Redefining Select: AI Physical Evaluation of the Fasig-Tipton July Sale
The Fasig-Tipton July Sale is often described as selective, but a review of the 2024 and 2025 July Sale samples through the Horse Sense AI Physical Evaluation™ model suggested something more precise: July does not simply reward elite physicals. July rewards alignment.
In this Horse Sense Pedigree Analytics™ study, AI physical scores, Black-Type Score™, and Sale Return Against Stud Fee™ are used together to evaluate what “select” means in the July yearling market. The data suggest that July rewards yearlings that are forward, commercially understandable, and well-timed around sire demand, rather than simply rewarding the most elite physicals in the crop.
Why Analyze the Fasig-Tipton July Sale With AI Physical Evaluation?
After completing the AI physical-evaluation model for Horse Sense Pedigree Analytics™, I wanted to test it against a real commercial marketplace. With the 2026 July Sale just around the corner, the Fasig-Tipton July Sale was the logical place to start.
July is early. The horses are still developing. The catalog is smaller than Keeneland September. The sale is often discussed as a select yearling sale, and I expected the physical scores to reflect that. I expected to see a deeper group of high-end, “select” physicals.
That is not what the data showed.
The horses were not bad. Many were useful. Many were forward. Many made commercial sense for July. But they were not, as a group, the type of elite physicals I expected to see when I thought about the word “select.”
That forced me to step back and ask a different question.
Maybe July is not selecting for elite physical quality in the same way Fasig-Tipton Saratoga or the top books of Keeneland September might. Maybe July is selecting for something more specific: yearlings that are forward, commercial, understandable, and ready early.
That distinction matters for breeders, consignors, and buyers.
What Does “Forward” Mean in a July Yearling?
In this context, a forward yearling is not necessarily the best long-term horse in the crop. It is a horse that appears more physically developed and sale-ready at an earlier point in the year.
A forward July horse usually has enough body, balance, maturity, muscle tone, and overall presence to look commercially prepared in early summer, before many later-developing yearlings have fully caught up. Forward does not mean perfect. It means the horse is mature enough, organized enough, and understandable enough for buyers to evaluate confidently at an early sale.
That distinction matters because a forward horse can be very effective in July without being a true premium physical.
How the Horse Sense AI Physical Evaluation Works
The Horse Sense AI Physical Evaluation™ model was developed as part of the Horse Sense Pedigree Analytics evaluation ecosystem to create a more consistent way to compare yearling conformation across a group of horses.
The goal is not to replace horsemanship. It is not to say a computer can look at a horse better than an experienced horseman. The goal is to reduce the amount of personal bias that enters the process by applying the same scoring structure to the same physical traits every time.
The model uses our proprietary 10-point scale and evaluates conformation through measurable structural relationships. The score is based on visible physical traits from the available photo, including:
- Shoulder angle. Length, slope, and functional relationship to stride mechanics.
- Pastern angle. Elasticity, balance, and shock absorption.
- Hoof-pastern alignment. Whether the lower limb presents as coordinated or mismatched.
- Shoulder-pastern match. Whether the front-end angles work together.
- Hip structure. Length, depth, and strength through the hindquarter.
- Femur length. Power and leverage through the rear assembly.
- Hock placement. Function, support, and hind-leg organization.
- Stifle strength. Athletic power and depth through the hind end.
- Overall balance. How the horse fits together as a complete individual.
- Commercial impression. Whether the horse presents as a saleable, buyer-friendly yearling.
The system is built around the same type of visual evaluation horsemen use every day, but it forces the evaluation into a repeatable framework. The same traits are scored in the same order. The same limitations are noted. The same scale is applied across the sample.
That is important because yearling evaluation is naturally subjective. One person may like length. Another may forgive pasterns. Another may value depth and hip more than shoulder. The AI model does not eliminate judgment, but it creates a more disciplined baseline. That fits the broader move toward structured pedigree and market analysis.
There is also an important limitation. These scores are based on single side conformation photos. They do not include front view, rear view, walking video, veterinary reports, scope, radiographs, or trainer inspection. The model scores the visible commercial physical from the available photo. It does not claim to predict soundness, racing class, or long-term development by itself.
The July-Specific Physical Scale
For this study, the physical scale was interpreted within the context of the Fasig-Tipton July Sale. That matters.
