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 logbookWhen the speed figure, the form, and the money all point the same direction, the read writes itself — and today, they do. The engine scanned all 39 races declared on 14 October 2026, across 346 runners in total, running each through a four-stage filter pipeline: first a race-state check (no runners dropped there), then an odds band of 2.0 – 50.0 decimal to strip out odds-on favourites and outsiders beyond 49/1, then a field-size gate of 6 – 14 runners to cut out the noise of tiny fields and the chaos of big ones, and finally a lay-of-day collision screen that removed 3 runners flagged by the opposing book. Feature weights used throughout: market_move at 1.6, speed at 1.5, form at 1.2, class_change and going at 0.8 each, trainer_strike at 0.8, and draw reserved for v2.
The day opened with 346 runners across 39 races. The odds band trimmed that to 292; the field-size gate brought it down to 232; the lay-of-day collision screen cut a further three, leaving 229 runners to be scored. From those, six horses cleared the threshold worth discussing, and one sat clearly above the rest.
The top six by composite were Frisco Lad at 8.15/10, Centurion's Sister at 7.29/10, Uomo Valeroso at 7.27/10, Viroflay at 6.82/10, and Philanthropist at 6.63/10. Frisco Lad is 0.86 clear of Centurion's Sister — well above the 0.4 separation bar the engine requires to call a clear lead. Centurion's Sister and Uomo Valeroso sit virtually level and would be a coin flip without the gap to the top. What tipped the balance at the head of the list is Frisco Lad's combination of a speed z-score that puts it two standard deviations clear of its field and a form string that currently reads a single, clean '1' — the tightest, most unambiguous signal the form model can produce. On a separate note, Fanion d'Estruval at 11/1 is the longest price in the shortlist and carries the widest model edge; that one is published as the value pick today, not the selection.
The read I keep coming back to with Frisco Lad is the convergence of three independent signals, each of which would be worth noting on its own. Speed first: an SR of 121 against a field mean of 87.8 gives a z-score of +2.05 — weighted at 1.5 that is the strongest raw speed contribution on the entire shortlist. This is not a horse scraping past the field average; it is operating at a genuinely different level to what it faces at Punchestown today.
Form next. A recent form figure of '1' with a decay-weighted average of +2.00 and a top-line score of +3.00 after the 1.2 weight is applied — that is the cleanest form reading of any horse I looked at today. Last time out it won. That fact, adjusted for recency, is doing serious work in the composite.
Then the market, and this is where it locks in. Frisco Lad is a steamer — a strict majority of bookies shortened it, none drifted — contributing +2.00 to the most heavily weighted feature in the model at 1.6. The model's estimated win probability is 41.1% against a market-implied 31.1%, a raw edge of +18.2%. At 15/8 (decimal 2.875), that is a price the model considers genuinely wrong. The overall composite lands at 8.15/10, which I would call strong — the bar for that label is 8+/10.
I am not going to pretend the gaps in the data do not matter. Class change is unavailable because today's race class is unrecognised, going is unavailable because Frisco Lad has no career runs on good or adjacent going on record, and the draw layer is not yet built. That is three features returning no signal at all, which means the composite is built on fewer data points than I would like. The trainer line is also mildly negative: Henry de Bromhead's 30-day strike rate of 9% (10 from 116) is flagged as on par, contributing -0.28 — not alarming, but not a tailwind. At 15/8 the odds are inside the tighter end of the band, so if the model is even slightly overstating the edge, the margin for error is thin. And Centurion's Sister at 7.29/10 is a live alternative if you want something with going form confirmed and a class drop behind it — though it is a different race on a different track.
Frisco Lad · 15/8 · 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.
