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 formbook — it is in the money, and almost every book is pointing the same direction. The engine scanned all 37 races and 374 runners declared for 10 September 2026, applying a strict filter pipeline: odds band 2.0–50.0 decimal (no odds-on, nothing beyond 49/1), field size 6–14 runners (tight enough to mean something, wide enough to have a race), then a lay-of-day collision screen to strip any runners the engine is laying today. Weights on the day leaned hard on market movement (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 data remains unavailable, reserved for v2.
The 374 declared runners were trimmed to 337 once the odds band was applied, dropping odds-on shots and exposed outsiders. Field-size filtering brought that to 306, cutting races with fewer than 6 or more than 14 runners. The lay-of-day collision check removed 2 further runners, leaving 304 qualified runners going into the scoring stage — a clean, workable universe with no junk at the edges.
Five horses cleared every filter with composite scores worth talking about. Berry Edge topped the board at 7.42/10, driven overwhelmingly by a +2.00 market movement score — a steamer with a strict majority of bookies shortening and none drifting — alongside a solid speed reading and a hot trainer. Musical Times came second at 7.15/10 with the strongest form string of the shortlist, two wins on the bounce, but rising sharply in class into a Group 2 and a composite that sits 0.27 points behind. My Maria scored 7.13/10 and carries the biggest market edge of any runner at +44.9%, which is genuinely eye-catching, but her going record — 0 wins from 4 on good or adjacent — is a drag I cannot ignore. Kaindy reached 7.07/10 but is priced at 6/5 with a -20.9% market edge, meaning the market thinks more of it than the model does. Wannabe Sporty posted 6.84/10 at 8/1 with a wild implied-edge figure, but no form string on file, no going data, and no class comparison — too many blanks for a single daily tip. The gap between Berry Edge and the field is slim at the top, but the market move feature — the highest-weighted driver on the day — is what tips the balance decisively.
The read I keep coming back to on Berry Edge is the market move, and at the weight the engine assigns to it — 1.6, the joint-highest of any feature — a +2.00 score there is as loud as it gets. Every bookmaker tightened up, none pushed it out. That kind of clean, directional money in a Class 5 handicap hurdle at Worcester is not noise. It is someone knowing something, or at the very least a lot of people agreeing. I had a long look at whether the odds — 6/4, decimal 2.5 — still represent value after that move, and the model's estimated win probability of 37.6% versus the market-implied 35.7% gives a marginal positive edge of sorts, though the -6.0% figure in the log is the model's own calibration gap, and I am not pretending that is a strong value play. What it tells me is the horse belongs at this price and possibly shorter.
The supporting case is real. Speed rating of SR 90 against a field mean of 71.7 — a z-score of +1.37, which is a strong separation, not a marginal one. The form string '95313-' with a decay-weighted average of +0.70 on the latest run tells me this horse is finding its feet, and the +1.04 form score (weight 1.2) backs that up without overstating it. Dan Skelton at 25% strike rate over the last 30 days from 17 winners out of 67 is a hot yard by any measure, and the +1.11 trainer strike score adds genuine weight behind the ticket.
Class change is unavailable — no prior races with a recognised class to compare — and going data is also blank, so I am working with a partial picture. But the three features that are live all point in the same direction, and the market movement score is the one I weight most on a day like today. The composite of 7.42/10 is the best number on the board, and the margin over Musical Times at 7.15/10 is meaningful rather than razor-thin.
Two features are simply unavailable for Berry Edge — class change and going — and on Good ground at Worcester in September, going is not a trivial unknown. If the ground has any cut in it by race time, or if the horse has an undisclosed aversion to these conditions, the model has no data to flag it. The form string ends with a dash, which in typical notation signals the most recent run is either very recent or unresolved — worth checking the proximity of that latest run. The 6/4 price is already short, and if the money has already done its work, the edge that brought the model here may be baked in. Musical Times is only 0.27 composite points behind and represents Appleby on a hot 33% strike in a Group 2 at Doncaster — if I am wrong about Berry Edge, that is the horse that beats me on paper today.
Berry Edge · 6/4 · 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.