An 8.0 threshold in this July study should not be treated as a universal standard for Saratoga, Keeneland September Book 1, or every yearling sale. Different sales have different depth, timing, and comparison groups.
For the July samples, the revised physical interpretation was:
| Physical Score | July Study Interpretation |
|---|---|
| 8.0+ | True select / premium July physical |
| 7.8–7.9 | Near-select / premium commercial physical |
| 7.3–7.7 | Good-to-strong July commercial physical |
| 7.0–7.2 | Forward but ordinary / needs sire or page help |
| 6.7–6.9 | Below preferred commercial standard |
| 6.6 or lower | Discount physical |
This is where the first major finding appeared.
A horse can be useful at 7.0 to 7.4. It can be forward. It can be commercially acceptable. It can sell well if the sire, page, and market timing are right.
But in this July framework, that horse is not a true select physical.
That is the distinction the data kept reinforcing.
Why the Word “Select” Can Be Misleading
When people hear “select,” they often think elite. They think a sale full of top physicals. They think a highly filtered group of yearlings that are better than the general population.
But the July data suggested something more nuanced.
July may be selecting for early readiness more than elite physical superiority. The sale rewards yearlings that are far enough along physically to present well in early summer. It rewards horses that are easy for buyers to understand quickly. It rewards commercial sire momentum. It rewards enough catalog-page support to make the buyer comfortable. It rewards an economic structure that gives the buyer or breeder a believable commercial outcome.
That is not the same thing as saying every good July horse is an elite physical.
The better July question may not be:
Is this one of the best physicals in the crop?
The better question may be:
Is this horse forward enough, commercial enough, and timed well enough to stand out early?
The 2024 and 2025 July Market Backdrop
Before looking at the scored horses, it is important to understand the overall sale environment. The 2024 and 2025 Fasig-Tipton July markets were very stable.
| Year | Offered Excluding Outs | Sold | RNA | Sold Rate | Average | Median |
|---|---|---|---|---|---|---|
| 2024 July | 231 | 156 | 75 | 67.5% | $111,288 | $90,000 |
| 2025 July | 218 | 159 | 59 | 72.9% | $106,717 | $90,000 |
The median was the same in both years: $90,000. The average changed only modestly, moving from $111,288 in 2024 to $106,717 in 2025. The number sold was also very similar.
That matters because the market did not radically change from one year to the next. The 2024 sale provides a useful comparison point, and the 2025 sale provides the current-market review. Together, they create a better baseline than looking at one year in isolation.
What the 2025 Physical Sample Showed
The 2025 photo-scored sample became the clearest test of the physical model because it included a broader working set of scored horses and a stronger statistical read.
Using the physical framework, price separated meaningfully by physical band.
| Revised Physical Band | Hips | Average Price | Median Price |
|---|---|---|---|
| 7.8+ | 4 | $283,750 | $267,500 |
| 7.3–7.7 | 20 | $166,550 | $155,000 |
| 7.0–7.2 | 15 | $111,200 | $100,000 |
| 6.7–6.9 | 15 | $95,600 | $95,000 |
| 6.6 or lower | 8 | $37,125 | $23,500 |
That is a clean and useful pattern.
The higher physical bands brought stronger prices. The 7.8+ group was the best-performing group. The 7.3–7.7 group also performed well. The 7.0–7.2 group was commercially usable, but clearly below the stronger physicals. The 6.6-or-lower group was a clear discount zone.
This supports the value of the physical model. Physical score mattered.
But the second finding is just as important: there were not many true premium physicals. In the 2025 sample, only four horses landed in the 7.8+ range. That does not describe a sale overflowing with elite physicals. It describes a sale where a small number of high-end physicals separated themselves, while many others were forward, usable, or ordinary commercial types.
The data was not saying July horses lacked value. It was saying July value came from more than elite physical quality.
What the 2024 Sample Added
The 2024 targeted photo sample supported the same broader interpretation.
In the 2024 sample, 47 horses were scored. The average revised physical score was approximately 6.96, and the median was approximately 6.9. None of the scored horses reached the 8.0 threshold for a true select July physical.
The 2024 sample had:
- 0 horses at 7.8+;
- 15 horses in the 7.3–7.7 good-to-strong commercial range;
- 5 horses in the 7.0–7.2 forward-but-ordinary range;
- 10 horses in the 6.7–6.9 below-preferred range;
- 17 horses at 6.6 or lower.
