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 logbookSaturday's market is telling a clear story, and the loudest voice in it belongs to a Class 3 handicapper at Newmarket — and I am listening. The engine scanned all 51 races declared today across British racing, covering 529 runners in total. From there, three filters trimmed the field: runners in abandoned or already-resulted races were removed first (no change at 529); then the odds band cut anything priced odds-on or beyond 49/1 decimal, leaving 464; finally, field-size limits of 6–14 runners stripped out the small noisy fields and the big chaotic ones, bringing the live universe down to 262 runners. Feature weights going into the composite are led by market_move at 1.6, speed at 1.5, and form at 1.2 — so a horse that scores on all three has a genuine structural advantage in the rankings.
The day opened with 51 races and 529 runners on the board. The race-state pass dropped nothing — all declared, none yet run. The odds-band filter (decimal 2.0–50.0) removed 65 runners, mostly odds-on favourites and long-priced outsiders, leaving 464. The field-size filter then removed runners from races with fewer than 6 or more than 14 declared, cutting a further 202, and the working universe settled at 262 runners across eligible races. That is the pool the composite scores were built from.
Five horses separated themselves at the top of the composite rankings. Sixpack led at 7.24/10, driven by a market_move score of +2.00, speed z-score of +1.79, and form of +1.75 — all three of the highest-weighted features firing together. Mwafaq came in second at 7.07/10 with similarly strong market and form signals, but a cold trainer at 4% over 30 days dragged the composite back. Theft and Romotoso — running in the same maiden at Doncaster — both posted 7.05/10 and 6.99/10 respectively, their scores almost entirely built on market move and a hot trainer, with form, going, and class_change all unavailable; that much missing data makes conviction hard. Romantic Symphony at 6.91/10 is classy and steamered, but a class rise of two ranks (delta -2) produced a class_change penalty of -1.20, and the market edge is actually negative at -16.5% — the model thinks she is marginally overbet. The feature that finally separated Sixpack from everything below it was having speed, form, and market all pointing the same direction with real data behind each one.
The read I keep coming back to is the combination of pace and money. Sixpack carries an SR of 98 against a field mean of 83.2, a z-score of +1.79 — that is the second-highest speed figure in today's entire shortlist, and it is the highest in a race with actual form data to back it up. The form string '313114' decays to a weighted average of +1.75, with the most recent run a win. That is not just good form — it is form that is improving on the last meaningful effort. The trainer James Owen has been running at 20% over the last 30 days from 26 winners in 128 runs, which the model correctly marks as hot. Going is good to firm at Newmarket and Sixpack has gone 2-from-5 on that ground with three placed efforts — a positive going score of +0.90 on a surface that suits.
But the feature that locks this in is the market_move. A score of +2.00 — the maximum the model can award — means a strict majority of bookmakers shortened, none drifted. That is not noise; that is coordinated money. The estimated win probability from the model sits at 36.8% against the market-implied 16.2% — an edge of +102.1%. At 9/2, you are getting paid more than double what the market thinks the chance is, yet the smart money has already moved. That combination — model edge plus money in — is the case in two lines.
The composite of 7.24/10 is a strong score by the criteria I use. Nothing in the shortlist significantly challenges it once you account for Mwafaq's cold trainer, the maiden debutants' data gaps, Romantic Symphony's negative edge, and Northcliff's weak form string at -0.26. Sixpack wins the score-off cleanly.
The honest risks here are worth naming. First, going: Newmarket's good-to-firm ground can ride very differently depending on where on the course they rail, and if there has been any overnight dew or change, that going call may soften slightly. Sixpack's going score is positive but based on only five career runs on this surface — not a huge sample. Second, class: the class_change score is flat at +0.00, meaning Sixpack is running at the same Class 3 level as last time out, which is fine but provides no tailwind from a drop in grade. Third, the draw feature is unavailable — the model has flagged this as reserved for v2, so any bias at Newmarket's July course over 1m4f is simply not priced in. Fourth, Mwafaq is a genuine rival at 7.07/10 and is also steamered; if it runs its race cleanly, the gap between the two is only 0.17 composite points — marginal rather than dominant. The engine conviction is set at MEDIUM, which I think is the right read: strong case, not a certainty.
Sixpack · 9/2 · 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.
