208f4a9092
j177: aim capture - log what the model BELIEVED, not what it did
SirStone2026-09-27 18:44:24 +02:00
55e92bc5e0
docs: the range-vs-approach ceiling - melee unavailable vs DrussGT, the 16 deg is a lead-model number, pre-multiply premise dead
pi
2026-09-27 15:05:14 +02:00
07866c99bd
merge j165-ring-commit: port tfil's arrival commitment to tfil_ring (default off)
SirStone2026-09-27 13:55:22 +02:00
51bfa57067
j163: RESULT - the firing floor is a clean negative, and the offline energy corpus missed the live game by 200x
SirStone2026-09-27 13:55:19 +02:00
fe77056459
j165: port tfil's arrival commitment to tfil_ring, default off
SirStone2026-09-27 13:27:02 +02:00
6cfb1698f1
j163: PRE-REGISTER the 2-arm firing-floor A/B (A floor=0 vs B floor=5) before any battle
SirStone2026-09-27 12:59:51 +02:00
64e23e29d1
j162: measure whether the bot ever runs out of energy (offline, no battles)
SirStone2026-09-27 12:58:14 +02:00
4a1f3e1c88
j159: live 2-arm A/B result — geometry draw weighting costs 8.8 damage/run, wins null; keep TR_TFIL_GEO_MODE=off
SirStone2026-09-27 12:44:38 +02:00
7c5bc9ccbd
j159: PRE-REGISTER the tfil geo A/B (geo draw weighting vs uniform) before any battle
SirStone2026-09-27 12:16:36 +02:00
2223ca667e
j154 (default-off): the DERIVED bounded hold TR_TFIL_HOLD_MAX_TICKS + panic release
SirStone2026-09-27 10:23:46 +02:00
38fbc6ecd1
j152 (default-off): geometry shapes the tile DRAW (TR_TFIL_GEO_MODE/TAU) + the offline sweep with the diversity cost
SirStone2026-09-27 10:22:24 +02:00
94ffc63160
j153: the open-loop ruler for "hold when trapped" + the A/B proposal (NOT run)
SirStone2026-09-27 10:16:53 +02:00
a01141c959
j151 (default-off): hard arrival bound TR_TFIL_ARRIVE_TICKS + the offline per-pick ruler that found it
SirStone2026-09-27 09:57:25 +02:00
5e213df8d9
j148: bound the corridor by speed x ticks (TR_{TFIL,STRAFE}_CORRIDOR_TICKS, default 0 = to the wall); UNTESTED
SirStone2026-09-27 08:23:16 +02:00
32ec040e90
j147 results: 180-battle live A/B on the frozen 15-opponent panel
SirStone2026-09-26 23:20:36 +02:00
d21f7ce5f5
j147: the 1-tick fire-detection lag is OURS — measure it, then back-date it (TR_FIRE_LAG)
SirStone2026-09-26 23:08:57 +02:00
de5d02ba3f
j146 ledger: the field shape restores a real safe set offline (filter broken 63.5%->30.4%) and is a live null on damage/run and round wins; 375 battles, 5 arms
SirStone2026-09-26 22:35:17 +02:00
298ea6d586
j146 TFIL: the bullet's own heat becomes tunable (TR_TFIL_BULLET_CORE/AURA, default 10/5 = today) and the field-shape A/B is pre-registered
SirStone2026-09-26 22:16:06 +02:00
5dd921c26b
j145 ledger: 300 battles, the turn tiebreak is a real but small mechanism with an under-powered outcome null
SirStone2026-09-26 22:13:03 +02:00
39c90fd930
j145 TFIL: a turn-cost tiebreak among the SAFE tiles (default-off)
SirStone2026-09-26 21:56:54 +02:00
0e7e124c7f
j144 ledger: the live A/B result, 600 battles, two independent blocks
SirStone2026-09-26 21:43:03 +02:00
0dc5552c73
j139 dotenv: strip trailing inline comments and warn on non-token values
SirStone2026-09-26 15:18:04 +02:00
10473fc211
j138 draw-only: yellow triangle (enemy-heading other point, enemy, us) in the geometry overlay
SirStone2026-09-26 15:14:48 +02:00
189d990812
j136 [modules] inventory: also list the always-on body API
SirStone2026-09-26 14:48:43 +02:00
36c4334487
bitbrain gun README: the gain warning was wrong and contradicted its own verdict section (it recommended the direction measured decisively harmful)
SirStone2026-09-26 14:48:23 +02:00
3903ec71b4