Again, the conclusion was not that the sale lacked useful horses. The conclusion was that the physical profile looked more like a sale of forward, commercially timed yearlings than a sale dominated by elite physicals.
The 2024 data also showed why physical score cannot be read alone. Several high-priced horses had ordinary or below-preferred physical scores but still sold well because sire demand, catalog page, or market timing carried them.
That is where the next two layers of the model became important: Black-Type Score and Sale Return Against Stud Fee.
What the Black-Type Score Measures
The Black-Type Score™ is separate from the physical score.
The physical score evaluates the horse’s body. The Black-Type Score evaluates the catalog page.
It is designed to measure the commercial strength of the female family by looking at the quality, closeness, recency, and repeatability of stakes production. The model gives weight to:
- first dam production;
- the first dam’s own race quality;
- second dam production;
- the second dam’s own race quality;
- graded stakes quality;
- closeness of black type;
- recency of production;
- repeatability through the female family.
This distinction is important because not all black type is equal.
A Grade 1 horse deep in the third or fourth dam can help a page, but it does not carry the same commercial force as a current stakes producer under the first dam. A family with one older outlier is different from a family that repeatedly produces stakes horses across multiple generations. A page with current graded activity is different from a page whose strongest runners are distant or dated.
The Black-Type Score tries to separate surface-level black type from commercially meaningful black type. This is closely related to the broader idea of Female Line Signal™, where repeatable production patterns can influence buyer confidence.
In the 2025 full catalog review, broad black type was too common to be highly predictive by itself. The Black-Type Score was more useful than a simple yes-or-no black-type label, but it still did not explain sale price as cleanly as physical score.
That was one of the more important findings.
Black type gave buyers confidence. It supported value. But by itself, it did not determine price.
Sale Return Against Stud Fee Explained in Plain Terms
The third layer is Sale Return Against Stud Fee™.
Sale Return Against Stud Fee compares the yearling’s sale price with the advertised stud fee used to breed the foal.
The formula is simple:
Sale price divided by stud fee equals the sale return against stud fee.
For example:
- A yearling by a $20,000 stallion that sells for $100,000 produced a 5-to-1 sale return against stud fee.
- A yearling by a $75,000 stallion that sells for $300,000 produced a 4-to-1 sale return against stud fee.
- A yearling by a $12,500 stallion that sells for $250,000 produced a 20-to-1 sale return against stud fee.
This is why raw sale price can be misleading.
A $400,000 yearling sounds better than a $250,000 yearling. But if the $400,000 horse is by a $100,000 sire, that is a 4-to-1 sale return against stud fee. If the $250,000 horse is by a $12,500 sire, that is a 20-to-1 return. The second horse produced a much more efficient commercial result relative to the stud fee.
But that still does not mean the second horse was automatically the better full economic result.
Sale Return Against Stud Fee measures Stud-Fee Efficiency™, not total profit.
In commercial breeding, a broad expectation is often that a yearling should bring roughly three to five times the stud fee to be considered workable. That does not mean 3-to-1 to 5-to-1 is pure profit. It is only a rough benchmark.
The breeder still has to account for mare care, foaling, board, veterinary work, farrier work, registration, sales prep, entry fees, commissions, insurance, transportation, and the risk of an RNA. A 3-to-1 return against stud fee may look acceptable on paper but still be thin once real costs are included. A 5-to-1 return usually gives the breeder more room, but the starting stud fee still matters.
That is why sale price, stud fee, and total cost basis have to be read together, especially when deciding whether a mating is being planned for racing, commercial resale, or both.
The Three-Layer View
The strongest way to understand July is to combine the three layers:
| Layer | Question It Answers |
|---|---|
| Physical Score | Is the horse forward, balanced, and commercially believable? |
| Black-Type Score | Does the catalog page give buyers enough confidence? |
| Sale Return Against Stud Fee | Did the sale price make sense relative to the stud fee? |
When those three layers line up, July can work very well.
The most attractive July horse is not always the horse with the highest raw sale price. It is the horse whose physical, page, sire demand, and economics are aligned.
That is the “perfect storm” July seems to reward.
Why Some Ordinary Physicals Still Sold Well
One of the more interesting findings was that some ordinary or below-preferred physicals still sold well.
That does not automatically mean the model missed them. It may mean the rest of the commercial stack carried them.
