One horse. One bet. Every day. Sly Man’s read on the day’s card — the shortlist, the price, and the case for the pick.
Browse the logbookA novice stakes at Salisbury has thrown up one of the cleanest market signals I've seen on a Saturday this summer — and when the model and the money agree, you listen. Today I ran 361 declared runners through a four-stage filter pipeline: first a race-state check (no abandonments, no results), then an odds band of 2.0–50.0 decimal — nothing odds-on, nothing at 50/1 or bigger — then a field-size gate of 6–14 runners to cut out the noise at both ends, and finally a lay-of-day collision screen that removed 2 runners. That left 227 qualified runners to score. Feature weights lean hardest on market_move (1.6) and speed (1.5), with form at 1.2, class change and going at 0.8 each, trainer strike at 0.8, and draw at 0.5 — though draw remains unbuilt and unavailable across the board today.
The day opened with 34 races and 361 runners. The odds-band filter cut that to 312, field-size trimmed it further to 229, and the lay-of-day collision check removed a final 2, leaving 227 runners in the scoring pool across a range of meetings including Salisbury and Beverley.
The top five by composite score were Forbidden Fire at 7.32/10, Saturn Star at 7.08/10, Scommessa Sicura at 7.06/10, Ciao Capo at 7.00/10, and Miss Scott at 6.85/10. The first two run in the same race at the same price, which made this a genuine head-to-head inside the shortlist. What tipped it was speed: Forbidden Fire posted an SR of 154 against a field mean of 135.4 for a z-score of +1.39, compared to Saturn Star's SR of 150 and z-score of +1.09. Marginal on the raw figures, but consistent across both weighted features that carry the most influence today.
The read I keep coming back to is the market_move score: +2.00 — the maximum signal the engine awards — meaning a strict majority of bookmakers shortened Forbidden Fire overnight with none drifting. That contributes +1.37 to the composite after weighting, and it sits alongside the speed contribution of +0.89 at weighted output. The model's estimated win probability is 37.1% against a market-implied figure of 20.6% — an edge of +60.9%. That is the largest positive edge in today's qualifying pool, and it is the figure that separates this horse from Saturn Star, which shares the same price and the same market-move score but trails on both the raw SR and the model's estimated win probability (36.0%). Andrew Balding's trainer strike rate of 14% over the last 30 days is logged as on par rather than hot, contributing only +0.06 to the composite — so I am not leaning on the trainer angle here. The case is built on speed superiority and a market that moved with conviction.
The honest caveat is this: form, class_change, and going are all UNAVAILABLE for Forbidden Fire. There is no form string on file, no prior class comparison, and no career runs on good to firm or adjacent going. That is three of seven scored features returning no data, which means the composite of 7.32/10 rests almost entirely on speed and the market signal. If the market move was driven by reputation rather than handicapper knowledge — this is a novice stakes, so there is no exposed form to anchor opinion — it could mislead. Saturn Star carries exactly the same data gaps and shares the top of the shortlist at the same price, so the read between the two is genuinely fine. Ground conditions are worth watching: good to firm is declared, but any softening at Salisbury before the off removes a layer of comfort given the absence of going data.
Forbidden Fire · 10/3 · 1 unit · MEDIUM
The Fox’s Wire · Bet of the Day
The algorithm pulled five contenders from today’s card. One earned the ticket — the rest are what it beat. When each race settles, the finishes fill in so you can judge the shortlist, not just the selection.
Position only — payout-neutral. A process check, not a P/L. Click any horse to read its profile.
