Strategies

PRE-REGISTERED SIGNAL — back the tennis match-odds favourite at the last pre-off snapshot when the exchange-implied probability is inside the frozen band. Found 2026-08-19 (issues #9/#12); this page is its out-of-sample test.

Bets

604

ROI/bet

+4.2%

Hit rate

+72.0%

Since

10 Aug

EXPLORATION BASELINE — back the Scandi football match-odds favourite (3-way) at the last pre-off snapshot. No edge claimed; tracks favourite-longshot bias while the sample grows.

Bets

21

ROI/bet

-4.0%

Hit rate

+57.1%

Since

14 Aug

EFFICIENCY CONTROL — back the WIN favourite in every settled race. The WIN book calibrated clean in research/h4_win_place, so this should bleed roughly the commission; it shows what no-edge looks like on the same axes as the signals.

Bets

264

ROI/bet

-14.9%

Hit rate

+34.1%

Since

10 Aug

PRE-REGISTERED SIGNAL — back every WIN runner priced 10-20 at the last pre-off snapshot. Found 2026-08-19 in an 11-way price/rank scan (in-sample +6.7%, n=508), so the scan-shaped risk is real: only the out-of-sample side of the marker counts.

Bets

482

ROI/bet

+4.0%

Hit rate

+8.5%

Since

10 Aug

PRE-REGISTERED SIGNAL — back the third favourite in every race. The strongest band in the 2026-08-19 scan (in-sample +16.1%, n=282, z=1.8) and also the most scan-shaped one; treat the in-sample number as bait and judge only out-of-sample.

Bets

260

ROI/bet

+12.4%

Hit rate

+17.3%

Since

10 Aug

EXPLORATION — lay every WIN runner priced under 4% implied at best lay, flat 10 DKK liability. The racing calibration's one one-sided finding (longshots win less than the book implies); spread-gated to skip empty ladders. Not pre-registered.

Bets

535

ROI/bet

+0.6%

Hit rate

+98.7%

Since

10 Aug

EXPLORATION BASELINE — back the favourite in settled politics markets. Long-dated markets settle rarely, so this accrues slowly; it exists because the politics niche has the deepest liquidity and no model yet.

Bets

2

ROI/bet

-100.0%

Hit rate

0.0%

Since

13 Aug