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 Group 1 at the Curragh is pulling the money one way, and I'm not fighting it. Today I ran the engine across 25 declared races and 397 runners, applying a seven-feature weighted model with market movement sitting at the heaviest individual weight (1.6) and speed second at 1.5. The odds band filter stripped out anything odds-on or wider than 50.0 decimal, keeping runners priced between evens and 49/1. Field size was tightened to between 6 and 14 runners — anything smaller is a coin-flip, anything larger is a lottery. Two runners were removed by the lay-of-day collision check. That left me with a clean shortlist built on form, speed, class, trainer strike rate, going record, and market behaviour.
The day opened with 397 runners across 25 races. The odds-band filter brought that down to 271. Field-size constraints cut it further to 140. The lay-of-day collision check removed 2 more, leaving 138 runners in contention. From those, the engine surfaced six horses above the composite threshold worth looking at properly.
The top six by composite were Dr Rascal (7.27/10), Valedictory (6.94/10), Langstone (6.74/10), Gert Lush (6.72/10), Eutropia (6.70/10), and Lilt (6.67/10). The gap between first and second — 0.33 points — is solid rather than emphatic, so I had a long look at Valedictory before settling. What tipped it was the combination of a perfect recent form string, a superior SR in a Group 1 field, and a market steam that the engine read as genuine. Lilt posts the largest raw market edge at +104.1% but the composite doesn't support it to the same degree — the speed z-score is softer and the form string is thin. Valedictory is the most credible alternative and I want the reader to know that.
The read I keep coming back to is the form string: '111'. That is three wins in a row, and the decay-weighted model scores that at +3.00 on the form feature — the highest of any horse in the shortlist by some distance. The weight on form is 1.2, so it moves the composite meaningfully. Pair that with a speed figure of SR 127 against a field mean of 109.8 — a z-score of +1.14, carrying the highest weight in the model at 1.5 — and Dr Rascal is bringing elite-level raw ability into this race.
The market is the other signal I am not ignoring. The engine logged a strict steamer — a strict majority of bookmakers shortened, none drifted. At 10/3 in a Group 1 field, the model estimates a true win probability of 36.9% against a market-implied 20.6%, an edge of +59.9%. That is a substantial gap. Markets in Group 1s at the Curragh are not soft — informed money moves these prices. When the model and the market are pulling in the same direction, I take that seriously.
Going is a further tick: 3-from-3 with 3 placed runs on good or adjacent going. Whatever conditions the Curragh serves up today, Dr Rascal has proven he handles them. The composite lands at 7.27/10 — that is a strong reading by any measure in this model.
The one drag in the data is the trainer. Oliver Cole posts a 30-day strike rate of 7% — 2 from 27 — which the model penalises at -0.38 on the trainer feature. A cold trainer in a Group 1 is not something to wave away. I am also aware that the 0.33-point gap to Valedictory is solid but not a blowout — if something about Dr Rascal's preparation or the draw (draw data is unavailable in this version of the model) is wrong, the engine does not catch it. The form string is elite but the sample is short — three wins in a row tells you about a rising horse, not necessarily one ready to step up to Group 1 opposition at the top of the season. Going can shift too; if the ground moves softer overnight the going record becomes less informative.
Dr Rascal · 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.
