You've probably seen (C# + Silverlight 1.1) clobbering JScript at Chess in the Silverlight Chess demo.

The Nintendo DS has a chess program, Clubhouse Games, which isn't very good, as I mentioned earlier here.

So then I wondered who would win at chess: Jscript or the Nintendo DS .


(Caveat: I recognize this isn't comparing anything useful. They have different hardware, computational abilities, the programs are using different algorithms, etc... Regardless, it seemed cute. )

To find out, I had one game on the Nintendo DS that was: DS vs Human. And then I had a second game in the browser that was JScript vs. Human. And then I manually entered the JScript moves for the DS's human opponent, and I manually entered the DS's moves for the JScript human opponent. Regarding keeping fair time per turn: The JScript program had a fixed upper time limit per turn. DS did not, but it had a difficulty setting that indirectly influenced time. I set the DS difficulty to get an approximate time around what the JScript program had.

On a sidenote, I actually had a few cases of the games getting out of sync because I'd mistakenly enter the wrong move. Sometimes, I could play for quite a while with them being out a sync without realizing it. This has some cute parallels to multi-player networked games.


Final score: JScript won 3 games, DS won 1.

The breakdown:
I played 4 games total. 2 for JSsript being white, 2 for DS being white.

DS was white;  JScript was black.
1) Jscript won, 22 moves.
2) JScript won in 48 moves.

JScript was white. DS was black.
3) DS won in 8 moves. This kind of looked like the DS got lucky. JScript gave away the queen.
4) JScript won in 64 moves. Slow game. JScript was well ahead early, but went for a strategy of advancing its pawns. It eventually had 3 queens. JScript was lucky it didn't end in stalemate.


Based on the playing style I observed, JScript seemed soundly better. If I had 100 data points instead of just 4, I'd still bet on JScript.