The memorable world of chess engines

by Harry Schüssler
9/5/2018 – The Swedish Grandmaster HARRY SCHÜSSLER was a commentator on the Computer Chess World Championships in Sweden. Schüssler admittedly does not understand much about computer chess, per se, but does knows about chess itself. The grandmaster looks back on the Computer World Championship and describes his impressions of the engine games and what lessons can be learned from them. | Photos: ICGA / Günes

Endgame Turbo 5 USB flash drive Endgame Turbo 5 USB flash drive

Perfect endgame analysis and a huge increase in engine performance: Get it with the new Endgame Turbo 5! This brings the full 6-piece Syzygy endgame tablebases on a pendrive. Just plug it in a USB socket and you are set!


Let yourself be inspired!

What a cultural clash! The scene was a trade fair in Stockholm were some of the best chess computers were battling it out for the world championship. I was there commenting on the games. The only problem: I don't know much about computer chess. When I retired as a professional chess player more than 25 Years ago computers were not widely used. ChessBase had just hit the world, and Karpov & Kasparov were still much better than any silicon player.

My lack of computer knowledge revealed itself when some of the spectators asked some very intriguing questions. Unfortunately for me the world championship was held in conjunction with an AI conference. Some 7,000 researchers from around the globe discussed the future of Articifial Intelligence. It was a suitable environment for the chess super machines., but I got a lot of questions from the well-informed spectators, and could only answer a few.

More or less everytime I entered the playing hall and approached the boards I whispered, as is the habit during a chess tournament. What a mistake. Of course, the machines are not disturbed by voices. If they are fed with a suitable dose of electricity they are more than satisfied.

Operators with their programs during the competition | Photo:

You may expect human chess players to be totally absent from a tournament like this. Not so. On each board, there are two players in action — just like in a traditional tournament. These players are called ”operators”. You might assume the computers move the pieces on the board all by themselves. However, they don't. They have highly developed brains — but no arms. 

Each of them needs an operator to read their choice of move on the screen and transform it into a piece movement on the board. The operator has to be alert. Just look at what happened to the newcomer Leela Chess Zero. This program is a "Deep Learning program" and follows the footsteps of the Google product Alpha Zero. It is programmed with basic knowledge of chess rules and then improves by playing millions of chess games against itself. However, Leela is a beginner when it comes to tough competitive play, and for their operators, the scene is new as well.


Komodo knew that this position was a theoretical draw. The machine had access to a database that gave her the answer. When the operator of Leela offered a draw, Komodo could safely decline. But why play on? Black can hardly win, or can she? Well, the operator of Leela had probably been a bit slow in executing the moves. Therefore there was a difference between Leelas' own "internal" clock and the real physical clock next to the board. Leela trusted her digital clock and thought that she had plenty of time. But it is the clock operated by the representative that is the valid one. So Leela — which also seemed to have lacked a built-in security margin — did not understand that she was in time trouble, pondered too long, and lost on time!


My first reaction was that this was a highly unethical behaviour from the Komodo team. But people convinced me that the operator is part of the game and that this is an acceptable manner to take advantage of a weak spot.

The different standard of ethics was also shown in the following position.


In this blocked position it would have been normal to agree to a draw — for players of flesh and blood, that is. The silicon monsters continued to play on for more than 50 moves. They are programmed to continue trying to win if they estimate their position to be just an inch more favourable.

White first carefully avoided three times repetition and showed that there were a lot of squares to put his rooks. After about 30 moves White gave up the pawn on b6 for nothing. This time it was the 50 move drawing rule which had to be avoided. And then she once again continued to manoeuvre the pieces around, apparently without purpose.


It was tediously boring. Surprisingly enough the operators didn't seem to mind at all. I think they have a romantic, very allowing, attitude towards their machine. Everything the machine does is witty and amusing. Like a newfound love.

For me, it was most refreshing to follow "Pauline", created by a chess enthusiast from Boston who has worked with the project in his spare time for a few years.

