Are there any positions where computers would play a bad move, but we humans know of the existence of a better move?

By “computer” I mean standard engines such as Stockfish, at full power.

Edit: By a “bad” move I mean a move that changes the "shape" of the game.

  • 1
    Depends on what you mean: in many endgame studies chess analysis software have revealed the existence of errors that 'humans' did not know about. Many things that computer programs fail to find are due to optimizations to improve playing strength, and programs fail to find things that are unimportant for normal game play, such as mate in shortest number of moves. So you are asking a complex question that depends on how you configure the computer software, and what truth you are referring to. It probably also depend on your definition of 'bad' -- is postponing a mate by one move bad?
    – user30536
    Mar 12, 2023 at 6:27
  • By bad I mean something that can change the "shape" of a game @user30536 Mar 12, 2023 at 7:07
  • Example: chess.stackexchange.com/questions/19107/… You can find lots of others on the site.
    – Allure
    Mar 13, 2023 at 1:13

1 Answer 1


A good example is the following position from a study by William Rudolph (which is also in the posting linked to by ASTA):

3B4/1r2p3/r2p1p2/bkp1P1p1/1p1P1PPp/p1P4P/PPB1K3/8 w - - 0 1

1. Ba4+! Kxa4 2. b3+ Kb5 3. c4+ Kc6 4. d5+ Kd7 5. e6+ Kxd8 6. f5

The problem here is that White seems to be completely lost no matter what they do, but it is in fact possible for them to lock up the position by giving up even more material.

In general, fortresses are still difficult for standard top engines. It is not hard to come up with positions that current top engines do not understand. It is, however, very difficult to achieve such positions in real games with any reliability.

It is also worth noting that there are engines that have specialised fortress-detection code. For instance, the Stockfish fork Crystal solves the position given above in a matter of seconds. However, these engines are weaker than the top engines in normal positions.

  • +1 but why can’t fortress detection code be added to usual Stockfish, making it stronger? Mar 23, 2023 at 16:52
  • 1
    Fortresses are rare. Even if it was free to handle of them completely correctly (it is likely an unattainable goal to handle them correctly even at high cost, because there are fortresses that are not obvious even to humans), the Elo gain from doing so would in all likelihood be very small. As it stands, additional fortress heuristics would simply take up computing cycles that Stockfish can use to find better moves in normal positions.
    – Polytropos
    Mar 24, 2023 at 10:03

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

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