In honor of the Olympics, I got my butt over to the park and played run-around-the-house chess for the first time ever.
As was discussed in the comments thread awhile ago, there seem to be three possible ways to play Turing chess:
1. You make your move and run around the house. The other player has to move before you return. Once you sit down to the table, the other player runs around the house. Then you have to move, etc. You lose if you are checkmated or if you fail to move before your opponent returns to his chair.
2. You make your move and run around the house. The other player has to move before you return, but he does not have to wait until you return to start running. He can start running once he’s moved. Then when you get back you have to move before he gets back, but you can start to run once you’ve moved, etc.
3. You make your move and run around the house. The other player takes as long as he wants and makes his move, then he runs. When you return, you take as long as you want and make your move, then run, etc. In this version, it’s possible to make two moves in a row. It would go like this: player A moves and starts to run, then player B immediately moves, runs so hard that he catches up to and passes A, then sits down and quickly moves before A gets back.
(I am excluding all possible rules that involve chess clocks, because to me a key feature of Turing chess is its beautiful simplicity, that the running provides the time control directly.)
I’m guessing that Turing played version 1. As I wrote in my original post on the topic, I’d guess that this version is more of a chess game than a running race, in that the benefit to running hard (thus very slightly reducing your opponent’s time to make a move) is small compared to the cost (making you too tired to think clearly about the chess game). I’d guess that in version 1 the dominant strategy would be for both players to be jogging, except maybe for a few points during the game where your opponent has a tough choice to make and it’s worth your effort to sprint to give him less time to think. In addition, version 1 has the fatal flaw, at least in theory, that at any time you can simply sit on a stoop halfway around the house and think as long as you want. Sure, this would give time to your opponent, too, but if you think you need the time you’d have that opportunity for delay. I’m assuming this never happened in Turing’s games because out of sportsmanship they’d always run, never stop or even walk.
Version 2 is what we played in the park yesterday (ok, not really yesterday, remember that this blog is on a lag). Also I was lazy so instead of doing the equivalent to “running around the house,” we just ran to a tree and back (about 40m total, I think). The result was a running race with essentially no chess, just exhaustion as we sprinted back and forth. Essentially this was bullet chess but it ended when one runner caught up with the other.
Version 3 could be worth trying but I’m worried about the whole two-moves-in-a-row thing. This could make things interesting but it changes the game on the board from chess to a chess variant. If you’re the faster runner and you’re close to catching up, you can time things just right and do your two moves. The other player can anticipate this, etc. The result is that it’s not quite chess anymore.
Right now I’m thinking that version 1 is the best, if supplemented by a rule saying that you have to run, so that walking, sitting, or standing are not allowed. This enforces some minimal level of effort and maximal level of time for each move. I’ll have to play it some more and see how it works out.