Crafty analysis output

[cyberfish]

I am running a bot on FICS that provides computer analysis to users using Crafty as backend, and in one of the games submitted (my code retrieves the game from FICS, puts it into PGN, and feed it to Crafty), a few positions are seemingly incorrectly evaluated.

http://cyberfish.wecheer.com/annotated_ ... 12699.html

After 26. Ra1 Ra8, Crafty suggests Rh1, and evaluates the position as even, when black is 4 pawns up.

Any idea why?

Many thanks!

[krazyken]

a repetition of position is drawish to crafty, if the opponent's best move before allows a repetition, why not believe it will happen a 3rd time as well?

[cyberfish]

But the position has never appeared before.

[krazyken]

if 26. Rh1 you would have the same position you had after 25. Ke2

[bob]

if 26. Rh1 you would have the same position you had after 25. Ke2

he might be making the classic mistake of thinking "repeated moves" rather than "repeated positions". This comes up from time to time.

[cyberfish]

I see. Thanks! Not sure how I missed that.