Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom

inaccuracies, mistakes, blunders based on winning chances #4705

Closed
niklasf opened this issue Nov 20, 2018 · 2 comments · Fixed by #5337
Closed

inaccuracies, mistakes, blunders based on winning chances #4705

niklasf opened this issue Nov 20, 2018 · 2 comments · Fixed by #5337
Labels
good first issue Narrow in scope (but not necessarily easy)

Comments

@niklasf
Copy link
Member

niklasf commented Nov 20, 2018

also for server side analysis
@niklasf niklasf added the good first issue Narrow in scope (but not necessarily easy) label Apr 18, 2019
@niklasf
Copy link
Member Author

niklasf commented Apr 18, 2019

Going from +100 cp to -200 cp is usually a huge mistake, but going from -600 to -900 hardly changes anything.

Therefore the "Learn from your mistakes" feature uses a difference in "winning chances" (as opposed to difference in raw centipawns) to indentify mistakes.

The formulas are here: https://github.com/ornicar/lila/blob/master/ui/ceval/src/winningChances.ts

Graph:
image

Meanwhile server side analysis still uses raw centipawn differences:

https://github.com/ornicar/lila/blob/d228fdcd3d66537da3ec921be56f536ead9f714a/modules/analyse/src/main/Advice.scala#L57-L61

That could be improved ...

abebinder added a commit to abebinder/lila that referenced this issue Jul 24, 2019
@abebinder
server analysis determines judgements based on winning chances
133a1eb 
resolves  lichess-org#4705
@abebinder abebinder mentioned this issue Jul 24, 2019
Merged
@abebinder
Copy link
Contributor

@niklasf following up on this

ornicar added a commit that referenced this issue Aug 9, 2019
@ornicar
Merge branch 'master' into print-ban
00d02fc 
* master:
  Fix the animation of the arrow in the sub-ratings list
  update stockfish.wasm (v0.5.10: emcc 1.38.41, follow official sf)
  fix typo
  consistent inaccurary threshold for retro
  tweak blunder threshold
  privatized method for cp to winning chance. Changed blunder delta to .5
  server analysis determines judgements based on winning chances resolves  #4705
RoepStoep pushed a commit to RoepStoep/lidraughts that referenced this issue May 22, 2020
@abebinder @RoepStoep
server analysis determines judgements based on winning chances
7f55f4c 
resolves  lichess-org#4705
@yuzisee yuzisee mentioned this issue Aug 29, 2023
Closed
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Assignees
No one assigned
Labels
good first issue Narrow in scope (but not necessarily easy)
Projects
None yet
Milestone
No milestone
Development

Successfully merging a pull request may close this issue.

server analysis determines judgements based on winning chances abebinder/lila
2 participants
@niklasf @abebinder