This worked example shows how to use BreedSmart to compare yearlings, identify potential opportunities before a sale, and set bidding limits against the price you would actually pay. For the reasoning behind probability estimates and uncertainty, read What Is BreedSmart?.
Buying a yearling at Keeneland September
Suppose your goal is to buy a yearling with a chance of achieving G1 success as a three-year-old. How would you compare an expensive shortlist when both the racing prospects and the prices differ? This example uses the ten yearlings selected in a completed BreedSmart comparison from the Keeneland September 2026 sale. The results below reproduce that saved report, rather than a new calculation.
The inputs
The racing goal is 3yo G1, with Male (M) selected. Every yearling is compared on that same male-specific basis. The hip identifies the catalog entry, while the pedigree shows the sire, dam, and broodmare sire. A different goal or sex selection would require a new comparison.
Campaign cost is $100,000 per yearling. This covers costs from purchase through the end of the three-year-old campaign, including breaking, training, veterinary care, and other expenses. It is additional to the purchase price, and is applied equally to all ten horses. All amounts in this example are USD.
Market assessment is 9 or 10. Into Mischief × Grazie Mille and Nyquist × Hot Cha have a score of 10; the other eight, including Nyquist × Factorofwon, have a score of 9. This is an input expressing market appeal beyond pedigree on a scale from 0 to 10. These high scores reflect the assumptions chosen for this example, not a claim that BreedSmart inspected the horses. The assessment affects both the estimated sale price and the estimated probability of racing success.
The report notes that the market-assessment adjustment was developed for the original three-year-old G1 model and is also applied here. Its use for this male-only goal has not been independently calibrated, so these results depend on that additional assumption.
The comparison
Click any column heading to sort; click it again to reverse the order. On a narrow screen, scroll sideways to see all the price columns. Report rank preserves the original ranking by fair price minus estimated sale price.
| 1 | 0181 | Not This Time × Danzatrice by Dunkirk | 9 | 4.093% | $2,486,834 | $2,722,001 | $3,004,201 | $3,349,113 | $3,780,252 | $1,537,445 | +$1,466,756 | $2,100,000 | +$904,201 |
| 2 | 0899 | Not This Time × Believe You Can by Proud Citizen | 9 | 3.740% | $2,263,534 | $2,478,401 | $2,736,241 | $3,051,379 | $3,445,302 | $1,585,806 | +$1,150,435 | $2,000,000 | +$736,241 |
| 3 | 0313 | Gun Runner × Luna Rosa by Malibu Moon | 9 | 4.563% | $2,784,016 | $3,046,199 | $3,360,819 | $3,745,354 | $4,226,023 | $2,343,314 | +$1,017,505 | $3,100,000 | +$260,819 |
| 4 | 0803 | Not This Time × Spell On You by Curlin | 9 | 3.443% | $2,075,730 | $2,273,523 | $2,510,876 | $2,800,973 | $3,163,595 | $1,802,632 | +$708,244 | $2,000,000 | +$510,876 |
| 5 | 0116 | Gun Runner × Amazement by Bernardini | 9 | 2.115% | $1,236,852 | $1,358,384 | $1,504,223 | $1,682,470 | $1,905,278 | $1,516,373 | -$12,150 | $2,300,000 | -$795,777 |
| 6 | 0084 | Gun Runner × Twenty Carat by Into Mischief | 9 | 4.587% | $2,798,728 | $3,062,249 | $3,378,474 | $3,764,971 | $4,248,092 | $3,524,166 | -$145,692 | $2,800,000 | +$578,474 |
| 7 | 0973 | Nyquist × Factorofwon by The Factor | 9 | 1.122% | $609,135 | $673,602 | $750,962 | $845,514 | $963,703 | $1,110,063 | -$359,101 | $2,000,000 | -$1,249,038 |
| 8 | 0697 | Into Mischief × Miss Costa Rica by Hit It a Bomb | 9 | 1.593% | $906,975 | $998,518 | $1,108,370 | $1,242,633 | $1,410,463 | $1,606,918 | -$498,548 | $2,000,000 | -$891,630 |
| 9 | 0642 | Nyquist × Hot Cha by Bernardini | 10 | 0.799% | $404,820 | $450,713 | $505,784 | $573,093 | $657,230 | $1,243,587 | -$737,803 | $2,250,000 | -$1,744,216 |
| 10 | 0242 | Into Mischief × Grazie Mille by Bernardini | 10 | 2.548% | $1,510,278 | $1,656,667 | $1,832,334 | $2,047,037 | $2,315,417 | $4,421,979 | -$2,589,646 | $3,700,000 | -$1,867,666 |
Sorted by original report rank, ascending.
What the outputs mean
Success probability is the estimated chance that the purchased yearling achieves the selected racing goal, including the chance of becoming a rated runner. The report displays a mean estimate for each horse. It does not display the full uncertainty distribution, so a small difference between two percentages should not be treated as a certain difference in ability.
Estimated sale price is what the model estimates the market may pay given the pedigree and the chosen market assessment. These figures are not the actual auction prices. Fair price answers a different question: what purchase price would give this horse the same estimated success per dollar as the group as a whole, after adding the campaign cost?
The benchmark pools the estimated success probabilities and the estimated purchase and campaign costs across these ten yearlings. It is specific to this group. Adding or removing a horse, changing a market assessment, or changing the campaign cost can change the fair prices. A fair price is therefore a relative comparison point, not an independent appraisal or a prediction of financial break-even.
