Thirty years! Happy Birthday Fritz (2)

by Frederic Friedel
11/9/2021 – In November 1991 ChessBase launched its first chess playing program for PCs. After a modest start it picked up strength and started challenging the best players in the world. In 1992 the reigning World Champion Garry Kasparov played 35 blitz games, on his notebook, a beta of Fritz 3. He won 31, the computer won four. In 1994 Fritz was equal first with Kasparov in the strongest blitz tournament of all time (17 GMs, Elo average of 2625). In a supreme effort, Garry won the playoff.

ChessBase 16 - Mega package Edition 2022 ChessBase 16 - Mega package Edition 2022

Your key to fresh ideas, precise analyses and targeted training!
Everyone uses ChessBase, from the World Champion to the amateur next door. It is the program of choice for anyone who loves the game and wants to know more about it. Start your personal success story with ChessBase and enjoy the game even more.


Kasparov and Fritz

In 1992, I took Fritz 2 (actually a beta version of Fritz 3) with me to Cologne, where World Champion Garry Kasparov was preparing for a match. We installed the program on his HP OmniBook laptop, and he played, compulsively, around 35 blitz games, against it. Fritz won four of them. It was an historic occasion: for the first time in his life Garry had actually lost a game against a computer, albeit at fast time controls and under informal circumstances.

Here are some of the games he played:


You can find the Fritz 2 games, as well as many others – against Hort, Piket, Henley,  Dzindzichashvili, etc. – in Mega Database.

In May 1994 the semiconductor company Intel staged the "World Chess Express Challenge." It was arguably the strongest blitz tournament of all time, with 17 GMs (Elo average of 2625), and one computer, running Fritz 3 on an Olivetti running the latest Intel processor, a Pentium 90 MHz.

To everyone's surprise Fritz beat GMs Chernin, Anand, Cvitan, Gelfand, Wojtkiewicz, Hjartarson, Kasparov, Kramnik and Short (in that order) to finish equal first with Kasparov. The performance calculated for the program was over 2800 points.

After a short pause there was a playoff between the world champion and the presumptuous Fritz. Garry spent the half-hour break pondering his strategy and appeared at the board in high spirits.

"I'm going to teach it a lesson," he said, "and I know exactly how I'm going to do it." In the first game he reached a promising position, but dropped the win in the endgame. Was the unthinkable going to happen? No way! In the following games Fritz met its master, who put it away with three convincing victories and a draw. The grandmasters following the games on a monitor outside the playing hall burst into enthusiastic catcalls as the computer was ground down to mate in the final game. You can find all the games of that memorable event in Mega Database.

The event made the title of a German computer chess magazine.

You can find all the games Fritz played in the Intel World Express Challenge in Mega Database

After the Intel Challenge Garry discussed the even on the premium sports program in German TV. There he played a quick demo game against Fritz – and lost. You can watch it on Youtube.

In 1995 Kasparov played against Fritz4 in London. With the black pieces, Garry went into a Classical Nimzo-Indian battle. He got a R+B+pawns versus R+N+pawns ending. He traded off his bishop for Fritz’s Knight. Then he played a very skilful endgame to win. In the second game, when Fritz had black, it chose to go into the Tarrasch Defense. Garry simplified into an ending with opposite colored bishops, got a draw and won the match.

But the program was getting stronger all the time. Frans Morsch explained:

"Fritz is built around so-called null-move search. In its search Fritz allows one side to move twice (the other side does a null-move). If the position after the null-move does not return a high value in the evaluation function, then clearly the first move did not contain a threat. Detecting such moves before they are searched to the full depth is an excellent method to speed up the search. In its latest version, Fritz manages a ten times speed-up. Selective search unavoidably introduces oversights, but these are few. In tournaments against humans and other programs, Fritz has proven to be a tough opponent when defending difficult positions."

Fritz vs. Deep Blue

Fritz was getting stronger. It had started participating in big opens – and scoring well. In the Bad Godesberg GM-Open it scored 5.5 points in 11 games, with an Elo performance of 2452. In May 1995 Fritz finished with a  performance rating of 2554 in the Katowice GM Open. 

In 1996 the EGON tournament was organized in the Hague, Holland – with fifty strong chess programs being paired with 50 strong GMs and IMs. In one round Fritz had to play a two-game match against former World Champion Anatoly Karpov. In the first game, with the black pieces, Fritz chose to go into a QGD Tarrasch, and the game ended in a draw. Karpov played positionally, tricking Fritz into giving up a little initiative here, and a little more there. He then simplified and took Fritz to an endgame, which he then won rather easily. In the second game, Karpov chose the French defense and held on to draw and won the match. The score, like Kasparov's, was 1½-½. Fritz ended in fifth place overall.

In the same year, playing in Baden, Switzerland, Fritz drew its games against Victor Kortchnoi and Women's World Champion Zsuzsa Polgar. In a Bled tournament, playing against the Slowenian Olympiad team, Fritz scored 8:2, with an Elo performance of 2727. It was slowly climbing the ladder of computer excellence.

In May 1996 Fritz participated in the Eighth World Computer Chess Championship, staged in  Chinese University of Hong Kong. There it played a memorable game against a prototype of IBM's Deep Blue – and won.

In the end Fritz, running on a 90 MHz Pentium, was the winner of the Eighth World Computer Chess Championship, open not just for micros, but for all computers, including giant mainframes.

