I was recently asked by a user new to
Fritz what I thought the best way to analyze is. DPA or IA?
OTB players will decide on two or three
candidate moves, then try to visualize the consequences of each move as far
ahead as he can. That’s how programs analyze in the Infinite Analysis mode. Correspondence players can create a tree
of analysis showing the possible consequences of all moves and this is, more or
less, how Deep Position Analysis
works. The engine makes one move at a time and each time it adds the top
evaluated candidate to a growing line of analysis so you end up with not just a
single best line of analysis but a whole tree.
I rarely use DPA simply because
the amount of analysis generated is, for me, overkill. Still, DPA has its advantages.
The most important is that it will analyze lines that look bad but sometimes, not often, turn out to be the
best.
DPA
gives a lot of information about the position. That’s good if you are looking for an opening
novelty because you can look at all the best replies. In fact, my suspicion is that a lot of
opening books these days make heavy use of DPA (or IDeA in Rybka). My problem with DPA is that reviewing all
that information is a royal pain. Just give me the best line.
Here’s a quick look at how you could use
DPA to make an opening book of your favorite opening lines though. In this case, after Fritz was done with the analysis I searched the database and copied in complete games where available.
Alekhine's Defense
1.e4
Nf6 2.Bc4 Nxe4
2...b5
3.Bb3 Bb7 4.d3 c5 5.Nf3 e6 6.e5 Ng8 7.0–0 Nc6 8.c4 b4 9.Bf4 Nge7 10.d4 cxd4
11.Nbd2 Ng6 12.Bg3 Be7 13.Ne4 0–0 14.Bc2 a5 15.Re1 Ba6 16.b3 f5 17.exf6 gxf6
18.h4 f5 19.Neg5 Bb7 20.Qd2 a4 21.Rac1 axb3 22.axb3 Ra4 23.h5 Nh8 24.Bf4 Qe8
25.g4 h6 26.Nxe6 dxe6 27.Bxh6 fxg4 28.Ng5 Rf5 29.Rxe6 Rxg5 30.Qxg5+ Bxg5
31.Rxe8+ Kf7 32.Rf8+ Ke7 33.Re1+ Be3 34.Rxh8 Ne5 35.fxe3 d3 36.Rh7+ Ke6 37.Rxb7
Nf3+ 38.Kf2 Nxe1 39.Bd1 Nf3 40.Bxf3 Ra2+ 41.Kg3 gxf3 42.Kxf3 1–0 (42)
Ivanovic,B (2480) -Kovacevic,V (2555) Sarajevo 1983]
3.Bxf7+
Kxf7 4.Qh5+ Kg8
4...g6
5.Qd5+ e6
a)
5...Ke8
6.Qxe4 d5 (6...Nc6 7.Nc3 d6 8.Nf3 Bf5 9.Qa4 Bg7 10.d3=0.16/10; 6...d6
7.Nc3 Nc6 8.Nf3 Bf5 9.Qa4 Bg7 10.d3=0.16/10) 7.Qe5 Rg8 8.Nc3 c6
9.Nf3 Bg7= 0.20/10 10.Qf4=-0.16/10;
b)
5...Kg7
6.Qxe4 d5 (6...Nc6 7.Nc3 e5 8.Nd5 d6 9.Nf3 Bf5= 0.00/10 10.Qe3=0.00/10;
6...d6 7.Nf3 Nc6 8.d3 Bf5 9.Qc4 h6 10.0–0=0.28/10) 7.Qe5+ Kg8 8.Nc3
e6 9.Nf3 Nc6 10.Qe2=-0.08/10; 6.Qxe4 d5 (6...Bg7 7.d4 Rf8 8.Nf3 d5 9.Qd3 Kg8
10.Bg5=-0.28/10; 6...Nc6 7.Nf3 Qf6 8.Nc3 Nb4 9.d4 Qf5 10.Kd1=0.20/10)
7.Qf4+ Kg7 8.Nf3 Bd6 9.Qe3 Rf8= -0.28/10 10.Nc3=-0.28/10]
...and
on and on it went. For a quick test I
went 10 moves deep and in maybe a minute got 8 pages of analysis. Of course it wasn't very accurate;
you would have to analyze for much longer, but it shows what you can do if
you have the time and are so inclined.
So, if you need and opening book on some offbeat opening or you don't want to spend any money, create your own. All you have to do is let Fritz analyze in the DPA mode, add selected games, copy and paste into Word then throw in some diagrams and you have your own book.
No comments:
Post a Comment