In July, sire momentum can be powerful. Freshman sires and young sires create urgency because buyers are still trying to figure out which sire crops they want exposure to. If a sire is hot, buyers may stretch further, even when the individual is not an elite physical.
The same can happen with page support. A horse with an ordinary physical may still sell well if the catalog page gives buyers confidence. If the female family is active, current, and commercially recognizable, buyers may be more willing to forgive a merely adequate body.
Stud-Fee Efficiency also matters. A $150,000 sale by a $10,000 or $15,000 sire can look much stronger on paper than a $250,000 sale by a $100,000 sire. But that does not mean six-figure sales from low-to-mid stud fees are routine. When they happen, they are strong commercial results. In many cases, they are the kind of outcome breeders are trying to create when they use a young or more affordable stallion before the market fully prices him.
That is why ordinary physicals can sometimes produce strong July results.
The horse was not necessarily elite. The timing was strong.
Why Good Physicals Can Still Underperform
The reverse is also true.
A good physical does not guarantee a strong commercial result.
If the sire is cold, the page is weak, the stud fee is too high, or buyers find something they do not like in the veterinary information, the price can still disappoint. This is especially important when a horse has a good physical and a good catalog page but sells far below expectation. In those cases, the model flags the result as a possible hidden or non-public issue.
That does not mean a veterinary problem is confirmed. Veterinary information is technically public to those attending the sale or with access to the sale repository, but in practice it is impossible for most buyers or analysts to review every horse in detail. In these situations, it simply means the publicly available information that was incorporated into this analysis does not fully explain the discount.
This matters because sales results alone can distort analysis. A horse may look like the market rejected the physical, when the real reason could be scope, radiographs, vet notes, reserve strategy, buyer perception, or another non-public factor.
That uncertainty is part of the broader market variance that has to be accounted for in any bloodstock analysis.
Why Sale Return Against Stud Fee Needs Context
I also looked at each horse’s Sale Return Against Stud Fee. This compares the yearling’s sale price with the advertised stud fee used to breed the foal.
For example, if a yearling sells for $100,000 and the stud fee was $20,000, the sale produced a 5-to-1 return against the stud fee. If a yearling sells for $300,000 and the stud fee was $75,000, the sale produced a 4-to-1 return against the stud fee.
That number is useful, but it is not the same as profit. It measures Stud-Fee Efficiency, not full economic return.
This distinction is important because the same return ratio can mean very different things depending on the starting stud fee.
A 5-to-1 return on a $10,000 stud fee produces a $50,000 sale. On paper, that meets the common industry expectation that a yearling should bring roughly three to five times the stud fee. But in real dollars, a $50,000 sale may not be enough to cover the full cost of producing and selling the horse. The breeder still has mare care, foaling, board, vet work, farrier, sales prep, entry fees, commissions, insurance, transportation, and the risk of an RNA.
So while a 5-to-1 return on a $10,000 fee is efficient against the stud fee, it may still be thin as a full commercial result.
The math changes at the upper end of the stud-fee market. A yearling by a $100,000 sire does not always need to bring three to five times the stud fee to be a reasonable result. A 1.5-to-1 or 2-to-1 return on a $100,000-plus stud fee can still produce a meaningful gross sale price, depending on the breeder’s full cost basis, the mare’s value, and the long-term plan for the family.
That is why sale price and stud fee have to be read together.
A $120,000 yearling by a $15,000 sire produced an 8-to-1 sale return against stud fee. That is highly efficient on paper and may represent an excellent commercial result. But it is still a $120,000 sale.
A $250,000 yearling by a $100,000 sire produced only a 2.5-to-1 return against stud fee. The ratio is much lower, yet the gross sale price is substantially higher.
This is where context matters. The best commercial outcome is not always the highest return ratio, nor is it always the highest sale price. It is the result that produces enough gross dollars, enough margin, and enough risk-adjusted return to justify the mating.
This is especially important with young sires. A lower stud fee can create opportunity when the market has not yet fully priced the stallion correctly. If a sire is standing for $10,000 to $20,000 and buyers decide the foals are better than the fee, the breeder can get a strong sale price before the stud fee catches up.
That window can close quickly.
Once a sire’s fee rises, the same sale price produces a different economic result. A $200,000 yearling by a $15,000 sire is a major commercial win. A $200,000 yearling by the same sire after his fee rises to $100,000 is a very different outcome.
