27

A few years ago Vishy Anand mentioned in an interview that computers had changed the way humans play chess. Obviously computers help a lot in opening preparation, but what other concrete examples are there?

  1. Is there a certain opening line or endgame whose evaluation has changed a lot due to computer analysis?
  2. Have new principles or strategies been found thanks to computers?
  3. Are grandmaster specifically training to play computer moves instead of relying on traditional principles?
  • "Have new principles or strategies been found thanks to computers" - new strategy of defending R+K vs. R endgame on 2nd rank, but I don't remember the details. – user11153 Dec 12 '16 at 14:44
  • I remember an example of an opening line busted by the computer program Fritz; but I don't have a handle to find the evidence. The bust was presented by Helmut Pfleger in the German weekly "die Zeit" some years ago. – jknappen Jul 27 '17 at 9:57
24
  1. Computers allowed the creation of endgame tables, which allow the user to know with 100% certainty if a position can be won, and how to do it. Currently all positions containing 7 or fewer pieces are 100% known. I am unaware of any opening line busted by computers.

  2. Not that I know of, but computers do find answers to some hard questions. So while computers don't change the principles, they can surely change the evaluation of specific positions. For example, it is common for computers to find flaws in pre-computer chess books, even books held in high regard.

  3. No. Grandmasters use computers as a training aid, however.

  • King's gambit was busted by computer but this was not unexpected as the main line was busted by Bobby Fischer 100 years ago. – Joshua Dec 10 '16 at 17:21
  • 1
    @Joshua it was a chessbase april fool joke (I guess) – dreamcrash Dec 11 '16 at 2:05
  • I would somehow disagree on 3. being a strong No. Computer moves aren't necessarily against traditional principles, in fact the evaluation of a move being good or bad is exactly based on some standard criteria (like figures advantage, king's safety, tricks and so on) just taking further the continuations that standard players can see at the board. – gented Dec 12 '16 at 13:48
  • @GennaroTedesco provide a concrete example and I will augment my answer. Thanks! – Tony Ennis Dec 12 '16 at 16:37
  • @TonyEnnis Examples are exactly the way computers elaborate moves, which are based on basic chess ideas, nothing more than that (one could look at the source codes of the major softwares in use, for instance Stockfish, but I believe that goes behind the scope of the questions). – gented Dec 12 '16 at 17:44
16

Anand is absolutely right. Computers have revolutionized the way we play chess, think about chess and prepare.

Tony Ennis mentions tablebases. That is just the tip of the iceberg. Computers have infiltrated almost every aspect of the game. The only place they're banned is when we sit down to play over the board (and perhaps on some online platforms).

At the risk of losing half the population let me make an analogy with the game of cricket. Actually Svidler, Anand, the rest of the population of India and the rest of the Commonwealth, for that matter, will understand. Computer tracking technology has revolutionized cricket and in particular the LBW law. As a result of the technology umpires now interpret the law differently, because computer technology has shown them that balls which hit the batter when he is well down the pitch would have hit the wicket, spin bowlers have a new lease of life and greater opportunities, and batters have to play spin much more carefully and skillfully than they did 20 or 30 years ago. The mindset and the way the game is played has changed.

Putting opening preparation and endgames with 7 or fewer pieces to one side, no serious book involving analysis can be published and succeed without at the very least computer checking the lines. Certain authors, like GM John Nunn, have made particularly skillful use of computers both in his endgame books and books containing analysis of old games.

But, like the computer and video technology in cricket, computers have changed the way we think about chess. We see the calm computer assessment in what looks like a horrendously risky position and realize that, yes, that apparently unsound move is playable. We realize that endgame can be drawn.

I think all the top players have learnt this to some extent but the doyen for me is Karjakin. Time and time again he over stretches and reaches positions which have the fans and expert commentators saying that he has gone too far and is going to lose (or his opponent outplays him in the middlegame) only for him to find a way to thread the needle through to a draw. This belief that apparently indefensible positions can often be defended, often in contradiction to established principles, is a mind-opening that has been brought about by computers and their use in study.

To a lesser extent I think this also happens with attacking moves. Very occasionally in one of the top games you will hear the expert commentators saying something like "Well, the computer is pointing out a fantastic line here. It involves a very unlikely piece sacrifice 3 moves into the combination. He is thinking a long time here. Will he see it? It is a computer move, not a human move" and then a few minutes later "OMG! He's played it!"

On the preparation front it is probably worth mentioning the democraticizing effect of computers. There is now a vast amount of chess knowledge which is available online which 30 years ago was only available if you subscribed to Informator and 40 years ago to Shakmatny Bulletin.

Above a certain level you can also find a lot of your opponents' games in online databases like chess-db. You can download them and quickly scan through them on your computer and hopefully identify lines you can investigate further with a view to playing against them. Even as recently as 10 years ago you mostly had to rely on your personal database of games you had played previously against them or that from co-operating friends.

Are players specifically trying to play "computer" moves? No, they are trying to play the best moves whether they accord with standard principles or not. I think this predates computers. One of the best books on this subject is John Watson's "Secrets of Modern Chess Strategy: Advances since Nimzowitsch" published in 1998. This makes it clear that top players have not been bound by dogmatic principles for quite some time. Computers are just the latest, most powerful tool to aid the human mind in finding chess "truth".

  • I disagree with your cricket analogy. Ball tracking technology is only used at the very top level of cricket (international matches only and, even then, some countries refuse to use it), and has zero impact at the grassroots level. That's not remotely like chess. – David Richerby Dec 10 '16 at 10:23
  • @DavidRicherby I agree ball tracking is only used at test level, but which countries refuse to use it? Just FYI England have today blown their 2 appeals within a few overs of taking the new ball against India. I would be very surprised if the mindset change it has brought about has not impacted the grassroots. Can batters still plant the front foot well down the pitch with impunity against spinners as they could 20 years ago? Or are umpires at that level also taking on board the new thinking and raising the finger? – Brian Towers Dec 10 '16 at 11:32
  • I had India in mind -- have they now accepted it? Umpires have always known that a ball that strikes the batsman half-way down the pitch could go on to strike the stumps but the point is that they're only supposed to give a batsman out if they're sure that it would. I don't see how knowing that ball tracking is available in test cricket would make a club umpire more certain that a particular ball that had just struck the batsman would have gone on to hit the stumps. If it has, they're a bad umpire. – David Richerby Dec 10 '16 at 12:26
  • @DavidRicherby DRS being trialed in current India v England test series. Disagree with your "bad umpire" comment. Umpiring at 1st class level has changed because DRS has shown the ball hitting the stumps when batter hit well down the pitch. That has opened the minds of umpires at that level to consider LBW and sometimes give out whereas that was automatically not out 20 years ago. Club umpires are incompetent if they have not had a similar change of mindset. – Brian Towers Dec 10 '16 at 13:18
  • Your argument makes no sense. The umpire has to determine whether the specific ball that hit my pads would have gone on to hit the stumps. The existence of ball-tracking technology in some other game makes literally zero difference to the question that the umpire in my game is trying to answer. The umpire must be certain that it would hit my stumps if they are to give me out. Ball tracking in some other game does not increase the umpire's certainty in my game. – David Richerby Dec 11 '16 at 16:05

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.