Average time per move question

This forum is for general discussions and questions, including Collectors Corner and anything to do with Computer chess.

Moderators: Harvey Williamson, Steve B, Watchman

Forum rules
This textbox is used to restore diagrams posted with the fen tag before the upgrade.
Post Reply
User avatar
appleshampogal
Member
Posts: 126
Joined: Sun Jan 12, 2014 8:53 am

Average time per move question

Post by appleshampogal »

Thought experiment; I'm hosting a match between someone and a dedicated chess computer on a sudden death level... Let's say 10 minutes each side. What is the best approach to programming the computer ahead of time so that the computer uses its time in an optimal way? Let's say, that this particular computer only has average time per move levels. One approach I could use would be to divide the average amount of moves in a chess game, which I understand to be 25 moves divided by the amount of minutes initially. Let's also say that my relevant options for average response times are 5, 10, and 30 seconds per move. If I calculate this out, then I get about on average 24 seconds per move for 25 moves in a 10-minute game. Some problems could arise however. For instance the opposing player might Blitz super fast and put the computer in time trouble, what are some good rules of thumb to best optimize time and adapt to circumstances, while allowing the computer to optimize its play in the most intelligent manner?
TracySMiller
Full Member
Posts: 865
Joined: Sun Mar 13, 2016 2:24 am
Location: Kingsport, TN

Post by TracySMiller »

I think this is going to be super-tough. For opponents that are somewhat evenly matched, the # of moves/game is going to be much higher than 25. Maybe more like 50-60. Your best bet might be to do something like give the human 30 minutes on his clock for a computer setting for 30 sec/move average. Of course, most computers will vary the time actually taken for its moves, making some instantly, and some 1 or more minutes, but will try to average 30 sec/move.
User avatar
appleshampogal
Member
Posts: 126
Joined: Sun Jan 12, 2014 8:53 am

Post by appleshampogal »

That is helpful to know. I'll try that out. What do you think about opponents that use super blitzing to try to flag the computer... When does one reassess in such an instance. Does the programmer compromise? Or hold true to averages through projected move count?
Post Reply