● The track record — open book
Top pick in the money 62% of the time. Our sharpest reads: 73%. The winner inside our three-horse shortlist nearly 2 of 3 races. Measured over roughly 4,900 settled starts — every number put alongside what you’d get picking blindly.
01The four numbers that matter
Every figure below is what happened in settled races. The grey bar shows the same outcome for a horse picked blindly — the floor a real read has to clear. These numbers clear it by a margin.
finishes 1st, 2nd or 3rd
6 of every 10 races, our top-rated horse finishes 1st, 2nd, or 3rd. That’s a pick you can count on to be in the race — not just near it.
Source: scorecard in-the-money 61.7% vs 37.8% random.
the actual winner was one of our three
The actual winner was one of TBredIQ’s three horses nearly 2 of 3 times. The shortlist is where the winner lives.
Source: scorecard 63.7% vs 39.9% random.
at least two of our three ran top 3
More than half the time, two of our three horses ran in the actual top 3. The shortlist finds the ones that matter — not just the winner.
Source: scorecard 56.6% vs 33.3% random.
our top-rated horse won the race
Our top pick wins roughly 1 in 3 starts. Picking blindly gets you 1 in 8. In our clearest reads (FOCUS), that win rate reaches 1 in 3 on the conservative measure — and higher in the sharpest tier. No pick wins every race; this one wins the right ones far more often.
Source: scorecard win 29.2% vs 12.7% random.
When a race calls for an exacta, our top two are the actual top two 13.5% of the time — versus 3.6% for a random pair: a 3.7× lift in combination accuracy. The engine builds the right pairing far more often than chance. (Pool takeout is real; hitting more often than chance doesn’t guarantee a return. We say so plainly below.)
Source: scorecard top-2 exacta-box hit 13.5% vs 3.61% random (3.7× lift).
02Confidence is the signal
TBredIQ’s confidence levels aren’t decorative. They track exactly how often the top pick wins. Four tiers, four real steps — the climb is clean and consistent across nearly 5,000 races.
In the sharpest spots of all: about 4 in 10. That’s our highest-confidence tier — and a number that’s about 3× what picking blindly returns in a typical field. This isn’t a claim about any single race; it’s the documented win rate across the full tier. The gradient is real. Use it.
13% · 24% · 33% win rate by confidence tier; sharpest spots ~40%. Compared to ~13% picking blindly across all races — about 3× at the peak. N=4,933 settled races.
02b● FOCUS reads — the engine’s clearest calls
FOCUS is the grade we give a race when the pace shape is defined, the form is readable, and our top pick stands apart from the field. These are the races worth your full attention — and the performance data show why.
our top pick finishes 1st, 2nd, or 3rd
vs ~23% in PASS reads — a 13-point lift
73% board (FOCUS) vs 48% board (PASS) — the grade separates the field
The grade earns its label: FOCUS outperforms PASS by 13 points on win rate and 25 points on board rate. Those gaps are real, measured, and the reason the grade exists.
Not every race earns a FOCUS. Most don’t. That’s the point. When the engine issues a FOCUS, it has a specific read: pace sets up for a particular horse, form is clean and recent, and the field hasn’t priced it correctly enough to close the angle. The 73% board rate is what that selectivity produces.
A clean PASS is as valuable as a sharp FOCUS. When a race doesn’t meet that bar, we call it PASS — plainly, with the reason. Knowing which race to skip is part of the read.
These numbers cover our highest-confidence setups — they don’t describe every race we read, and no grade guarantees a result. Source: race-grade actuals, FOCUS n=1,742 / PASS n=1,160 (board 73.2% vs 48.2%; win 35.7% vs 22.5%).
03What these numbers mean
Our top pick finishes 1st, 2nd, or 3rd — "in the money" — in about 6 out of every 10 races. A randomly chosen horse finishes in the money only about 4 times in 10. That gap is what the engine delivers: a top pick you can trust to be competitive, race after race.
In nearly 2 out of 3 races, the actual winner was one of TBredIQ's three highlighted horses. We're not just naming a single horse — we're building a shortlist where the winner usually lives. And in more than half of all races, at least 2 of those 3 horses finished in the actual top 3, which means the shortlist consistently identifies the horses that matter most.
The engine's top-rated horse wins the race in roughly 1 out of 3 starts. For comparison, if you picked randomly from an eight-horse field, you'd win about 1 in 8 times. TBredIQ finds the winner more than twice as often as chance — though it won't win every race, and no system does.
04Where the analysis is sharpest
This is where the engine reads a race most reliably — not a map of where to wager. The form lines are cleaner in these spots, so the read is sharper.
Sources: dirt sprint 6–6.5f in-the-money 66% / win 32%; small field (5–6) in-the-money 74% / 2-of-3 capture 80%; stakes & graded races show the biggest lift over picking blindly of any class; highest-confidence tier top-pick win ~40%.
05Built to read the whole card
TBredIQ works the full card the way a sharp handicapper does: it tells you which races are worth your time, which to skip, and what each race's shape calls for. The read is the product — you make the call.
And when a pool carries over — money left on the table from prior days — we flag the one spot where the math, not the handicapping, is on your side. See today's carryovers →
06What we don't claim
TBredIQ reads races — with documented accuracy — and stops there. These numbers describe field-reading quality. They are not a profit claim, not a market-beating guarantee, and not advice on what to wager or how much. No system wins every race. We’d rather show you what the data says than sell you a version of it that doesn’t hold.
So where is the value? Not in out-hitting the market — in the structural spots the math hasn't priced: carryover and forced-payout pools, where money left by past players is on the table for today's. That's the one edge we'll actually point you at.