The eight numbers

One mile. Eight numbers. One edge.

Every claim verified, every number footnoted with sample size and p-value. No hand-waving.

01 · The marquee
4.7×

Prize money per dollar

$100k the market's way → $52,200. Our way → $245,710. Within-band, tier-filtered (sub-$50k × top biomech).

n = 875
02 · The trust line
60–86%

Lift, in 72% of the market

Same-band coin-flip: top vs bottom biomech tercile. +86% in $0–50k, +71% in $100–200k. Survives a 5% outlier trim.

n = 5,528 · p < 0.0001
03 · Durability
+36%

More starts

15.1 vs 11.1 career starts at the same per-race ability. They race more often — and longer.

n = 8,827
04 · Longevity

Still racing at 30 starts

13.8% vs 4.6%. Triple the chance a horse is still on the track when the deep career money is paid.

n = 8,827
05 · Multi-winners
+8 pts

47% vs 39%

Eight more percentage points of multi-race winners. Not flukes — repeat winners.

n = 8,827
06 · Stability
1.4×

Across 7 vintages

PM-per-dollar lift holds steady across seven yearling vintages. Not a single-year anomaly.

n = 3,195 mature
07 · Rarity
1 in 13

Net +6 = top 7.5%

The signal is rare on purpose. Net +8 = top 2% / 1 in 48. Few horses clear it — and that's the point.

Rarity framing
08 · Both sexes
1.69× / 1.54×

Fillies / Colts

Works on both sides of the catalogue. Sex-consistent — not a fluke of one population.

n = 8,258
Where the model doesn't work
“Above $300k the model adds nothing — the market is too distorted at the prestige end. For the 72% of yearlings that sell below that ceiling — where 7 out of 10 bids land — we're 60–86% better than chance.”
The institutional trust line
Talk to us

Bring eight numbers to your next sale.

Briefings for bloodstock managers, syndicators and trainers. Eight Furlong is private and bespoke — we don't publish picks.