I would certainly expect that all SOUND openings will lead eventually to balanced positions, and that White's centipawn advantage derived from moving first is (far) short of a winning advantage and will gradually shrink towards zero as the game goes on.
... but that's just my own belief. I certainly can't offer any sort of hard proof.
Right, I did not try really nonsensical openings. I tried E4, D4, C4, NF3, G3, B3, E3, and A3. All led to an opening score of 0 at some depth (and all greater depths).
Right, I did not try really nonsensical openings. I tried E4, D4, C4, NF3, G3, B3, E3, and A3. All led to an opening score of 0 at some depth (and all greater depths).
You could try a genuine nonsense opening such as 1. f3 just for comparison.
It may (or may not) trend towards zero as well.
Since all opening moves CPs are <=0 at a depth of approximately 50 (let's call this X), perhaps it is too arbitrary of a value to be useful. Perhaps a better metric might be some sort of sum-product of the CP and the inverse of the depth for all depths less than X. If no one else has named this value, then I claim this as the Partlow Opening Score, and I will attempt to publish it for the major openings in the near future.
Certainly if all openings eventually equalize, then one way to distinguish between them is just how laborious and difficult the road to equality is in one opening compared to another.
Would this be (the conceptual equivalent of) the first differential of CP? The rate of change of CP over time?
Thanks for the suggestion. Will try that. Is there an opening move so bad that it is impossible to achieve a draw even with optimal play following the bad opening. I think there are 20 possible first moves. I will hypothesize that at least a few of these are that bad. Will post after analyzing.
Exactly. The integral, or area under the curve. I think this can be determined with brute force, and not that much brute force is really required.
Lets assume one move has been played by white- the first move.
If the engine gives an opening that is equal at a depth of X
and another opening at a depth of > X
then it is probable that the other opening is commonly regarded as "bad " as it is unclear that it is equal or possibally hard to know to moves that prove equal unless you are an engine
I'm new at online chess, so this might just be my own misunderstanding of the concepts.
My assumptions (based on things I have read) are that the deeper the analysis, the more accurate the Centipawn Score. The second assumption is that engines determine CP scores based on optimal play by both White and Black from a given position.
I wanted to see what opening had the best score, so I tried them all on 365 chess which uses stockfish and goes to a depth of 43. Then in an attempt to be even more accurate, I took the optimal position line - which gives the position to a depth of about 10 turns (20 plies), and evaluated those to a depth of 43. So this should give the opening score to a depth of about 63 if my assumptions are correct. What I found is that in all cases the score was 0, since optimal play by both black and white results in a draw for all openings. Hence my conclusion that all openings have a CP of 0 - you just have to go to a depth of between 40 and 63 to get that answer. Is this a correct understanding, and if not, where is the gap in my understanding?