Repository navigation
Publish current Test ratings from GAM, Stan, Glicko-2 and TrueSkill - #4
Merged
Merged
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Current ratings were only available from Elo; the other model links led to a January 2024 forecast experiment. Add current Test ratings from GAM, contextual Stan, Glicko-2, and a separate forward-filtering TrueSkill implementation, using all eligible completed matches through the prepared data cutoff.
The website has readable tables, 90% strength intervals, charts, and downloads under Ratings → Other models (Tests). Research keeps the historical evaluation. The current fit uses 1,157 Tests through 2026-10-25, with complete venue coverage after two audited ground mappings. Acquisition, fitting, visualization, publication, and evaluation remain separate commands. Weekly updates retain the previous model publication if a fit fails, while allowing Elo to update and flagging divergent cutoffs.
Validation: 205 local assertions and lint passed; four-chain Stan simulation recovery and opponent-swap checks passed; the actual current fit passed diagnostics with no divergences. All nine website pages and publication downloads validated locally. Elo's numerical CSVs are unchanged after rebuilding dependency provenance. Agy and Claude Code independently reviewed and re-reviewed the changes; neither found remaining substantive blockers. See docs/current_model_review.md.
Current publication is Tests only. Methods retain their own scales and uncertainty assumptions; this change does not select a predictive winner or add player ratings.