j134 ledger: note phantom_meteor is dead code (not selectable), out of scope
SirStone2026-09-26 13:01:36 +02:00
3237b65f40
j134 ledger: propagated fire fix to all 5 movers via one shared fire_tracker (TR_FIRE_FIX), per-mover catch table 98.888->100, live tick alignment measured (correction belongs on event_tick+2)
SirStone2026-09-26 12:59:35 +02:00
882733819b
j134 fix: tracker must take the mover's switch per-call (instance flag was silently off on the object-literal construction path) + apply the correction on the reading that carries the server's turn-N+1 energy change (two-slot buffer); live-verified alignment
SirStone2026-09-26 12:56:26 +02:00
6ad5d99922
j134 fire fix: share ONE fire_tracker across tfil/ring/strafe/learned/surf (TR_FIRE_FIX, default on); env-gated TR_FIRE_DIAG alignment trace
SirStone2026-09-26 12:40:19 +02:00
bd57db6d63
j133 strafe: correct the comment's measured fire count (67065, catch 98.888%->100.000%)
SirStone2026-09-26 12:25:06 +02:00
3bee4353b2
j133 ledger: appended 'Missed fires + the label question' (catch 98.888->100%, 746/67065 shots were blind: 456 by the server +3*power bonus, 290 by our own same-tick damage; exact label still -0.230, state-conditional model -0.347 -> the observable state is the constraint) + report fixtures
SirStone2026-09-26 12:22:31 +02:00
799438bbe9
j133 strafe fire detection: correct the enemy energy delta for the server's +3*power hit bonus and our own damage, never drop a too-large drop (TR_STRAFE_FIRE_FIX, default on); catch 98.89%->100% of enemy fires on the 70-battle corpus
SirStone2026-09-26 12:18:21 +02:00
0566264b06
j132 strafe display check: also assert the chosen tile's ramp border and value plate reach the SVG
SirStone2026-09-26 11:54:14 +02:00
cc332138b3
j131 learned movement: real bullet-endpoint resolution (TR_LEARNED_REAL_EVENTS, default off) + exact-geometry Gate A/B (inversion NOT fixed; state still the constraint)
SirStone2026-09-26 11:54:11 +02:00
d5047598b5
j132 strafe display check: also assert the chosen tile's heat number reaches the SVG with the heat grid ON and OFF
SirStone2026-09-26 11:53:10 +02:00
4db46f02dd
j132 strafe pick quality: chosen tile's heat in the [strafe] log line (heat=h/10.0 alt=best-other ok=0|1) and ON the tile (green/yellow/red ramp + value), legible with TR_STRAFE_HEAT_GRID=0; deterministic display check vs pathMaxHeat
SirStone2026-09-26 11:51:33 +02:00
8dd9b3b3b5
j130 learned movement outcome label: live panel results (no arm beats strafe; label swap is a dead heat; gap is information) + Gate A report
SirStone2026-09-26 11:35:47 +02:00
a5c893ace6
learned movement: record the pre-registered prediction as partly wrong (right on wins, wrong on the hit-rate mechanism)
SirStone2026-09-26 10:56:02 +02:00
4f332dcb95
learned movement (SBC): live panel results and the direct answer (module is a wash vs strafe on wins; state conditioning is a CI-separated hit-rate gain vs the same mover without it; the label is the wrong quantity)
SirStone2026-09-26 10:55:14 +02:00
ee98827a62
learned movement: danger-map alignment diagnostic (corr(P(arrival bin), P(hit|bin)) = -0.342) in the gate + ledger
SirStone2026-09-26 10:45:26 +02:00
b4e54fc3cd
j127: allocation batch results - forced shares do not beat pattern; allocation not a lever (closed axis)
SirStone2026-09-26 10:45:10 +02:00
2e336452e6
HeadOn gun colour: red -> pure white (#FFFFFF)
SirStone2026-09-26 09:50:08 +02:00
e25a00175a
Pattern gun colour: orange -> green (#00CC44 / #66FF99) so it is not confusable with BitBrain's pink
SirStone2026-09-26 09:40:22 +02:00
2d8b7d7875
