No cherry-picked screenshots. We publish the full card, mark our pool against the actual winner, and show you the misses. Here is exactly how the model has done.
Straight talk: these are coverage and accuracy numbers, not a profit guarantee. The edge we sell is disciplined structure and total transparency — you always see the whole card and every result.
Our confidence tier is locked before post. It barely moves pool coverage (that's budget-sized ~87% at every tier) — what it tells you is how often the single stands up.
The fulcrum is the horse we tag as the field's fastest competitive early pace. Two honest questions: did it actually dictate the pace, and when the crowd bet it to ≤2/1, was our "don't single it" call right?
Our pool for every race, locked before post, graded against the chart. 🏆 is our top single; In pool means the winner landed inside the pool we published.
Separate from the coverage pool, we flag horses the market underrates. Recent overlays that hit.
We publish overlays win or lose and grade the whole set — the losers are in the record too.
Because a tout only shows winners. We run a real post-race debrief on every card — coverage, rank, and a named reason for each miss — and it all lives in the same database the picks come from. You are not buying a hot tip; you are getting a disciplined, self-correcting engine that grades its own homework in public.