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 logbookThe sharpest signal today is not in the form book — it is in the money, and for once the model agrees with the crowd. The engine scanned all 45 races and 468 runners declared on 28 September 2026, applying a four-stage filter pipeline. Stage one preserved the full 468 after removing abandoned or already-resulted races. Stage two cut to 413 by keeping only runners priced between 2.0 and 50.0 decimal — no odds-on shots, nothing beyond 49/1. Stage three narrowed to 285 by restricting to races with 6–14 declared runners, eliminating both noisy small fields and chaotic big-field lottery races. Stage four dropped four runners whose selections collide with the Lay of the Day, leaving a clean 281-runner pool for scoring. Feature weights applied were: market_move at 1.6 (heaviest), speed at 1.5, form at 1.2, class_change at 1.0, trainer_strike and going each at 0.8, and draw at 0.5 — though draw-bias data is reserved for v2 and contributed nothing today.
The day opened with 45 races and 468 runners. The odds-band filter removed 55 runners priced odds-on or beyond the 50.0 decimal ceiling. Field-size restrictions cleared out a further 128, keeping only races of 6–14 runners. Four runners were then dropped at the Lay of the Day collision stage, leaving 281 runners across a workable set of races to score against the weighted feature matrix.
Five horses separated themselves clearly from the 281-runner pool. Scott's Legacy led the scoring at 7.90/10, driven by a market_move score of +2.00 (a strict steamer — every bookie shortened, none drifted), a speed z-score of +1.85 against its field, and a 24% trainer strike rate from Nigel and Willy Twiston-Davies over 30 days. Annie Edson Taylor came second at 7.29/10 — a genuine steamer in her own right, with solid going form (3-from-9 on soft), but her estimated edge was marginally negative at -12.1% versus market implied, which capped her case. The Anthony Gover scored 7.19/10 with the field's best speed z-score of +1.94 and strong recent form, but a negative going record (0-from-5 on good) and a class rise both weighed against him. Ladies Well at 7.09/10 showed a market edge of +56.1% and clean good-ground form, though class data was unrecognised and the trainer runs at par. Ladrona at 7.06/10 had the day's single largest estimated edge at +97.4%, but no form string on file and no going data — that volume of unknowns holds her back, even with the money arriving. Scott's Legacy separates from the pack on the combination of edge (+59.7%), a clean speed reading, and a trainer operating genuinely hot over 30 days.
The read I keep coming back to with Scott's Legacy is the market_move score. A +2.00 on a weight of 1.6 is the ceiling — strict majority of bookies shortened, none drifted. At 3/1 decimal 4.0, the engine's estimated win probability is 39.9% against a market-implied 22.3%, a positive edge of +59.7%. That gap is not noise. Either the layers know something the morning prices did not, or a horse priced at evens-on territory in terms of true probability is sitting at 3/1 in the market. Either way, I follow the money when the model also backs it up. Speed is the other pillar. An SR of 155 against a field mean of 124.1 produces a z-score of +1.85 — that is among the better speed readings in today's filtered pool. The form string '51-' is short, but the engine's decay-weighted average of +1.18 on the latest first position is solid, and the trainer record of 5/21 = 24% over 30 days is genuinely hot, not just on par. The composite of 7.90/10 is a strong score. The nearest rival, Annie Edson Taylor, is 0.61 points behind. In a model where a point represents material feature separation, that is not a marginal lead. The engine conviction is rated HIGH, and I am not arguing with it today.
Two features returned unavailable for Scott's Legacy: class_change (no prior races with a recognised class to compare) and going (no career runs on good or adjacent going to draw from). That last one is the honest risk I need to sit with — Newton Abbot on good ground today, and the horse has no form on the surface. We are backing a steamer with no going evidence, only speed and market signal. The class_change absence is less troubling given it is a novice hurdle, but the going gap is real. If the ground has changed since morning assessments, or if the horse simply does not handle good going, those two absent features represent a genuine blind spot in the case. Annie Edson Taylor is the closest challenger in composite terms and she does have soft-going evidence — if you need a fallback, that is where I would look. But I am on Scott's Legacy today.
Scott's Legacy · 3/1 · 1 unit · HIGH
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.
Value pick 0 of 5 won · -5 units at advised prices · since 11 October 2026. Kept separate from the bet’s record above.
