Skip to content

Drop crewmate duo and killed-by boards from the guild stats page - #73

Merged
denverquane merged 1 commit into
developfrom
trim-pair-stats
Oct 6, 2026
Merged

denverquane merged 1 commit into
developfrom
trim-pair-stats

Conversation

@denverquane

Copy link
Copy Markdown
Contributor

The API stops sending bestCrewmateDuo, worstCrewmateDuo, and killedBy on GET /guild/stats, and now ranks duo and teammate boards with the number of games in mind (Wilson bounds) rather than raw winrate.

  • Remove the three boards, their parsers, sample data, and English strings. The other locales drop the keys on the next Crowdin sync.
  • The "Best duo" hall-of-fame card now shows the best impostor duo.
  • Add a hint to the impostor duo and teammate boards explaining why a 75% pair can place above a 100% one.
  • The parser ignores the retired boards, so this can ship before the API change; a test pins that.

The API stops sending bestCrewmateDuo, worstCrewmateDuo, and killedBy on
GET /guild/stats, and now ranks duo and teammate boards with the number
of games in mind (Wilson bounds) rather than raw winrate.

- Remove the three boards, their parsers, sample data, and English strings.
  The other locales drop the keys on the next Crowdin sync.
- The "Best duo" hall-of-fame card now shows the best impostor duo.
- Add a hint to the impostor duo and teammate boards explaining why a
  75% pair can place above a 100% one.
- The parser ignores the retired boards, so this can ship before the API
  change; a test pins that.

Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
@denverquane
denverquane merged commit c86f014 into develop Oct 6, 2026
4 checks passed
@denverquane
denverquane deleted the trim-pair-stats branch October 6, 2026 01:49
@denverquane
denverquane restored the trim-pair-stats branch October 11, 2026 15:08
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant