Fritz 12....Christopher you will like this....

[Mike S.]

But engines will not evaluate "Bishops on opposite squares" position as 0.00, will they?

Sometimes in such endgames, one side is one or two useless pawns up, causing wrong evals and thus preventing the draw. This can typically lead to a "delay of the 50 move rule" repeatedly, e.g. engine sacs one of the two extra pawns to avoid a 50-draw, and here we go another 50 moves...

[Dr.Wael Deeb]

It depends if you can afford some more (wasting of) testing time. I set draw to "early" in Fritz, because I think the number of games where there are

1. a couple of 0.00 evals from both sides, AND
2. if continued, it would NOT end in a draw anyway

must be extremely small. So I prefer to save some time here, instead of forcing such games to an "ultimate" end.

I also use game length limits because of the engine's tendency to cause huge waste of time in drawn "bishops on opposite square color" endings and similar nightmares. If the limit isn't too small, almost all stopped games are easy to adjudicate.

(I'm not interested in testing for silly bugs or time losses, only for relative chess strength. - A beta tester may need to use other configurations.)

P.S. I set resign to "late" because I think, some fortress-like situations have a better chance to end in draws then, instead of being resigned due to wrong evals. But actually I've never checked if it really has this effect, and how often. But engines definately do resign defendable fortresses sometimes, more or less due to the material balance only. I have no example at the moment, but I guess it's a common observation.

Thanks Mike for the input,I've taken some serious notes here regarding the testing under ChessBase GUI....I am much more experienced with Arena as it's tournament features beside installing winboard/UCI engines are fantastic....apart from that it is outdated in many aspects....
Dr.D

[Dr.Wael Deeb]

Two engines start to play and after they leave the book and play a couple of moves,suddenly boom....Draw accepted...

Many human super GMs do the same in the tournaments, don't they?
Just another step to AI

An interesting point of view I must admit
I think that the GUI here is messing things up and hence my question....
Dr.D

[Dr.Wael Deeb]

But engines will not evaluate "Bishops on opposite squares" position as 0.00, will they?

Sometimes in such endgames, one side is one or two useless pawns up, causing wrong evals and thus preventing the draw. This can typically lead to a "delay of the 50 move rule" repeatedly, e.g. engine sacs one of the two extra pawns to avoid a 50-draw, and here we go another 50 moves...

Yep,happens very frequently and when I have a lucky day I am near the pc to adjudicate the game manualy....
Dr.D