j126: read bot settings from a .env file (default ./.env, --env-file flag, TR_ENV_FILE); file wins over shell leftovers, boot report labels (source: .env)
SirStone2026-09-26 09:08:39 +02:00
8e109bae0b
closing j125: verify movement ship end-to-end (clean build SuccessX, full 16-test guard suite all counts match, real default+revert battles) and restructure gun/movement ledgers outcome-first
SirStone2026-09-26 05:06:49 +02:00
5fe28574ab
gun j123 Batch 2 outcome + phase-2 final answer: len6 does not replicate (wash on 33 opponents); the shipped Pattern is already tuned on its match-length/history parameters; no default changed
SirStone2026-09-26 04:54:40 +02:00
1b59b65810
gun j123 Batch 1 outcome + Batch 2 prereg: len6 is the only arm to beat pattern on wins (15 opp), but all 5 arms are wins-positive incl. the bearing-invariant rad_offset; pre-register the 33-opponent confirmation with a structural control
SirStone2026-09-26 04:22:49 +02:00
2a98aba91b
gun j123 Task A+B prereg: expose Pattern's TR_PATTERN_LEN/TR_PATTERN_DEPTH (default parity) and pre-register the 6-arm match-parameter sweep
SirStone2026-09-26 04:01:51 +02:00
467e07a6a4
gun ledger: fix the arm count in the final answer (six configurations, not seven)
SirStone2026-09-26 03:55:32 +02:00
2a5ea5c17b
gun j121 Batch 2 outcome: the wider panel kills the damage hint (bitbrain wash, tmhorizon WORSE); onlyPattern confirmed; phase closed
SirStone2026-09-26 03:54:52 +02:00
c343c00aaf
gun j121 Batch 2: correct the SpinBot path in the wider panel before any battle (it is not under /tmp/tr_bots)
SirStone2026-09-26 03:35:11 +02:00
3fd6db97e7
movement ship: gate v2 fresh-data primary passed (sign-flip p=0.045, CI [+0.02,+0.58]); default TR_MOVEMENT flipped to strafe
SirStone2026-09-26 03:19:33 +02:00
087e26955f
gun j121 Batch 1 outcome: onlyPattern CONFIRMED across the 15-opponent panel; pre-register Batch 2 wider-panel calibration
SirStone2026-09-26 02:58:41 +02:00
514674886d
movement gate v2: pre-register the sign-flip primary test on fresh 300-battle data (before any battle)
SirStone2026-09-26 02:44:22 +02:00
feefc1912e
movement final: confirmation result and ship decision (gate failed on the sign-test leg; default NOT flipped)
SirStone2026-09-26 02:41:08 +02:00
1d8143a15e
gun j121 Batch 1: pre-register the 6-arm Pattern-panel tournament and the decision rules
SirStone2026-09-26 02:22:38 +02:00
ff03e81591
movement final: pre-register the ship criterion before the confirmation battles
SirStone2026-09-26 02:21:09 +02:00
47ce244740
movement j119 Batch 3+4 results: the reversal/dwell and heat-field axes are a clean negative; nothing beats strafe
SirStone2026-09-26 02:18:22 +02:00
1256357f07
movement j119 Batch 3+4: pre-register the reversal/dwell and heat-field arms
SirStone2026-09-26 01:40:03 +02:00
7311aaef5a
movement j119 Task A: make the shipped TFIL heat shape env-overridable (default path byte-identical)
SirStone2026-09-26 01:38:51 +02:00
c07e6d9599
movement ledger: the post-hoc structure of the strafe win (60 points, both batches)
SirStone2026-09-26 01:33:24 +02:00
44e2d191e3
movement ledger: fix two claims in the Batch-1/2 prose
SirStone2026-09-26 01:32:28 +02:00
7d3645da5a
movement Batch 2: the strafe win over shipped tfil replicates; the range target decides nothing
SirStone2026-09-26 01:31:08 +02:00
07766303f5
movement Batch 1: pure strafe (range tilt OFF) beats the shipped tfil on round wins across a 15-opponent panel
SirStone2026-09-26 01:13:29 +02:00
8efa627c05
gauntlet: BitBrain vs Pattern across 32 legacy opponents (does not generalize)
SirStone2026-09-26 00:55:57 +02:00
SirStone
pi