FIRST IMPRESSIONS OF FRITZ5

by Steve Lopez

Back when I was a full-time ChessBase staffer (pre-Fritz4), I used to recieve phone calls from beginning and intermediate players who were (paradoxically) simultaneously happy and unhappy with the strength of Fritz. Happy because the strong engine provided them with excellent analysis of their own games. Unhappy because the program was too strong an opponent for the typical club player and because the aforementioned numerical analysis was often difficult to properly interpret.

There's now some great news for my fellow non-titled players: Fritz5, in addition to being a strong playing program with excellent database features, now includes several new tutorial features and introductory playing levels that make it the most valuable training tool that any chessplayer can own!

I've taught many people how to play chess and I've noticed a common problem they've all had: when it's their turn to move, the number of possibilities often boggles them: "So many possible moves...how do I choose?" Fritz5 has a new feature called "Explain all moves", and that's exactly what it does; Fritz generates a list of all the possible moves in the position and provides a brief text explanation of the "why" of each move. The explanations are concise and to the point: "gives check" or "protects b5". Fritz will present the list in order of preference, with the strongest move it finds listed first. The order of the list changes as Fritz searches deeper and finds new possibilities in the position.

This is obviously a wonderful feature for the novice, but even veteran players can make use of it. We've all had the experience of playing through master games and being baffled by some of the seemingly unfathomable moves. "Explain all moves" can often provide simple pointers to help explain why Kasparov or Anand made a particular move.

One of the downfalls of playing exclusively against computers is that they make one tactically lazy. After a while, one stops looking for tactics in one's games. Why try? The program will see the possibility of a future Knight fork three moves before you even spot the pattern and will take steps to prevent it. It's frustrating for the human player; it's said that "98% of chess is tactics", but the human never even gets the chance to sping a tactical shot against a computer opponent.

Fritz5 covers this with a new level called "Sparring" level. I first heard rumors of this feature in 1993, and it's great to finally see it implemented in the program. As you play against Fritz in Sparring mode, Fritz will still look for tactical possibilities for you. But instead of shying away from them, Fritz will actually steer toward them, setting them up for you and even alerting you to the presence of a tactical shot. You can also turn off the "alarm" so that Fritz will still set these combinations up for you, but will not alert you to them as they arise (in order to more closely simulate tournament play).

Another recent development in computer chess is adaptive algorithms. This is a technique by which a computer chess program adjusts itself to your level of play. Face it, if you've ever belonged to a chess club you know who your favorite opponents are: the players who are about equal to you in strength. You get a competitive game and you feel that you have a reasonable chance to win. Adaptive technology allows a chess program to become that favorite opponent, one who challenges you while not overwhelming you.

Fritz5 has a new level, called "Friend" mode, that uses adaptive technology to attempt to match your level of play. The more you win, the stronger it gets. The more you lose, the weaker it becomes. And it plays very human-like chess the entire time. This is not simply a handicap mode in which Fritz plays (as Andre Schulz termed it in ChessBase Magazine 59) "weird chess". Fritz's Friend level plays very natural chess while giving you a real chance to win.

The analysis functions of Fritz have been greatly improved as well. I know I risk incurring the wrath of my comrades at ChessBase by saying this, but Fritz4's natural language analysis left me cold. It was often repetitive and unhelpful. But Fritz5's analysis has been greatly improved. There is now a huge variety in Fritz's commentary, and the annotations are very enlightening.

Fritz does more than just comment on your games now. If you're using the new .CBH file format (introduced in ChessBase 6), Fritz will add colored squares and arrows to the game to illustrate key points, and will place "medals" in the game list to alert you to important themes that occurred in a game it's analyzed. It will also generate "training questions" (like those in ChessBase 6) to quiz you and score your progress as you play back over a game. And it does all of this automatically. Just highlight the game in the game list, click "Analyze", check off the appropriate boxes, and away it goes!

If all of this hasn't whetted your appetite, perhaps a couple of examples will. The following is my very first game with Fritz5, played in Sparring mode:

SL - Fritz5 [C68]
Sparring, 1997
[Fritz 5.00 (180s)]

1.e4 e5 2.Nf3 Nc6 3.Bb5 a6 4.Bxc6 dxc6 5.d4 exd4 6.Qxd4 Qxd4 7.Nxd4 c5 8. Nb3 Nf6N

