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 logbookEvery steamer on today's card had a reason, but only one combined a top-end speed rating with a bookmaker consensus that looked like genuine information rather than casual money. That is the read I keep coming back to with Sonic Angel. The engine scanned all declared runners today under the following filters: odds band 2.0–50.0 decimal (no odds-on, nothing beyond 49/1), and field size 6–14 runners only — tight enough to be competitive, loose enough to avoid the lottery of big-field handicaps. Feature weights placed market_move at the top of the hierarchy at 1.6, with speed at 1.5 and form at 1.2; class change, trainer strike, going, and draw weighted below those. The market-move weight is the sharpest tool in the box — when it fires alongside a speed z-score above 1.0, I listen hard.
The day opened with 49 races and 497 runners. The odds filter removed 59 runners priced either odds-on or beyond 49/1, trimming to 438. The field-size filter then cut anything running in races below six or above fourteen declared, bringing the active universe down to 252 runners — the pool the engine actually scored against.
Six horses reached the top of the scoring table. Sonic Angel leads on 7.38/10 — a clear step ahead of the rest. Into The Light came second on 6.85/10 with the most rounded profile in the shortlist: form, class drop, and a hot trainer all contributing. Thornaby Pearl sits third at 6.81/10, with a compelling +30.1% model edge and strong form figures, but Adrian Nicholls running at just 4% in the last 30 days is a flag I cannot ignore. Blue Victory and Lady Dora Mae both scored 6.78 and 6.76 respectively — neither had enough feature depth to push through. Sexy Chick at 6.67/10 was the closest thing to a wildcard, with an SR of 155 and a two-class drop, but a cold trainer and a winless record on good to firm kept her below the cut. What separated Sonic Angel from all of them was the combination of the maximum market_move score and a speed rating that stood well clear of the field mean — two of the three highest-weighted features both firing hard.
The read I keep coming back to is simple: Sonic Angel returned the highest composite in today's entire filtered universe at 7.38/10 — and the two features that matter most to the engine both scored at or near their ceiling. The market_move came in at +2.00 on a weight of 1.6 — a strict majority of bookies shortened, none drifted. That is as clean a steaming signal as the engine records. Alongside it, the speed score logged +1.36 on a weight of 1.5, with an SR of 154 against a field mean of 130.8 — a z-score of +1.36. Speed more than a standard deviation clear of the average field in a six-furlong Class 4 novice stakes at Newcastle AW is a legitimate edge. The trainer Simon and Ed Crisford is at 17% in the last 30 days — on par rather than hot, but steady — and that contributed a modest +0.14 to the composite. The overall picture is a horse the market is pricing tighter than its opening mark while the speed data supports that confidence independently.
The caveats are real and I will not paper over them. Form is unavailable — there is no form string on file, which almost certainly means this is a lightly-raced or first-time-out runner. Class history is also unavailable for the same reason. Going data is blank — no recorded runs on good to soft or adjacent ground. Three of seven features returned no signal. The composite score of 7.38/10 is built almost entirely on speed and market move, with trainer strike as a minor third leg. I am backing two features, not seven. That is the honest read.
What keeps me on board is precisely the market-move weight the engine assigns. When bookmakers shorten a horse with no public form record to lean on, that money is coming from somewhere private — work watchers, connections, stable sources. At 5/4 decimal 2.25, this is not a price that leaves room for error, but the model's estimated win probability is 37.4% against a market implied 39.7% — a small negative edge of -15.8% on raw figures. The engine still selected it because the composite score dominates the shortlist and both its top-two weighted features are maxed out. I accept the thin market edge; the informational signal is strong enough.
The model edge is negative at -15.8% — the market has this horse slightly shorter than the engine's own estimate, which means I am not buying value in the traditional sense; I am backing the signal strength of the composite. Three features — form, class change, and going — returned no data, so the 7.38/10 score is built on a narrow evidence base. No recorded run on good to soft means there is a genuine unknown about how Sonic Angel handles any ease in the ground. The field-size filter passed this race, but a novice stakes at Newcastle AW can attract unexposed horses with no public form whatsoever, and any one of them could be ahead of the market's knowledge. Into The Light at 6.85/10 is the runner-up I would watch — a more complete profile, a hot Appleby yard, and a class drop. If the race had been at a slightly longer trip I might have been tempted to flip the selection.
Sonic Angel · 5/4 · 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.
