I wonder if people did the following test for evaluation

[smirobth]

In order for the material values to equal zero, you would have to see clear to the end of the game (so 12000 plies, give or take).

For instance, when you can see clear to the end of the game (1,1/2, or 0) then the material is "immaterial".

Hi Dan,
Would not a program that saw 12000 plies and evaluated every position as only 1, 1/2, or 0 actually play worse in some respects than current programs? The reason is that it would suffer from the same blindness that some programs using tablebases suffer from, such as giving away material for no reason (material is "immaterial") in vastly superior positions merely because with perfect play it is a draw. Of course such a program will never lose, but it usually won't win either, even if playing against weak opponents. Such a program would tend to get a terrible position and then just hang onto a draw by a thread.

[Uri Blass]

I agree with you that evaluating every position as 0 0.5 1 is not good
but a program that search to the end and evaluates only positional advantage may play well.

Uri