Proof of Chess Position?

[Gamesfreak13563]

Good idea/bad idea: a coin that isn't primarily generated through mining, but generating "chess positions" on your computer in an effort to solve chess. Bitcoin or other SHA-256 coin miners could merge mine to secure the network, but generate less coins.

[Barnacle_Ed]

I'd say "interesting" rather than good or bad, per se.


You'd have to define what "solved" looks like fairly explicitly, and the constraints about what constitutes a legal "solved" game (ie: Can pieces be missing from the board? Are only checkmates a legal "solution"? What about pawns making it to the end of the board?). A chess game is complicated enough that it should work for mining purposes though. That's just theorycrafting rather than any actual mathematical backing atm though.

[ymer]

Sounds like an interesting idea, you mean calculate all solutions and possible scenarios for a random chess game?

Anybody know how many different games can be played using all possible movements? chess could be hacked that way lol

[Gamesfreak13563]

I'd say "interesting" rather than good or bad, per se.


You'd have to define what "solved" looks like fairly explicitly, and the constraints about what constitutes a legal "solved" game (ie: Can pieces be missing from the board? Are only checkmates a legal "solution"? What about pawns making it to the end of the board?). A chess game is complicated enough that it should work for mining purposes though. That's just theorycrafting rather than any actual mathematical backing atm though.

Solved as in "generate every possible chess position, and then find the best move to play in any situation." A 1 MHz processor would take 10^90 years to solve chess, so we'd have no shortage of positions. My only problem is that there's not really a way to store all the possible legal positions, so it'd have to be really well distributed -or- a central supercomputer.

[Barnacle_Ed]

I'd say "interesting" rather than good or bad, per se.


You'd have to define what "solved" looks like fairly explicitly, and the constraints about what constitutes a legal "solved" game (ie: Can pieces be missing from the board? Are only checkmates a legal "solution"? What about pawns making it to the end of the board?). A chess game is complicated enough that it should work for mining purposes though. That's just theorycrafting rather than any actual mathematical backing atm though.

Solved as in "generate every possible chess position, and then find the best move to play in any situation." A 1 MHz processor would take 10^90 years to solve chess, so we'd have no shortage of positions. My only problem is that there's not really a way to store all the possible legal positions, so it'd have to be really well distributed -or- a central supercomputer.

It wouldn't be too difficult to limit the "solution" to a particular subset of moves, rather than every single possible position. Then you could even set the solution at a particular number of moves deep, which would help increase difficulty by making the solution more challenging to find.

[jimhsu]

Sounds like an interesting idea, you mean calculate all solutions and possible scenarios for a random chess game?

Anybody know how many different games can be played using all possible movements? chess could be hacked that way lol

Between 10^40 and 10^10^50, depending on your definition of a "game". Fairly big I'd say.

[MaGNeT]

ChessCoin vs Magnus Carlsen 

[jimhsu]

Go, however, is more future proof. Orders of magnitude more complex than chess (at least for a computer).

And "GoCoin" sounds cool.