Here's the decisive game against Deep Blue:


And on it went:

  • In Junuary 1997 in Greece, Fritz won a live TV match against GM Spyridon Skembris 2½-1½
  • In December 1997 in Cologne, Germany, Fritz played in the international IMPULS GM tournament, finishing in third place with 10/13 points and an Elo performance of 2700
  • In May 1998 Fritz beate GM Epishin 2-1 in Kuppenheim, Germany
  • In June 1998 Fritz won the Frankfurt Chess Classic ahead of world class players Ivanchuk, Beliavski, Kortchnoi and Hübner, with an Elo performance of 2780
  • In March 1999 at the CeBit Hannover trade fair Fritz played 1-1 against Garry Kasparov
  • In April 1999 Fritz beat Judit Polgar 5½-2½ (Elo performance 2810)
  • In July 1999 Frankfurt won the Chess Classic 99 ahead of world class GMs PeterLeko, Veselin Topalov, Peter Svidler, Judit Polgar, Christopher Lutz, Alexander Morosewitsch and Michael Adams. The Elo performance was 2825.

​In 1999, a Fritz DVD was sent up to the Russian space station “MIR”, where the cosmonauts could play against it. Fritz burnt up with MIR station when it was de-orbited two years later. 

  • In May 2000 in Rotterdam Fritz was the first program to take part in an official national championship (of humans!) and finished in third place.
  • In September 2001 Fritz was installed into an industrial robot and played in the Danish Industrial Fair in Herning.

In November 2003 Garry Kasparov played a very special match against Fritz in New York. This will be the subject of our next Fritz birthday article

All Fritz history articles


Editor-in-Chief emeritus of the ChessBase News page. Studied Philosophy and Linguistics at the University of Hamburg and Oxford, graduating with a thesis on speech act theory and moral language. He started a university career but switched to science journalism, producing documentaries for German TV. In 1986 he co-founded ChessBase.
Discussion and Feedback Join the public discussion or submit your feedback to the editors


Rules for reader comments


Not registered yet? Register

Z80 Z80 11/11/2021 01:03
@Frederic: I am curious to know what hardware Kasparov may have been using during these informal games against (beta) Fritz 3. I'm sure that if he took the games seriously there was no contest, but it was all so new! Lots of fun reading these stories again, Thank you!

postscript: Fritz 3 runs very nicely on an HP 200LX :)
Frederic Frederic 11/10/2021 11:06
@lajosarpad: Yes, we did a lot for AI in chess, but also had 35 years of great fun doing it. And adventures galore. I will be narrating them in the next four (at least) instalments. And all the interesting people you meet... Hope I am not boring anyone with these "war stories."
Denix Denix 11/10/2021 01:42
It was because of Fritz I became a national champion.
lajosarpad lajosarpad 11/10/2021 11:32
@Frederic yes, times have changed. Back in the day, when super GMs were considered to be stronger than engines and an engine holding its own against the World Champion was something new and fascinating. Fast forward to 2005 when Michael Adams was running for his money, achieving a single draw against the Hydra, if memory serves me well and just a year later the Classical World Champion was accused of using Fritz during the WC match. Nowadays if somebody plays his/her best tournament and shows fantastic chess, suspicion of getting assistance arises immediately. My ironic joke was directed to this kind of suspicion, when good performance is ab ovo suspicious. To have a more serious note, what you and others did in the field of AI, applied to chess is fantastic. Thank you!
Frederic Frederic 11/10/2021 10:17
"Fritz was probably receiving computer assistance..." Very funny, lajosarpad. But you know, at the time a journalist friend actually speculated that there might be clandestine HUMAN intervention, i.e. some very strong grandmaster guiding its moves. After the Kasparov-Deep Blue match possible human assistance was still on some people's mind. How times have changed: today nobody suspects that in Elo 3500 engine ranking tournaments a human may be somehow interfering with a computer's moves...
Fritz2 Fritz2 11/10/2021 12:11
soon2800 soon2800 11/9/2021 07:41
Hi everyone. I am an old engine collector. How do I get Fritz 8 an 9?
physica physica 11/9/2021 03:30
Very nice chronological article. I would certainly buy Fritz 1-13 bundle by F. Morsch converted to the standard UCI protocol with the addition that they would run with modern SMP ("lazy SMP") code, at least the latest version(s).

@lajosarpad The most curious feature with these matches by today's norms were that the engines were allowed to use opening books, which easily boosted their performance at least by ~100 elo imo. Any engine back then couldn't reproduce human opening theory OTB very far on their own. They were too materialistic, "antipositional" and passive.
Davidx1 Davidx1 11/9/2021 01:33
@goeland : you're right, why don't we call Franz Morsch for a special edition?
lajosarpad lajosarpad 11/9/2021 01:27
"To everyone's surprise Fritz beat GMs Chernin, Anand, Cvitan, Gelfand, Wojtkiewicz, Hjartarson, Kasparov, Kramnik and Short (in that order) to finish equal first with Kasparov. The performance calculated for the program was over 2800 points." Fritz was probably receiving computer assistance...
goeland goeland 11/9/2021 10:26
Chessbase has been the precursor wih its database, opening dvds and fritz. Best of the best. Not sure about the evolution tho. Much competition. My only critic about CB is that they never organized a clear interaction, public forum or whatever, with the users : question, feedback, innovation. Nowadays, with digital, all companies know on how to be customer-centric.
Z80 Z80 11/8/2021 11:04

As far as I know the first of the HP Omnibook production was in 1993,

and that laptop in the picture surely is no Omnibook,

does anyone know the model of laptop/notebook is in the picture?

Davidx1 Davidx1 11/8/2021 06:16
The program on dos is much sweeter than the one on windows, it is a different way to play and it feels: you can play with Fritz 3.
Is there a way to translate the first 3 Fritz to uci? I want to compare it with other engines.
Already Fritz 5 for 16 bit is too strong and not very funny, it is already an analysis program.
Theochessman Theochessman 11/8/2021 06:06
The Deutsch overdub is annoying. Maybe add some English subtitles? The Kramnik hairstyle is epic, though!