[8...Be6 9.Nc3 (9.Be3 b6 10.N1d2 0-0-0 11.a4 a5 12.Nc1 Ne7 13.Nd3 f6 14. Nf4 Bf7 15.0-0-0 Nc6 16.Nb1 Bd6 17.Nc3 Rhe8 18.f3 Nd4 19.Kb1 c6 20. Nce2 g5 21.Bxd4 cxd4 22.Nd3 c5 23.Nf2 h5 24.h4 Bc4 25.Nd3 Rg8 26.b3 Bf7 27.hxg5 Rxg5 28.g4 hxg4 29.f4 Rgg8 30.Rh6 Rg6 31.Rdh1 Re8 32.Ng3 Bc7 33.Kc1 Kb7 34.Kd1 Reg8 35.Rh7 R8g7 36.Rxg7 Rxg7 37.Rh6 Rg6 38.Rh1 c4 Biron,R-Datu,I/FEBTC Open, Greenhills PHI 1997/[TWIC]/0-1 (50)) 9... 0-0-0 10.Be3 b6 11.0-0 Ne7 12.Rfd1 Rxd1+ 13.Rxd1 Nc6 14.Nd5 1/2-1/2 Voiska, M-Micic,J/Dresden-W 1995/EXT 95;
8...f6 9.Nc3 Bd6 10.Na5 Ne7 11.Be3 b6 12.Nc4 Be6 13.Nxd6+ cxd6 14. 0-0-0 0-0-0 15.a4 Kc7 16.Rhe1 Kc6 17.Rd2 Rd7 18.Bf4 Rhd8 19.Red1 d5 20. exd5+ 1/2-1/2 Muehlens,R-Kos,S/GER-chJT Leipzig 1995/JUN 95]

9.Nc3 Be6 10.0-0 0-0-0 11.Bg5 h6 12.Bxf6 gxf6 13.f4=/+

[13.Rad1 Bd6=]

13...f5=

[13...c4 14.f5 Bd7 15.Nd2 Bxf5 16.Nxc4=/+]

14.e5-/+ White wins space

[Better is 14.Rad1!?= and White could well hope to play on]

14...b6=

[Better is 14...c4!? 15.Nc1 Rd2-/+]

15.Rad1 Bg7 16.h3=/+

[16.Rfe1 Rxd1 17.Rxd1 f6=/+]

16...Rhe8=

[16...f6 17.exf6 Bxf6 18.Rfe1=/+]

17.a4=/+

[17.Rxd8+ Rxd8 18.Kf2 c4=/+]

17...f6 18.Rxd8+ Rxd8 19.Rd1

[19.Re1 fxe5 20.fxe5 Bf7-/+]

19...fxe5 20.Rxd8+ Kxd8 21.fxe5 Bxe5 22.a5

[22.Nd1 Bd7 23.a5 c4-+]

22...Bxc3 23.bxc3 Ke7

[23...Kd7 24.Kf2-+]

24.Kf2 Bd5

[24...c4 25.Nd2-+]

25.axb6 Bxb3??+- lets it slip away

[25...cxb6 26.g3 Bxb3 27.cxb3 c4 28.bxc4 Kd6-+]

26.bxc7-+

[Better is 26.b7 and White can celebrate victory 26...Bxc2 27.b8Q+-]

26...Ke8+-

[Better is 26...Be6-+ and Black wins]

27.c8Q+ Ke7 28.cxb3 c4 29.Qxc4

[29.bxc4 f4 30.c5 f3 31.Qxa6 h5 32.Qd6+ Kf7 33.c6 fxg2 34.Kxg2 Kg7 35. c7 h4 36.c8Q Kh7 37.Qf5+ Kg7 38.Qdf8#]

29...a5 30.Kf3 Threatening mate... how? 30...Kd8 31.Kf4

[31.Qc6 Ke7 32.Kf4 a4 33.Kxf5 axb3 34.Qc7+ Ke8 35.Ke6 b2 36.Qe7#]

31...Ke7

[31...Kd7 cannot change destiny 32.Qc5 h5 33.Qb6 Ke7 34.Ke5 Kf7 35.Qf6+ Kg8 36.Qg6+ Kf8 37.Ke6 a4 38.Qf7#]

32.Kxf5

[32.Qd5 Kf8 33.Qd6+ Kg7 34.Kxf5 h5 35.Qg6+ Kf8 36.Ke6 a4 37.Qf7#]

32...Ke8

[32...Kd6 cannot change what is in store for ? 33.Qa6+ Kc7 34.Ke5 a4 35. Qa7+ Kc6 36.b4 a3 37.c4 a2 38.b5#]

33.Ke6 Kd8 34.Qa6

[34.Qc6 a4 35.Qd7#]

34...Kc7 35.Qxa5+

[35.Qa7+ Kc6 36.c4 a4 37.b4 a3 38.b5#]

35...Kc6 36.Qd5+

[36.Qa8+ Kc7 37.Qa7+ Kc6 38.b4 Kb5 39.Kd5 h5 40.Qa5#]

36...Kc7 37.c4 h5

[37...Kb6 a fruitless try to alter the course of the game 38.Kd7 h5 39. Qb5+ Ka7 40.Kc6 h4 41.Qb7#]

38.c5 Kb8 39.Kd7

[39.Qd7 h4 40.c6 Ka8 41.Qb7#]

39...h4 40.Kc6 Ka7 41.Qd7+ Ka6 42.Qb7+ Ka5 43.Qb5#

1-0

Other than my changing high ASCII Informant-style evaluations to a form visible in HTML, I made no changes to the game. The commentary you see is exactly how Fritz presented it.

