Stridometrics
Sign inTry it free

When we say 30%, it happens about 30% of the time.

We publish a win probability for every horse, then grade every prediction in public. This page is the running report card.

Historical calibration, 2023 to 2025

Three seasons of held-out data: more than 900,000 runners, graded against what actually happened.

0%0%20%20%40%40%60%60%80%80%actual %predicted %0% to 5%: predicted 2.9%, actual 2.4%, n = 237,9245% to 10%: predicted 7.3%, actual 7.2%, n = 225,53710% to 15%: predicted 12.3%, actual 12.6%, n = 150,91715% to 20%: predicted 17.3%, actual 17.8%, n = 98,93220% to 25%: predicted 22.3%, actual 23%, n = 65,94025% to 30%: predicted 27.3%, actual 27.8%, n = 44,11330% to 40%: predicted 34.4%, actual 34.7%, n = 49,44740% to 50%: predicted 44.3%, actual 44.4%, n = 22,02250% to 60%: predicted 54.3%, actual 54.8%, n = 9,46860% to 70%: predicted 64.2%, actual 61.8%, n = 3,93270% to 100%: predicted 76.5%, actual 76.3%, n = 1,866

121,714 races · 910,098 runners

Live calibration, accumulating since launch

Every prediction we've published since going live, graded in public as results come in. It grows every day.

0%0%20%20%40%40%60%60%actual %predicted %0% to 5%: predicted 2.8%, actual 3.2%, n = 9,7115% to 10%: predicted 7.3%, actual 7.8%, n = 8,39910% to 15%: predicted 12.3%, actual 12.9%, n = 5,88715% to 20%: predicted 17.3%, actual 18%, n = 4,08320% to 25%: predicted 22.3%, actual 22.3%, n = 2,79125% to 30%: predicted 27.4%, actual 25.2%, n = 1,88630% to 40%: predicted 34.3%, actual 34.1%, n = 2,11840% to 50%: predicted 44.3%, actual 41.3%, n = 1,04850% to 100%: predicted 59.9%, actual 53.9%, n = 770

5,237 races · 36,693 runners

The toteboard's implied win chances add up to ~123%. The house edge hides in every price. Ours add up to 100%, and they match reality bucket by bucket.

2023 to 2025
121,714 races · 910,098 runners
This chart starts empty on purpose. Every prediction we publish gets graded and added here, one day at a time.
5,237 races · 36,693 runners

The live chart is young. Early buckets ride on small samples and will swing; that is what honest data looks like at the start. Every day adds races, and the chart fills in as they grade.

Last 30 days

The same live record, restricted to recent races. Shorter windows swing more; the diagonal is still the target.

0%0%20%20%40%40%60%60%80%80%actual %predicted %0% to 5%: predicted 2.7%, actual 3.2%, n = 7,0735% to 10%: predicted 7.3%, actual 8%, n = 5,90110% to 15%: predicted 12.3%, actual 13%, n = 4,12015% to 20%: predicted 17.3%, actual 18.5%, n = 2,82820% to 25%: predicted 22.3%, actual 21.9%, n = 1,93925% to 30%: predicted 27.4%, actual 25.9%, n = 1,28230% to 40%: predicted 34.3%, actual 33.7%, n = 1,50640% to 50%: predicted 44.3%, actual 39.6%, n = 76250% to 100%: predicted 60.1%, actual 52.6%, n = 582

3,732 races · 25,993 runners

Last 90 days

The same live record, restricted to recent races. Shorter windows swing more; the diagonal is still the target.

0%0%20%20%40%40%60%60%actual %predicted %0% to 5%: predicted 2.8%, actual 3.2%, n = 9,7115% to 10%: predicted 7.3%, actual 7.8%, n = 8,39910% to 15%: predicted 12.3%, actual 12.9%, n = 5,88715% to 20%: predicted 17.3%, actual 18%, n = 4,08320% to 25%: predicted 22.3%, actual 22.3%, n = 2,79125% to 30%: predicted 27.4%, actual 25.2%, n = 1,88630% to 40%: predicted 34.3%, actual 34.1%, n = 2,11840% to 50%: predicted 44.3%, actual 41.3%, n = 1,04850% to 100%: predicted 59.9%, actual 53.9%, n = 770

5,237 races · 36,693 runners

Why not just bet the favorite?

Because the crowd is genuinely good at ranking horses. Across our 2023 to 2025 test data, the public favorite won about 39% of races, while our model's top-rated horse won about 36.3%. We publish that number because it is true, and because it explains what this model is for. The market concentrates its confidence on one horse; our model spreads probability honestly across the whole field. Both can be calibrated at the same time. Favorites win most often and still lose money over the long run, because the takeout is built into every price. Our job is not to out-pick the favorite. It is to price every horse honestly and to flag the spots where the market's price looks wrong. Those flags are the edge bets, and the archive grades every one of them in public.

Hit rate is the wrong scoreboard

Win percentage is the number every tout leads with, so here is ours, told straight. Our no-edge fallback picks hit at 54.2% over their first graded stretch and still lost 17.9% of flat-stake money, which is roughly the track takeout. Hitting often is easy if you take short prices; the payoffs are just too small to cover the house's cut. That is why we don't sell hit rates, why no-edge picks are labeled and kept out of the claims ledger, and why the numbers that matter here are calibration and the graded edge-bet record.

Ready to handicap with an edge?

Start free on a selection of races, then unlock every track.

Try it free