That is why this part of the July analysis cannot rely only on the sale return against stud fee. The better question is not simply, “What was the return ratio?” The better question is, “Did the sale produce enough gross dollars, enough margin, and enough risk-adjusted return to make the mating worthwhile?”
For July, the best opportunity often comes when several things line up at once: a forward physical, a sire the market wants to evaluate early, enough catalog-page support, and a stud fee that has not yet fully caught up to buyer demand.
That is the real economic edge. Not just a high return ratio. A well-timed mating before the market fully reprices the sire.
What 2025 Said Compared With September
The 2025 July sample also became more interesting when compared with same-sire Keeneland September results.
September had a higher average because of its elite Book 1 and Book 2 top-end depth. That is expected. But the July photo-scored sample had a higher median than the overall September population, which suggests July can be a more concentrated commercial group.
Comparing horses by the same sire is a way of creating a fairer comparison. Instead of looking at all July horses against all September horses, I looked at how July yearlings stacked up against September yearlings by the exact same stallion. In simple terms, it asks: “Did this July horse sell better or worse than other offspring of the same sire later in the season?”
That same-sire comparison showed that some July horses performed very well relative to later September horses by the same sires. Strong physicals in July could land high relative to same-sire September medians. Horses in the 7.8+ band were especially strong. The 7.3–7.7 group also performed well.
That supports the idea that July can be a good placement for the right yearling.
But it also reinforces the main point. July is not necessarily a deeper elite physical sale. It is a sale where a forward, well-timed horse can stand out before the larger September marketplace creates more competition, which is why foal-crop timing and sales placement matter.
A horse that looks strong in July may have more room to separate. A similar horse in September may become one of many.
The Practical Breeder Takeaway
For breeders, the July question should not be, “Can this horse simply get into July?”
The better question is, “Can this horse stand out in July?”
The ideal July horse should have several things working together:
- a forward, mature physical;
- at least a good usable commercial body;
- enough balance and presence to be believable early;
- a sire buyers want to evaluate now;
- enough Black-Type Score support to give the page credibility;
- a realistic sale return against stud fee, interpreted in the context of full costs;
- enough gross-dollar upside to make the mating worthwhile;
- clean enough veterinary presentation;
- enough individuality to avoid blending into the group.
July does not require perfection. But it does require alignment.
A yearling with a 7.0 physical may be acceptable, but it needs help from sire and page. A 6.7 or lower physical is already in the commercial danger zone unless something else is very strong. A 7.5+ physical can create real separation, especially when matched with the right sire and a workable stud-fee base.
This is where the Horse Sense Pedigree Analytics model becomes useful. It separates the question into parts.
- Is the body good enough?
- Is the page strong enough?
- Is the sire commercially active enough?
- Does the sale return against stud fee make sense?
- Is there enough gross-dollar return to justify the risk?
- Does the result match what the public information suggested?
That is a better way to evaluate July than simply asking whether the horse sold high or low.
Redefining “Select” for the Fasig-Tipton July Sale
The useful takeaway from this study is not that physical quality does not matter. It clearly does. In the 2025 sample, better physical scores separated price more clearly than Black-Type Score alone. The stronger physical groups brought higher average and median prices, while the discount physical group was punished.
But the data also showed that July select is not a simple label for elite conformation. A yearling can be useful, forward, and commercially well placed without being a true premium physical. That is the important distinction.
For breeders and consignors, the question is not simply whether a horse is good enough to enter in July. The better question is whether the horse has enough early commercial alignment to stand out there.
That means looking at the horse in layers:
- Is the physical forward and commercially believable?
- Does the catalog page give buyers enough confidence?
- Is the sire one the market wants to evaluate early?
- Does the sale return against stud fee make sense in real dollars?
- Is the horse likely to stand out in July more than it would in a deeper September marketplace?
That is where the Horse Sense Pedigree Analytics model becomes useful. It does not reduce the horse to one number. It separates the visible physical, the catalog page, the sire market, and the economics so each part can be judged more clearly.
July does not reward perfection. July rewards alignment.
As the 2026 July Sale approaches, that is the framework I will be using. The goal is not just to find the best-looking horse in isolation. The goal is to identify which horses have the right combination of body, page, sire demand, timing, and economic structure to make sense in the July marketplace.
Horse Sense Pedigree Analytics