Notice the symbol for "novelty" at Black's 8th move, followed by a number of game citations. This was generated by Fritz. When I gave Fritz the command to analyze the game, I designated that it should use my Ruy Lopez Exchange database as a reference. It scanned the database and discoverd the novelty in our sparring game, plus pointed out alternatives from master play.

Since this was sparring mode, Fritz looked for tactical shots for me. The blue light started flashing after 25...Bxb3. My goal was to then find the winning tactic. Sad to say, I missed it. My understanding was that a tactical combination starts with a check or capture. Unfortunately, I forgot that it also can be a threat. Looking for a capture, I played 26.bxc7. Fritz immediately alerted me to the fact that this was the wrong move. Upon replaying the game, after 26.bxc7, Fritz again asked me to find the shot. As it turns out, the proper move is 26.b7, guaranteeing a coronation for the lowly pawn.

How about another game? This one was played at five seconds a move:

Fritz5 - SL [A29]
9/20 5"/move, 1997
[Fritz 5.00 (240s)]

1.c4 e5 2.Nc3 Nf6 3.Nf3 Nc6 4.g3 d5 5.cxd5 Nxd5 6.Bg2 Nxc3N

[6...Nb6 7.0-0 Be7

A) 8.a3 0-0 (8...Be6 9.d3 0-0 10.b4 f5 11.Bb2 Bd6 12.Rc1 Qe7 13.Na4 Nxa4 14.Qxa4 a6 15.Rxc6 bxc6 16.Nxe5 f4 17.Qxc6 fxg3 18.hxg3 Kh8 19.f4 +/- Hansen,C-Hector,J/Reykjavik (zt) 1995/Inf 63) 9.b4 Re8 10.d3 Bf8 11. Bb2 Bg4 12.h3 Bh5 13.Rc1 Qd7 14.g4 Bg6 15.Nh4 a6 16.Ne4 Rad8 17.Qc2 Nd5 Gulko,B-Waitzkin,J/San Francisco 1995/Inf 63/[Gulko,B]/1-0 (44); B) 8.Rb1 8...f6 9.d4 exd4 10.Nb5 Bf5 11.Bf4 Rc8 12.Rc1 d3 13.exd3 Nd5 14.Nfd4 Nxf4 15.Nxf5 Nxg2 16.Rxc6 Khalifman,A-Tiviakov,S/Linares 1995/Inf 63/[Khalifman,A]/1-0 (74)]

7.bxc3 Consolidates b4 7...Bf5

[7...e4 8.Ng1 f5 9.f3+/=]

8.Qa4 Bd6 9.Nxe5 Bxe5 10.Qb5 0-0 11.Bxc6 Bxc3 12.dxc3

[Not 12.Qxf5 Bxa1 13.Bxb7 Rb8-/+]

12...Qf6 13.0-0

[13.Bxb7? doesn't lead to the expected results 13...Qxc3+ 14.Kf1 Bh3+ ( 14...Qxa1?! 15.Kg2 Qf6 16.Bxa8 Rxa8 17.Bf4=) 15.Bg2 Qxa1 16.Bxh3 Qxc1+ 17. Kg2 Qc2-+]

13...bxc6 14.Qc4 Rab8 15.Qd4 Qxd4 16.cxd4 Rb1+/=

[16...Rfd8!?= is worth looking at]

17.Rxb1 Bxb1 18.Bf4=

[18.a3 Bc2+/=]

18...Bxa2 19.Ra1 Bd5+/-

[19...Bc4!?=]

20.Rxa7 Rb8+/-

[20...Re8 21.Bxc7 c5 22.dxc5 Rxe2 23.Ra1+/-]

21.f3 Rb1+ 22.Kf2 Do you see the mate threat? 22...Rb8+-

[Better is 22...h5!?+/-]

23.Bxc7

[23.Bxc7 Re8 24.e4+-]

1-0

What do you think (of Fritz's work, definitely not my play)? I particularly wanted you to notice the comments "Do you see the mate threat?" and "Threatening mate...how?". This, plus the fact that Fritz generated two training positions for me after 9...Bxe5 and 10.Qb5, make Fritz a very human-like chess coach. Each training position was a multiple-choice question in which I was to select the best move in the position. Fritz5, upon saving the game, also automatically marked it with a "tactics" medal in the game list, designating it as a game full of tactical themes.

I've had Fritz analyze many more games for me, and some of them contain much more extensive commentary than you see in these sample games. I'm already starting to see patterns to my weaknesses (for example, the first game demonstrated that I really need to tighten up my endgame) and I'm learning what I have to work on to improve my game. I'm sure that the new levels and features in Fritz5 can do the same for you!

I hope this information has whetted your appetite and that you want to know more. In the weeks to come, we'll delve much deeper into the new features of Fritz5. But don't wait! Go ahead, dive in, and explore the new features for yourself. Above all, have fun!

Help me out! I want to know your thoughts and comments about Electronic T-Notes. Please post comments to our ChessBase Users Group or e-mail me directly.