Building Breeders’ Cup Saturday Exactas: Where Probability Meets Coverage

Perhaps the most compelling result in our three-part Saturday study: a structured exacta ticket covered 12 of 45 races and returned +52.3% historically, while the 2025 card returned +166.5%. The sample is small, so ticket cost and selectivity still matter.

By Jeff Bessa · CHV Insights · Breeders’ Cup Saturday analysis · October 22, 2026

This is the third and final article in our Breeders’ Cup Saturday series. The win-wager study asked how modeled win probability and price worked together. The Daily Double and Pick 3 study asked how far to expand a multi-race ticket. Here we use the model to build an ordered, single-race exacta: who belongs on top, and which horses merit coverage underneath?

Across five Breeders’ Cup Saturdays, 2021–2025, this review covers 45 races. Every ticket uses $1 per unique ordered combination. ROI equals payouts collected minus the cost of all combinations, divided by that cost. A horse may qualify for both sides of a ticket, but it is never paired with itself. Overlaps and ties can change the number of distinct runners and combinations from race to race.

Start with early speed underneath

The clearest simple build was Top 2 W% + Top AML over Top 2 ES. It caught 7 of 45 exactas, a 15.6% hit rate, and returned +51.2% ROI on 175 $1 combinations. W% is CHV's modeled win probability; AML is the Adjusted Morning Line. The top side usually contains two or three distinct horses because the AML favorite can overlap a top-W% selection. The top two Early Speed (ES) horses supply the underneath side.

One possible ticket-shape explanation is that an early-speed horse can lose the lead yet still hold second. That is a plausible rationale, not a mechanism established by this study.

Adding a third top-W% horse raised the combination count without improving the hit count over the highlighted build. This is a historical cost comparison, not an instruction to play that same shape in every future race.

Other underneath signals

PFIG and ENY offered alternatives, but their stand-alone results were less consistent than ES. We hold the same top-side choices in the next two tables so the underneath factors can be compared directly.

Top 2 PFIG underneath

Top 2 ENY underneath

The practical reason to keep PFIG in view is complementary coverage, especially for international runners from Japan or Europe who may lack an ES score. The data here do not show that PFIG alone was a stronger replacement for ES.

Widen the underneath pool—deliberately

With Top W% + Top AML on top, combining first-ranked horses across underneath factors produced the strongest returns per dollar in this comparison. Adding a second ES horse increased coverage but reduced ROI. These are overlapping ticket variants, so their hits and returns cannot be added together.

The three-factor pools illustrate the tradeoff: the tighter version returned more per dollar, while the broader ES version caught one additional exacta. For the main five-year test, we used Top 2 W% + Top AML over Top 2 ES + Top 1 PFIG + Top 1 ENY. Widening the top side raised coverage from 9 to 12 hits, while ROI fell from +78.9% to +52.3%. That broader ticket is the featured build below—not an optimized rule established for future cards.

Distance and surface

Three categories were profitable in this sample, with turf routes strongest. Turf sprints are a caution: the win-wager study's top-W% selection did not win a turf sprint, and this broader exacta build caught only one of five while losing money. Five races are far too few to settle a category's true profitability.

Applying the rule to 2025

The supplied 2025 chart gives each race's top and underneath pools, actual result, $1 payoff, and whether the ticket covered it. The same combined rule hit three of nine exactas, requiring 49 $1 combinations and returning $130.60 on $49 staked, for +166.5% ROI.

That year is a useful illustration, but it is part of the 45-race review, not an independent holdout. The $79.60 R9 payoff accounts for more than half of the 2025 return. The ticket missed six of nine exactas, including the $141.40 R5 result. Those facts belong alongside the headline ROI.

A promising test, not a promise

This may be the most compelling article in the series because the featured exacta build paired a 26.7% historical hit rate with +52.3% ROI. The comparison is directional: win bets, exactas, Doubles, and Pick 3s have different costs and payoff distributions. The exacta edge could shrink or disappear in a new sample. With only 45 races, a few payouts can move the return sharply, and many of the variants above were examined on the same data.

Treat this ticket as a framework for organizing contenders, not a blind bet. Confirm the final field and scratches, inspect the actual prices and pool, count every unique combination, and decide whether the cost fits the race. Passing is a valid choice. Wager only what you can afford to lose.

Read the win-wager study and the Daily Double and Pick 3 study for the first two parts of the series. The CHV Glossary explains the factors, and the CHV handicapping application lets you evaluate the current card and price your own tickets.

All returns above are exploratory historical results. They are not guaranteed betting advice or a prediction of future performance.