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 steam on a mares' hurdler at Downpatrick is the read I keep coming back to today — everything else in the top six has a flaw loud enough to hear, and this one does not. The engine scanned all 27 declared races and 246 runners for Saturday, applying a price filter of 2.0–50.0 decimal (no odds-on), a field-size gate of 6–14 runners, and a lay-of-day collision check. Weights used throughout: market_move 1.6, speed 1.5, form 1.2, class_change 1.0, trainer_strike 0.8, going 0.8, draw 0.5. Market movement carries the heaviest coefficient in this engine, and today it is doing the most work.
The pipeline opened with 246 runners. The odds band trimmed that to 225, the field-size gate brought it to 171, and the lay-of-day collision check dropped a further four runners to leave 167 eligible runners going into composite scoring. That is the pool from which the shortlist was drawn.
Five names came out of the score-off with composites above 6.6/10. Sun Goddess topped the speed column with an SR of 121 and a composite of 7.23/10, but she is priced at 2.2 decimal and the market read is actively negative — the engine puts her estimated win probability at 36.7% against a market-implied 40.6%, a -19.2% edge. That is a fade, not a follow. Sea Cookie at 7.19/10 has a hot trainer in Tom Clover (22% strike rate, 30 days) and a meaningful class drop, but the composite trails by nine points and the odds are a shorter 10/3. Tropical Twist and Pralognan share the Downpatrick 15:08 card and both steam, but Gordon Elliott is cold at 7% in the last 30 days and Pralognan has not placed once in three runs on good ground. Thesilverbridle carries a trainer strike rate of 2% — I do not back horses with a trainer in that kind of form unless the rest of the card is overwhelming. Gangster Granny leads on composite at 7.28/10, and the feature that tipped the balance is the market move, which registered the maximum available signal.
The number I had a long look at is the market edge. The engine estimates Gangster Granny's win probability at 36.9% against a market-implied 11.9% — that is a +177.1% edge, comfortably the widest positive gap on today's card. The market-move feature scored +2.00 — the maximum — meaning a strict majority of bookmakers shortened overnight and none drifted. At 13/2, that steam is telling a story, and this engine gives market movement the highest single weight at 1.6. I do not ignore that kind of consensus.
The supporting case holds up. The speed figure is SR 105 against a field mean of 97.5 (z = +0.80), which is a solid reading without being outlandish. The form feature scored +0.66 — the string '0-1521' with the latest win carrying a decay-weighted average of +1.12 — meaning the recent form is genuine and freshly earned, not a distant memory. Going is a real positive too: 3-from-11 with seven placed runs on good or adjacent ground, scoring +0.68. This horse has form on the surface she meets today.
Two features are unavailable: class_change is unrecognised for this race type, and trainer strike rate is below the noise threshold with only four runners in 30 days. I do not pretend those gaps are not there. But the features that do fire — speed, form, going, and the steaming market — fire together, and the composite of 7.28/10 is a strong reading by the engine's own scale. The market is doing something deliberate here, and I am following it.
The honest risks are these. Class is unscored — the engine cannot tell me whether this is a drop, a rise, or a lateral move, and in a mares' handicap hurdle that matters. Trainer data is below the noise threshold, so I have no handle on current stable form. The going is Good at Downpatrick, and while the form on that surface is encouraging at 3-from-11, it is not watertight. The runner-up on composite is Sun Goddess at 7.23/10, only five points back — if I am wrong about the market fade there, that is a horse with a significantly higher SR (121 vs 105) and cleaner recent form. And at 13/2, the price is generous for a reason: this is a provincial mares' hurdle, not a deep-data environment. The steam is the case. If that steam is noise, the rest of the evidence is solid but not commanding.
Gangster Granny · 13/2 · 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.