The other four price levels show how value changes around that benchmark. Great price provides 20% more estimated success per dollar than the group benchmark, and Good price provides 10% more. High price provides 10% less, while Too expensive provides 20% less. These percentages refer to success per dollar after campaign costs; they are not simple percentage discounts or markups on the fair purchase price.
Difference is fair price minus estimated sale price. It is a reference for spotting potential opportunities before bidding: a positive number suggests the horse might be available below the group-relative fair price, while a negative number suggests it might cost more. The original ranking uses this estimated dollar gap to guide further attention. It is not a purchase recommendation. When deciding whether to bid or buy, compare the actual price you would have to pay with your chosen price limit.
Actual sale price shows the reported sold price in the sale results supplied for this example. Fair minus actual compares the report’s fair-price limit with that actual price. A positive amount means the horse sold below that limit. This is the relevant comparison for assessing the purchase at the reported price. The original Difference column and report rank compare fair price with the model’s estimated sale price only to help identify opportunities in advance. Before the sale, the estimate helps you plan; in the ring, the actual asking price or bid must guide the decision.
Reading the example as a buyer
Hip 0181, Not This Time × Danzatrice, leads the male-specific report. Its estimated success probability is 4.093%, its estimated sale price is $1,537,445, and its fair price within this group is $3,004,201. The report therefore shows a positive difference of $1,466,756. The estimated-price gap would flag it for attention before the sale. The buying decision, however, depends on whether the actual price falls within the buyer’s chosen limit.
Sorting by success probability tells a different story. Hip 0084, Gun Runner × Twenty Carat, has the highest displayed estimate at 4.587%. But its estimated sale price is $3,524,166, above its group-relative fair price of $3,378,474. Its difference is negative $145,692, placing it sixth in the original ranking. The horse with the highest estimated chance of success does not automatically offer the strongest value comparison. This negative forecast gap does not rule it out: as the actual-price comparison below shows, it sold within its fair-price limit.
Hip 0242, Into Mischief × Grazie Mille, illustrates the other side. Its estimated success probability is 2.548%, but its estimated sale price of $4,421,979 is substantially above its group-relative fair price of $1,832,334. The report shows a negative difference of $2,589,646. This does not mean the horse cannot become exceptional; it means that the predicted market price would make it a lower priority for advance investigation under these assumptions. Whether to buy still depends on the actual price: its reported $3,700,000 sale price was above the $1,832,334 fair-price limit.
These results help frame a bidding decision. You still need to decide whether the market assessments and campaign budget fit your own knowledge, and compare the report with the price you would actually have to pay. The calculation compares racing success per dollar; it does not value every possible source of earnings, future breeding value, or personal appeal. This is a retrospective illustration of a saved comparison, not evidence that the model predicted auction prices before the sale or that these horses will achieve their racing goals.
Which horses would a buyer following the report pursue?
A buyer using the report’s Fair price as a maximum purchase price would find five of these ten yearlings within that limit at their reported sale prices. This assumes the same 3yo G1 male goal, market assessments, and $100,000 campaign budget used in the report.
- Not This Time × Danzatrice, hip 0181: sold for $2,100,000 against a fair-price limit of $3,004,201, leaving $904,201 below the limit. The sale price was also below the report’s Great price.
- Not This Time × Believe You Can, hip 0899: sold for $2,000,000 against a fair-price limit of $2,736,241, leaving $736,241 below the limit. This was also below the Great price.
- Gun Runner × Luna Rosa, hip 0313: sold for $3,100,000 against a fair-price limit of $3,360,819, leaving $260,819 below the limit. Its price fell between the Good and Fair levels.
- Not This Time × Spell On You, hip 0803: sold for $2,000,000 against a fair-price limit of $2,510,876, leaving $510,876 below the limit. This was also below the Great price.
- Gun Runner × Twenty Carat, hip 0084: sold for $2,800,000 against a fair-price limit of $3,378,474, leaving $578,474 below the limit. Its price was just above the Great level of $2,798,728 and below the Good level of $3,062,249.
A more conservative buyer insisting on the Great price or less would narrow that list to Danzatrice, Believe You Can, and Spell On You. The other five yearlings—Amazement, Factorofwon, Miss Costa Rica, Hot Cha, and Grazie Mille—sold above their Fair limits and, in this comparison, above even the Too expensive levels.
Twenty Carat makes the distinction between a forecast and a bidding decision especially clear. The report’s estimated sale price of $3,524,166 was above its fair-price limit, producing a negative original Difference. But its actual sale price was $2,800,000, below that limit. A buyer should compare the price being asked in the ring with their bidding limit, rather than reject a horse solely because the estimated market price looked too high.
These are candidates a buyer following those limits would pursue, rather than purchases BreedSmart would make automatically. Veterinary findings, inspection, budget, and the ability to secure the horse within the limit still affect the decision. The comparison holds the saved report’s benchmark fixed; it does not recalculate fair prices using the actual sale prices. It also does not establish whether a bid at the reported price would have secured the horse, or whether any purchase will prove successful.
Source: the completed BreedSmart report supplied for this example and the Keeneland September 2026 catalog. Actual sold prices are transcribed from the sale-results screenshot supplied for this example. The table retains the report’s displayed rounding, so its original differences can vary by one dollar from subtraction of the rounded prices; Fair minus actual uses the displayed fair price.