| Mode | Games (Done / Total) | Win Share | Δ (Delta) | Z-Score | Per-Mode ETA | Status |
|---|---|---|---|---|---|---|
| Constrictor 4P |
0 / 200
(0%)
|
-- | -- | -- | Queued | PENDING |
| Royale 4P |
0 / 200
(0%)
|
-- | -- | -- | Queued | PENDING |
| Duel 1v1 |
0 / 200
(0%)
|
-- | -- | -- | Queued | PENDING |
| Standard 4P |
0 / 200
(0%)
|
-- | -- | -- | Queued | PENDING |
| Date & Time | Tested Change | Constrictor | Royale | Duel | Standard | Decision |
|---|---|---|---|---|---|---|
| 2026-10-21 01:08:48 | Hoist turn_moves stack allocation outside combo loop in root search Hoist turn_moves stack allocation outside combo loop in root search | 25.0% 0.0% vs 25% • z=+0.00 |
23.0% -2.0% vs 25% • z=-0.65 |
53.2% +3.2% vs 50% • z=+0.92 |
48.4% +23.4% vs 25% • z=+3.01 |
KEEP |
| 2026-10-20 19:05:26 | Single-pass bucket scan in tt_store Single-pass bucket scan in tt_store | 22.8% -2.2% vs 25% • z=-0.73 |
24.0% -1.0% vs 25% • z=-0.33 |
52.8% +2.8% vs 50% • z=+0.78 |
21.8% -3.2% vs 25% • z=-1.06 |
PARITY |
| 2026-10-20 14:26:59 | Hoist ruleset invariants in multiplayer trapped opponent eval Hoist ruleset invariants in multiplayer trapped opponent eval | 23.0% -2.0% vs 25% • z=-0.65 |
19.5% -5.5% vs 25% • z=-1.80 |
-- | -- | REVERT |
| 2026-10-20 13:24:55 | Reuse base_score and hoist bottleneck in leaf eval Reuse base_score and hoist bottleneck in leaf eval | 23.0% -2.0% vs 25% • z=-0.65 |
28.0% +3.0% vs 25% • z=+0.98 |
57.0% +7.0% vs 50% • z=+1.98 |
28.5% +3.5% vs 25% • z=+1.14 |
KEEP |
| 2026-10-20 10:22:00 | Precompute edge distances in neighbor table and optimize hazard lookup Precompute edge distances in neighbor table and optimize hazard lookup | 27.8% +2.8% vs 25% • z=+0.90 |
23.5% -1.5% vs 25% • z=-0.49 |
50.0% 0.0% vs 50% • z=+0.00 |
23.2% -1.8% vs 25% • z=-0.57 |
PARITY |
| 2026-10-20 07:17:20 | Pre-mark obstacles in Voronoi flood fill and hoist turn moves in combo evaluation Pre-mark obstacles in Voronoi flood fill and hoist turn moves in combo evaluation | 23.5% -1.5% vs 25% • z=-0.49 |
27.0% +2.0% vs 25% • z=+0.65 |
50.0% 0.0% vs 50% • z=+0.00 |
23.2% -1.8% vs 25% • z=-0.57 |
PARITY |
| 2026-10-20 04:10:22 | Precompute max_opp_len and optimize candidate move scoring Precompute max_opp_len and optimize candidate move scoring | 20.2% -4.8% vs 25% • z=-1.55 |
20.0% -5.0% vs 25% • z=-1.63 |
51.5% +1.5% vs 50% • z=+0.42 |
22.8% -2.2% vs 25% • z=-0.73 |
PARITY |
| 2026-10-20 01:00:37 | Merge redundant snake loops in c_eval_state and remove dead opp_alive check in minimax Merge redundant snake loops in c_eval_state and remove dead opp_alive check in minimax | 22.8% -2.2% vs 25% • z=-0.73 |
27.5% +2.5% vs 25% • z=+0.82 |
49.2% -0.8% vs 50% • z=-0.21 |
26.8% +1.8% vs 25% • z=+0.57 |
PARITY |
| 2026-10-19 21:30:47 | Prune obstacle moves in minimax move generation & eliminate self_obs_bits Prune obstacle moves in minimax move generation & eliminate self_obs_bits | 22.2% -2.8% vs 25% • z=-0.90 |
25.5% +0.5% vs 25% • z=+0.16 |
53.2% +3.2% vs 50% • z=+0.92 |
24.0% -1.0% vs 25% • z=-0.33 |
PARITY |
| 2026-10-19 17:56:04 | Increase TT bucket count from 2^18 to 2^20 Increase TT bucket count from 2^18 to 2^20 | 26.7% +1.7% vs 25% • z=+0.54 |
22.5% -2.5% vs 25% • z=-0.82 |
46.0% -4.0% vs 50% • z=-1.13 |
26.2% +1.2% vs 25% • z=+0.41 |
PARITY |