Skip to content

Lookahead: the model gets the drones' flight (-0.47 to +0.05) - #136

Merged
ptescher merged 1 commit into
masterfrom
actors-lookahead
Nov 15, 2026
Merged

ptescher merged 1 commit into
masterfrom
actors-lookahead

Conversation

@ptescher

Copy link
Copy Markdown
Owner

Why the lookahead lost

The lookahead model had no flight. In the model a drone counted in a tower's circle as soon as it was within 60 m. In the arena it flies at its orbit height (55-87 m) and comes down at 3 m/s only within 60 m of the tower, and only a drone below the treetops counts. That is 12-22 s of nothing at every arrival, which the model did not know. Moving a drone that sits low in a circle to another tower was always a gain in the model and a loss in the game.

Measured on the old model (40 seeds, both sides; each taken proposal replayed from the seed with and without it, nothing else proposed for 45 s): 45 proposals made and 25.5 taken per game; promised +17.6, realised -1.9. Drones that were capturing: promised +10.9, realised -16.5 ± 1.3 (204 moves).

What changed

  • lookahead.rs: the model has height, climb, descent, take-off and charging drones (a Model of switches; drift and redeploy are there but off, because they measured worse).
  • Lookahead::with_model, with_cooldown, Model::spare_capturing: for measuring; the defaults do not use them.
  • New example cognition_towers --example lookahead (calibrate, variants).
  • Unit tests for the flight, drift and redeploy; 18-advice.expect: the stale-proposal expectation now says west tower (the old lookahead used to move the drone to the centre there) and the text says what the model keeps.
  • Findings, the rule-by-rule mismatch list and all tables in plans/actors.md step 4.

Numbers (400 seeds, both sides, vs no advisor)

Lookahead Margin Won Lost
as it was -0.47 ± 0.04 268 532
flight +0.02 ± 0.04 420 379
flight + charging (default now) +0.05 ± 0.04 430 370
+ drift -0.04 ± 0.04 404 396
+ redeploy -0.10 ± 0.04 380 419
flight, margin 5 (not in the code) +0.18 ± 0.04 466 333

Oracle enemy positions: +0.05; blind: -0.03 (belief is not the problem). Delays now: 2 s -0.16, 5 s -0.27, 10 s -0.55, 30 s -0.94.

judge --game captain --champion timid --challenger default --seeds 24: lookahead as it was +0.175 (lower bound -0.132, kept out); with the fix +0.384 (+0.048, promoted); SIM001_LOOKAHEAD=0 +0.268 (+0.031).

What did not change

The lookahead is level with no advisor on this arena, not a gain. I did not tune the margin (5 points would give +0.18) and did not touch the default; whether the default commander keeps the lookahead is the owner's call. Still unexplained: why moves of capturing drones still lose (-6.4 ± 1.5), and why drift and redeploy measure worse.

Tests: cargo test --release -p recreation_towers -p cognition_towers passes; clippy clean for the repo's code.

🤖 Generated with Claude Code

…ndings

The model counted a drone in a tower's circle as soon as it was near; in the arena it
comes down from its orbit height first. The lookahead went from -0.47 to +0.05 against a
commander with no advisor. Drift, redeploy and charging are switches of the model; the
example measures each.

Co-Authored-By: Claude <noreply@anthropic.com>
@ptescher
ptescher changed the base branch from actors-advice-seat to master November 15, 2026 02:49
@ptescher
ptescher merged commit a8b88e6 into master Nov 15, 2026
2 checks passed
@ptescher
ptescher deleted the actors-lookahead branch November 15, 2026 05:36
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