# Computer Chess Club Archives

## Messages

### Subject: Re: Maximum Number of Legal Moves

Author: Uri Blass

Date: 08:16:05 05/09/05

Go up one level in this thread

On May 09, 2005 at 10:55:52, Vincent Diepeveen wrote:

>On May 08, 2005 at 23:44:31, Andrew Shapira wrote:
>
>>What's best known (the smallest) upper bound on the number of legal moves, taken
>>over all chess positions?  An obvious bound is 9*7*4+2*14+2*14+2*8+8 = 332.
>>This comes from 9 queens, 2 rooks, 2 bishops, 2 knights, and 1 king.  Is a
>>better bound known?
>
>220 moves is theoretic maximum. I managed myself to not construct a position
>with more than 216 moves, a year or 12 ago, a few years later when i tried once
>i came to 218 moves.

I believe that it is not possible to generate a position with more than 220
moves and even that it is not possible to generate a position with more than 218
moves but I know no proof for it(it was not important for me to try to prove
it).

>
>In Diep for safety reasons i assume 220 to be the maximum.

practically for games even if you assume 200 moves there will be no problem in
games because I believe that positions with more than 200 legal moves never
happen in the tree.

>
>No one will manage ever to construct obviously a position with more than 220
>legal moves :)

I believe that you are right but it is not a proof.
>
>Instead of asking the maximum number of legal moves, you should be more
>concerned however with the number of semi-legal moves as they can happen in your
>engine.

It is dependent on the engine.
In Movei it cannot happen because it generates only legal moves.

Uri