (after 11...Nd5) was played. Note that you only get this section if the reference database has an opening key attached to it.

The opening report feature also provides a statistical game tree in a separate window, which we'll return to in a later article.

So now that you have all of this info, what do you do with it? There are a number of approaches at this point:

1) Click on the first link for all the games in which 11...Nd5 was played. This pulls up a complete list of the games. Look for games with annotations. You can easily pick these out by the "V" and "C" symbols in the right hand column of the game list. I'll typically go through these in chronological order (in general, older games are easier to understand), but do them in any order you like. Play through each game in its entirety and study the notes carefully (and note that in CB8 you can play through the games right in the search result window if you like without having to load each game into a separate window. This obviously works best on big monitors with higher screen resolutions). The important part is not memorization, but getting a feel for the plans and ideas exhibited by each player.

2) Study the options given for White and pay particular attention to the middlegame piece placements provided for both players. The handy-dandy links will take you to the applicable games.

3) Look at the list of strong players who have used the variation. If you see a favorite, review that player's games. If you don't see a fave, pick someone whose name you know.

In all three of these cases (and you really should do all three, in whatever order you choose, though I'm partial to doing them in the order given) you're not looking to memorize opening variations. If you play the kind of people I play, you'll never see these exact variations played anyway. You're looking to learn and assimilate the ideas behind the opening variation, by repetitive viewing of the games if nothing else. When you play through a lot of games you're bound to see patterns develop, patterns that you can remember and use in your own games.

If you're playing along at home, you should now be well on your way to being a complete expert on this variation of the Queen's Gambit Accepted. But what about the other 2,000 variations of this opening? Heh. We'll get to that. But I think this is enough for you to play with for one week -- I'd hate to spoil you.

Until next week...awwwww, you know what to do.

You can e-mail me with your comments, suggestions, and analysis for Electronic T-Notes.