Pauline was named after the programmer's mother and had some tactical abilities. So I think his mother can be pretty proud! Still, the apparatus was probably more than a thousand Elo points weaker than its rivals. The scene was set for a massacre.


The black position is in ruins. "Baron" executed the poor monarch in style with 24.Rf7+! Nxf7 25.Nxg6+ Ke8 26.Nf6 mate 1-0


But Pauline was an exception that allows us humans to feel pretty strong. In general, the competitors at the World Championship were extremely forceful. And the most brutal of them all was Komodo, bearing the name of the Indonesian Komodo Dragon, a lizard that can grow up to three meters long and often kills its prey by attacking it at the throat. 

Komodo 12 — the computer version — won all the competitions in Stockholm (blitz, the software world championship where all engines run on the same hardware and the "absolute" championship where there are no limits on the machines' hardware).

Have human players anything to learn from these monsters or should we better try to avoid them? Of course we can learn a lot.

My guess is that Magnus Carlsen has studied these programs quite thoroughly. The machines often avoid theoretical main lines, just like the human World Champion. The winner, Komodo, tried to achieve strategically unbalanced situations, relying on its ability to evaluate positional factors and choose the right path in calm waters. To name it after a lizard was perhaps not entirely fitting. Rather, the resemblance to Carlsen was, in my view, striking.

Less advanced players than Carlsen can also learn a lot. Play against them! Don't be frustrated when they make ugly moves and still win. They play seemingly unaesthetic moves because these moves work. They find the exceptions to general rules. This can benefit the play of even strong players. Don't forget that a move like 1.e4 c5 2.Nf3 Nc6 3.d4 cxd4 4.Nxd4 e5 was considered awkward until Evgeny Sveshnikov and others opened our minds.

The top-level chess machines are the creative minds of today. Be inspired! Dare to take the step into the peculiar world of chess programs.

Komodo 12

In computer chess there is no getting past Komodo, a two-time ICGA Computer World Chess Champion. Find out how Komodo can take your game to the next level!



Harry Schüssler is a Swedish Grandmaster. He twice won the Swedish Individual Championship and played six times for Sweden at the Chess Olympiad. At the 1980 Olympiad, he won individual silver on board four | Photo: Lund Akademiska Schackklubb
Discussion and Feedback Join the public discussion or submit your feedback to the editors


Rules for reader comments


Not registered yet? Register

jsaldea12 jsaldea12 12/19/2018 05:33
Reprint with due respect to Dr. Dennis Hassabis, please see if the following is possible (a) or is corredct (b):
(a)Can self-searching Alphazero search and recognize if the visible light that envelops all over Milky Way originates from the center of Milky Way. Light has both positive and negative. Alphazero operates on the same positive and negative. If the light originates from the center of Milky Way , then it proves light escapes , there is no black hole. Black hole is the greatest hunt in science for the past 50 years. Dr. Stephen Hawkings and all astronomers of the world clai9m it exists. But the search for this black hole is almost impossible because no light can escape black hole, thus it is invisible. Dr. Albert Einstein, the author of black hole, until his death in 1956 vehemently denied the existence of black hole. Undersigned agrees. The self-searching neural brain of Alphazero has proven it is deadly accurate in chess and Go . Can it be that such capability of Alp[hazero be used to solve this greatest enigma in science in this 50 years.
(b) Why stockfish lose to Alphaqzero. It appears like this, subject to correction by Dr, Dennis Hassabis : Alphazero is programmed/commanded to self-analyze some 80,000 moves a second from each position as played at that TIME on the board. It means that every position is considered new by Alphazero but is subject to the same programming//command. Then self-searching Alphazero selects the best move thereof. While stockfish has storage of moves per second, by the millions, all activated all the time chess is played. But not all positions are on the stated storage millions of stockfish because there are billions of variant moves in just the first 10 moves, thus Stockfish is at a lost on positions not in its storage.
Jose S. Aldea
December 15, 2018