fix(sports): offer the idle back-off only the kickoffs each board would show

[ChuckBuilds]

What

Football 3.18.7 (#622) stopped a favourites-only board from waking for other teams' kickoffs. Before that, the live loop offered every game on the scoreboard to core's idle back-off. Each kickoff holds the poll at the live cadence for 15 minutes, so on a busy slate the board fetched the whole scoreboard about every 30 s for games it then filtered out.

This PR applies the same rule to every other scoreboard that offers kickoffs. Each one now offers only the games its own live filter would show, and excluded teams are honoured the same way. Since core #772 (in 3.8.1) every offered kickoff gets its own grace, so offering only the games a board shows matters more than it did.

plugin	version	filter used for the offer
afl	1.37.0 -> 1.37.1	_classify_live_game(home_abbr, away_abbr)
baseball	1.59.0 -> 1.59.1	new _classify_live_game() (was an inline chain in update())
basketball	1.44.0 -> 1.44.1	_classify_live_game(..., is_tournament=...): tournament games still offered in tournament mode
hockey	1.44.0 -> 1.44.1	_is_live_game_included()
lacrosse	1.37.0 -> 1.37.1	new _classify_live_game() (was inline)
nrl	1.36.0 -> 1.36.1	_classify_live_game(home_id, away_id): team IDs, since NRL abbreviations collide
soccer	2.41.0 -> 2.41.1	_classify_live_game(home_abbr, away_abbr)
ufc	1.20.0 -> 1.20.1	new _classify_live_game() (was inline; ufc has no excluded list)

• odds-ticker offers no kickoffs, so it is untouched. football already has the rule.
• Baseball, lacrosse and ufc used to filter inline inside update(). That chain moved into a _classify_live_game() that both the live list and the kickoff offer call, so there is one rule per plugin. The behaviour is identical; only the LIVE_PRIORITY_DEBUG reason string is simpler.
• The favourite-matching behaviour pinned by test_favourite_matching.py (test(scoreboards): pin every plugin's favourite matching before family 6 reconciles it #634) is unchanged: all 204 rows pass.
• Boards that show every live game are unchanged.
• Each plugin has a new versions[] entry and a CHANGELOG entry. update_registry.py has been run and --check passes.

Proof

All results below are against plugins main 19fc4a2 and core main e745ae80 (3.8.1 + #774).

Unit tests. Each plugin has a new test_kickoff_offer_follows_the_live_filter.py. It runs a real SportsLive.update() over a canned payload and records what gets offered. It checks that:

• a favourites-only board offers only the favourite's game;
• an excluded team's game is never offered;
• show-all, favourites-only-off and no-favourites boards offer every game;
• the call is still getattr-guarded.

basketball adds checks for tournament mode. For nrl, the favourite and the other game share abbreviations, so matching on abbreviations instead of IDs fails.

Mutation check:

Mutant	Result
main's sports.py	fails in all 8 plugins
soccer offers nothing on show-all	fails
basketball drops is_tournament	fails
nrl matches abbreviations	fails
baseball loses the exclude check	fails
ufc loses the no-favourites fallback	fails

Replay. This reuses football's frozen-time fake-ESPN harness, generalised, on recorded payloads served gzipped. Each run covers 18 h (09:00-03:00 ET). Counts are scoreboard requests.

run	main	this PR
MLB Fri 2026-09-25, favourite LAD (one 22:15 game)	927 req / 19.6 MB wire / 273 MB dec	620 / 13.1 / 183
MLB same day, favourites NYY + LAD	1,361 / 28.7 / 400	1,293 / 27.4 / 381
MLB same day, show all live	1,683 / 35.0 / 489	1,683 / 35.0 / 489 (same games shown)
NHL Sat 2026-10-03 (preseason), favourite LA (one 22:00 game)	645 / 12.7 / 192	513 / 10.1 / 155

The favourites' games were seen live within one poll of going live in every run.

Guards.

• check_version_bump.py --base origin/main with all 8 ids passes.
• Every other check_* passes, including drift.
• All 50 scripts/test_*.py, including test_favourite_matching.py, pass against a fresh core main.

Plugin tests. run_plugin_tests.py for each touched plugin, branch vs main, with PYTHONIOENCODING=utf-8 on Windows:

• Every plugin has +1 pass, which is the new test.
• The failures are the same on both. The only one is baseball's test_baseball_activity.py (KeyError 'plays' on a real-feed fixture).

Previously stacked on #620, which has now merged; this PR has been rebuilt on main.

Related core PR: ChuckBuilds/LEDMatrix#772 (merged).

🤖 Generated with Claude Code

[coderabbitai]

Important

• 🔍 Trigger review

This repository does not receive automatic reviews because it has fewer than 10 stars.

⚙️ Run configuration

• Configuration used: Organization UI
• Review profile: CHILL
• Plan: Advanced
• Run ID: 3cebebbd-ffb8-491d-869e-5fb58c2a4943

• Autopilot · Keep fixing CodeRabbit findings and required CI, and resolving merge conflicts

---

Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.

❤️ Share

• X
• Mastodon
• Reddit
• LinkedIn

_{Comment @coderabbitai help to get the list of available commands.}

[codacy-production]

Up to standards ✅

🟢 Issues 0 issues

Results:
0 new issues

View in Codacy

🟢 Metrics 128 complexity

Metric	Results
Complexity	128

View in Codacy

_{NEW Get contextual insights on your PRs based on Codacy's metrics, along with PR and Jira context, without leaving GitHub. Enable AI reviewer
TIP This summary will be updated as you push new changes.}