I know that the old good question about engines vs humans has been asked multiple times, e. g. here.

But what about random chess (aka Fischer chess or chess960)? Can humanity feel confortable about struggling machines in this game?

Do you know of plays or championships that throws any light about the question?

  • You would need to know how an engine works; how it analyses moves, positions, ect. Does an engine only work in a standard start position? Asking questions like these would help refine your question and result in better answers. – theeppright Dec 19 '14 at 17:48
  • 1
    This is really no different than asking who is best at chess, humans or engines? – Travis J Dec 19 '14 at 20:50
  • 1
    @TravisJ Not necessarily. Removing all opening knowledge could be to the benefit or disadvantage of either side. – David Richerby Dec 20 '14 at 12:15

You have to understand, that a top ten player is superior to an engine in the opening, because he used engines to analyze openings for many years, and because engines usually don't have an opening book on the level of the preparation of a top player.

In chess 960 you basically take away the opening book for both players and at the very top this actually benefits the engine.

So I think there can be no doubt, that engines would be similarly superior in chess960 as in normal chess. But the only Human-Engine match I know of was still won by the human player: Peter Svidler vs Baron, 1,5:0,5 in the year of 2005. Of course the big jump in engine strength only came with Rybka in the very same year 2005.

  • you must understand that top 10 players can not remember all the openings all the time with all the complexity of the openings, but even for the stupidest machine this is a super easy tasks. With some lines spawning for more then 20 moves, even top 10 can make mistakes. Here are some of the good players (not top ten making terrible mistakes (Fisher, Spassky, Petrosian): chessgames.com/perl/chessgame?gid=1018015 chessgames.com/perl/chessgame?gid=1044389 chessgames.com/perl/chessgame?gid=1084279. – Salvador Dali Dec 20 '14 at 7:22
  • 1
    Yes, obviously. But without high quality input, there is nothing useful for the machine to remember. If Kramnik would feed his opening preparation to an engine, obviously the machine's perfect recall would make any claims of a human superiority in the opening moot. – BlindKungFuMaster Dec 20 '14 at 10:48
  • @SalvadorDali Completely irrelevant, these were players who had no access to chess engines in their time. Even the game with Anand took place in 1988. – bucketman Dec 20 '14 at 13:17
  • Good point. The Anand game isn't even an example for misremembering. I think I read somewhere, that the line was published somewhere and he just played it without checking it first. – BlindKungFuMaster Dec 20 '14 at 14:47
  • @bucketman I am not getting it. So you mean that if there were no chess engine at the time, when the games were played, no chess player was preparing openings and this is why they have done such terrible mistakes? Do you really think that GMs started to prepare openings only after the rise of engines? Also I added here the games, where majority 1600 rated players would see that there was mistake in the opening, mistakes happens now as well, but it is not obvious that they were made. – Salvador Dali Dec 20 '14 at 20:42

The machine, hands down.

These days the machine is so incomparably stronger than a player, that even if a player comes up with a better long-term strategy, they will still lose to inevitably missing tactics. Even if they don't outright drop material, they will overlook threats and walk into a situation where they need to concede positional advantage to not lose that material.

It's good to understand the role of Engines in human opening preparation. Players would play through various lines against a computer looking for a good position, and using the engine to spot and avoid tactical problems for themselves, while laying tactical problems for the opponent. Thus when they get to the board, if the line played out matches the studies of one player more than the other, they will instantly know which moves to play and the other is in danger of falling prey to the tactics uncovered by the machine.

This is why Fischer invented Random, to get away from excessive analysis. It is impossible for a player to memorize 960 opening schemes. Yet a computer is capable of memorizing, playing itself thousands of matches AFK. But even if you throw that opening book out (and with the new generation following Alpha Zero, this doesn't even make sense) and elite players win the strategy of the opening, yet machines will crush them in the execution.


The basic difference between Chess 960 and Classical Chess is that by switching the pieces around at the start, you've in essence removed the computer's opening book. That would eliminate the potential initial advantage that had afforded, since most humans wouldn't have matched that. However, during the "960" opening and subsequently, the machine would go about business as usual, developing quickly and then searching for all possible tactics. Depending on the depth of search, it should prove superior to the vast majority of humans in that respect.

  • 2
    -1 This is just wrong because it's well known removing opening knowledge benefits the engine considerably. The other answer here is correct. – SmallChess Aug 9 '15 at 13:24
  • Well, let me ask you one question. Why was chess 960 invented in the first place if not to negate the advantage of the opening book, which would be an integral part of a chess engine's program? As I stated, the vast majority of human players couldn't match that. – CConero Aug 9 '15 at 13:58
  • The point is indeed to ignore the effects of opening book. But it was NOT DONE to make it human play stronger against computer. The main point is to avoid opening preparation where players can just exercise imagination and tactics without memorizing like 30 moves into the Sicilian Najord. – SmallChess Aug 9 '15 at 14:00
  • Without any opening pattern, human players would lose one of the most important advantage against a computer. The humans would need to compete with the machine in brutal tactic on move 1. Your answer is totally wrong. Please refer to the other answer by @BlindKungFuMaster, he correctly pointed out the fact about opening book. – SmallChess Aug 9 '15 at 14:01
  • I'm certainly no computer expert, but if you re-read my answer it would be hard to categorize it as "just wrong". It's hard to see how "removing opening knowledge benefits the engine considerably". And I did say that, "Depending on the depth of search, it should prove superior to the vast majority of humans in that respect." – CConero Aug 9 '15 at 14:33

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.