Skip to content

Goal: Rules for Turing chess

Daniel Murell has more thoughts on Turing chess (last discussed here):

When I played with my brother, we had it that if you managed to lap someone while running around the house, then you got an additional move. This means that if you had the option to take the king on your additional move, you could, and doing so won you the game. He was fitter at the time so he slipped in two additional moves over the course of the game. I still won :) I am much better at him at chess though, so I’m sure he would have beaten me had we been more even.

W.r.t. dsquared’s comment and your response, I’m not overly concerned about the first move, because you can enforce that white must reach a halfway point or that some time interval elapse before black makes his first move.

This version though does have one significant weakness that is evident to me. If you wait a little for your opponent to return to make his second move in a row against you, you get your breath back. He couldn’t plan for this tactic since it was your decision to wait. So he’s probably not played his first move to check your king. If he does play is second move and start running and you are in the position to take his king in two moves you can easily do it since you have had a short rest. His only defense would be to see exactly what you are up to when he comes back for his second move, and wait for you to move before he does so he can respond and so that you don’t get two moves in a row. If you also don’t respond, then this give time for him to catch his breath back too :) and so he will now be harder to lap when he finally does move. If you can out sprint him naturally though then he still has a problem and you would both wait indefinitely if both playing optimally. Can you see any solution to this problem? I’m still trying to decide on a rule set that works.

I agree that it would be great to have a standard set of rules for this game, a set of rules that are non-arbitrary (that is, no clock, no tuning parameters other than the length of the circuit to be run) for which the game is playable and is a reasonable balance between chess and running skills.

P.S. A google search turns up this idea:

The rules are the same as standard chess but the chess clock is set up some distance away, say 10 meters, on a separate table.

That’s ok, I guess, but the use of the chess clock makes it less beautiful. I want a pure version where the running is its own clock.


  1. Rahul says:

    Keep a jug of milk (or a pitcher of beer) or some other task handy. If you are waiting at the table and your opponent’s dawdling too long, you get to try and finish it. Then he loses.

    The second task should be such that even the fastest person should need much longer to do it than the slowest conceivable runner around the house. Thus it’d serve as a weapon only against intentional dawdlers and nothing else.

    • oz says:

      This idea reminds me of a much more fun variant: beer chess

    • Daniel says:

      Hi Rahul

      Thanks for this idea, it’s pretty good. I assume it only applies to the player that arrived second at the table, is that correct? Its pretty unpleasant to run with a belly full of beer though. Maybe it could be something like a Sudoku puzzle. This only partly solves the problem. While it prevents the player reaching the table first from waiting, playing just after his opponent plays and sprinting to lap him for a sure 2 in a row move, it doesn’t stop the fitter player from matching the speed of his opponent as he gets close to lapping him. Since he’s fitter, after say two laps of matching his opponents speed, he should find it easy to come to the board second, play his move and then lap his opponent. Can anyone see any solutions to this? This problem might just be inherent in the nature of the game where players run simultaneously.

      There might be a slight problem with your solution I can see though. If the task is, say, a 5 minute task, the fitter player could match his opponents speed, always come to the board second, and use the time his opponent takes to move to do the puzzle, probably completing it before the game is over. How long do you think this task should take?

  2. Peter says:

    Every lap gets you a move. That is, it’s not your opponent’s move which causes you to be able to move, it is making a move and then doing a lap that gets you the ability to move. For the first move, white moves and starts going, and then black moves once white is halfway around the course.

    • Andrew says:


      Will that work? I’m worried that then the faster player can just get extra moves. Suppose you’re just a bit faster than the other guy. Then, at the beginning, you just make a really fast move and run like the wind, then you’ll get 2 moves in a row. This seems contrary to the sprit of the game. Or am I missing something?

    • konrad says:

      This seems like the most natural rule set to me, except I would have both players make their first move simultaneously, then make a move every time they complete a lap. The faster runner effectively has the advantage of playing white, but will have to lap the opponent before gaining another whole move.

      • Daniel says:

        I really like this idea. If the players are unequally matched in fitness (normally the case) it makes the first lapping situation occur in twice the time it would if the first black move was played when white was half way round the track. If they are very close in fitness then lapping won’t be too much of an issue anyway.

    • Andrew says:


      I remain unsatisfied. My image of Turing chess is that player A moves, then runs, then B must move before A returns, then B runs, etc. That is, you think about your move while the other player is running. I’m worried that, under the rules proposed by you and Peter, it’s just running all the time, never a chance to reflect upon the move.

