A confidence tier on each race, the coverage pool sized to the market, a one-line read on every horse, and the validated overlays β from the WagerCast MCP. Graded honestly afterward on The Record.
The free preview shows the field, the morning line, and each connection's win rate in the race type. A pass opens the part that took the work:
Which horses to actually cover, sized to reach ~70/80/85% of the market's win probability β plus our single π top pick in each race.
The note behind each runner β why it's in the pool, why it's out, the trip or class angle the crowd is missing.
Value overlays (Debut-gate, Dirt-relief, Comp-exit), Banker win-rate reads, and Fade traps β each backtested, not vibes.
LOCK single-friendly, STANDARD, or CHAOS wide-open β so you know when to press and when to spread.
Where our number beats the market's β the horses flagged as genuine price value before the odds move.
Every analyzed card is scored against the official chart on The Record β hits, misses, and what we learned. No cherry-picking.
Not tips β a repeatable, backtested workflow that reads the whole card the same way every day and grades itself in public.
The full field's prior performance history β speed and pace figures, class, the workout tab, jockey/trainer form, and every trip line.
The WagerCast MCP reads the whole field at once: a coverage pool sized to the market's win probability, the pace and fulcrum picture, class moves, and the angle flags.
A full mechanical backtest β 500+ charted races, 3,700+ runners β showed no model out-ranks the odds. So the market sizes the pool, and handicapping is a small, high-confidence adjustment on top.
Every flag β turn-time lone-speed, declining-form fade, debut-gate, dirt-relief, company-class, trip-authority β is proven out with date-splits and same-priced-peer controls. The nulls get rejected.
Every read is written down pre-race and timestamped. What you see is exactly what was called before they ran β nothing edited after the result.
Each card is graded against the official chart β coverage hits, misses, and value-play returns. Every result, kept in public.
I'm an avid pace handicapper and a data analyst β and honestly, I'm in it for the puzzle.
The daily question of which horse the crowd has wrong, and why, is the most fun problem I know. WagerCast is the tool I built to answer it for myself: a way to turn a mountain of past-performance data into an honest, repeatable read on every race β and to prove out, or kill, each angle with a real backtest instead of a hunch.
I use it every day. I built it to help me; if it helps you too, that's the whole point. The blog is where I document the journey β the research that worked, the ideas that didn't, and what the data keeps teaching me. Follow along on X at @williammmink for the day's card reads and the occasional bad beat.
One-time access β no auto-renewal, no card kept on file. Secure checkout via Stripe.
Log in with Google or Discord, then pay via Stripe β we never see your card.
Already have access? Enter your access token
Sign in to buy your pass and unlock the full card.
One-time passes, no subscription. We only use your email to restore access.
Already purchased? Enter your access token