VVLTC search tune and STC regression #5173
PavelChess
started this conversation in
Ideas
Replies: 0 comments
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
-
Recent VVLTC patches give very good results at VVLTC, but at STC they have significant regression. Why does this happen ?
These patches change some parameters/values. Most of changs were in formulas that don't include depth.
Maybe if more formulas were including depth than STC results were better.
But the main idea of this is not to improve STC results, but to create a graph of optimum values for different TC, for example, 10s+0.1s, 25s+0.25s, 60s+0.6s, 150s+1.5s, and then to find formulas for optimal extrapolation of values. Maybe this can get better results at longer time controls.
Beta Was this translation helpful? Give feedback.
All reactions