      • Don’t allow two moves in a row, but make it so that if you don’t make your move before the opponent returns from the run you lose immediately. to keep track of whose move it is, put some token on the table that you move back and forth. Say a red checker and whoever has the checker on their side is supposed to move next. So you make your move, move the checker to the other side, run around the house. If you have the checker when you get back you need to move, otherwise you can wait until the opponent comes back and makes his move, and runs around the house before you have to run again. What is wrong with this?

        • Andrew says:


          Now I’m getting confused….

          • It’s like a clock that resets after each move and runs for however long the opponent takes to run around the house.

            Person A moves, and begins running, if person A gets back before person B has moved, person B loses. If person B makes his move, he begins running, and if when he gets back person A hasn’t moved, person A loses. So long as you move before your opponent gets back, you can continue playing until the end by regular chess moves.

            The checker is there just to keep track of whose turn it is so that you aren’t confused when you get back to the table.

            • Daniel says:

              Hi David

              Am I missing something? By this logic, if one player laps the other (lets say they each take 5 seconds to move but A runs faster on average than B), then the game would be lost by B. This is a far worse result than A getting A getting two moves in a row as it means that one player can win the game just by running faster and not worrying about the chess so much.

      • Ben Murrell says:


        Daniel Murrell’s mentioned brother here. I was under the impression that the whole point of the game was to learn to reflect on the moves *while you’re running*. Of course you can’t reflect on the current game state, because you haven’t yet seen your opponent’s latest move, but maintaining conjectures about the moves your opponent will likely play and countering them mentally while trying not to trip down the stairs is what makes the game interesting! I wasn’t a very good chess player, so I’d frequently arrive exhausted at the board thinking “oh crap I didn’t expect that”.

        And I don’t think the “two moves in a row” issue is at all contrary to the spirit of the game – it injects a little stochasticity, giving poorer players the occasional chance at a chance victory, but you shouldn’t expect more sanitary from a game that involves chess and running laps!

  3. John Barrdear says:

    You have, I assume, come across Chess Boxing?

    • Andrew says:


      Yes, every time this comes up on the blog, someone brings up chess boxing. But I don’t like the chess boxing rules. They seem too arbitrary, involving an additional clock. What I’m looking for is a set of rules that are pure.

  4. Paul says:

    How about this: The first player to reach the board is allowed to play, or wait and insist his opponent plays next. Multiple plays by the same player in a row are allowed. BUT, play still alternates between black and white. That means that after lapping your opponent, you’ll both be working to destroy each other’s positions. All rules of chess remain the same, so you can’t put your opponent’s piece actively into check. But if you’re getting ahead, you can put them a move away from checkmate and then beat them in a race to play that checkmate.

    • Daniel says:

      Hi Paul

      Doesn’t this sentence contradict itself? “Multiple plays by the same player in a row are allowed. BUT, play still alternates between black and white.”

  5. noyb says:

    I’m constantly amazed that there are individuals who feel it necessary to change the rules of a game that human beings are incapable of master perfectly in the first place. Enjoy it’s practical infinity and just play the game!

  6. noyb says:

    sorry, *mastering

  7. Daniel Murrell says:

    Considering all the responses here and after playing a few test rounds, we (Eoin and I) have come up with a solution that may work. Unfortunately Andrew, it does require the use of a small timing device but we couldn’t see a way around all the problems inherent in the stronger runner deciding when his two moves in a row will come. The solution is to use a 30 second sand timer and works like this: if a player (usually the weaker runner) feels like he is being abused (stronger runner is tailing him waiting for a chance to move lap and move again), then he can wait for the stronger runner to reach the board and start he timer on him. He has 30 seconds to make his move and loses the game if it’s not made in that time. The player that started the timer has to wait until either the 30 seconds are over or his opponent is some (could be the same amount as at the start of the game) way around the track before playing another move. In this way, it is (usually) the weaker runner that gets to decide when the stronger runner gets his second move in a row, stopping the stronger runner from abusing the rules. If the stronger runner intentionally runs slowly so that he can start a timer on the weaker runner, he gets no advantage from this as he should not be able to lap the weaker runner if he played promptly after the start of the timer. Please could you give feedback on this solution. Of course, if something that doesn’t involve timers works instead here, this would be a lot better, so if you can think of anything, please suggest it.

  8. […] is intriguing. Though it is not entirely clear how the game was intended to be played, and there is some discussion about the rules, it looks like the version played on Sunday makes two moves in a row possible if you manage to […]