The Real Edge Was Never the Vet Data. It’s Knowing What a Barn Won’t Tell You.
Y is worth considerably more than X
A confession first, because it earns the rest of this piece the right to be taken seriously.
Long before any of this, before I even heard of InCompass, before EquiTAPS, before anyone was building custom past performances with embedded proprietary figures I was sitting at my kitchen table with a printed past performance sheet and a pen, and my wife Blondie was reading me Thorograph numbers off the page so I could write them into the margins by hand. That’s how far back my own use of supplemental figures goes. Legally, for my own personal handicapping, the same as plenty of serious horseplayers have done for decades. A full card done that way took an hour, sometimes more. We used to joke that we wished someone would just build it this way and sell it to us, we’d have gladly paid for the finished product instead of doing it by hand at the kitchen table. I tell you that not to be cute, but because it’s the honest starting point for everything else in this piece. We’ve been doing some version of “enhanced past performances” since long before anyone called it that, and we wanted exactly the kind of product this story is now investigating. What’s changed isn’t the idea. It’s the scale, the automation, and as this story has made clear, who has access to which parts of it.
I haven’t printed a physical past performance sheet in years now. Everything is digital, instant, and infinitely more sophisticated than a pen and a phone call to my wife. We were ahead of the curve once, in our own small way. This story has taught me how far the curve has actually moved, and how much of it sits in places most horseplayers, including this one, didn’t fully understand until now.
What Looking Into This Has Actually Taught Me
I want to be direct about something before going further. Pursuing this investigation across the last several days has been genuinely valuable to me personally, separate from anything it’s produced in print. I know more now about how custom past performance products get built, how easily proprietary data can be merged into a private tool, and how a sophisticated operation thinks about information than I did when this started. That knowledge is powerful, and I’d be doing you a disservice if I didn’t say so plainly.
Here is the most important thing I’ve learned, and it reframes the entire story.
The vet data, the Shockwave entries, the intra-articular injections, the off-dates was never the real edge. It’s a real edge, and a real violation if it was obtained the way this investigation suggests. But it is the least sophisticated version of something much bigger, much older, and much harder to detect. Vet data tells you something about physical condition. It does not tell you what the people who know the horse best actually think.
That distinction is the whole story, and it comes down to two numbers I’ll call X and Y.
X and Y
A trainer who likes their horse is worth X. That information has some value, trainers know things the public doesn’t, and a barn that’s confident going into a race is worth knowing about. But X is a limited, often unreliable signal, for a reason every serious horseplayer already understands intuitively even if they’ve never put words to it: trainers, like owners, like handicappers, fall in love with their horses. They watch them work every morning. They’ve invested months of hope and effort and money into them. That proximity creates bias, and bias clouds judgment. A trainer’s optimism about their own horse is, more often than not, a losing proposition to bet on blindly. History bears this out again and again, the most expensive mistakes in this game often come from believing the people closest to a horse over the evidence in front of you.
A trainer who does not like their horse, or who knows something is off, is worth Y. And Y is worth dramatically more than X, for the simple reason that it cuts through the bias entirely. Nobody talks themselves into doubt about their own horse without a real reason. If a trainer or the assistant trainer who actually has eyes on that horse every single morning knows it isn’t ready, has a hidden issue, or simply isn’t the horse the past performances and the public money think it is, that is not an opinion clouded by hope. That is closer to a fact, quietly held inside a barn, and it is worth far more than any number of people telling you who they like.
This matters most precisely where past performance models and public handicapping are weakest: well-bred, high-profile first-time starters, especially two-year-olds, where there is the least public data to draw from and the most public money chasing a name, a pedigree, and a barn’s reputation. A computer model has very little to work with on a horse that has never raced. A person standing in the barn every morning has everything to work with. That asymmetry is exactly where a sophisticated operation would look for an edge, and it is not an edge that requires beating anyone on who is going to win. It only requires knowing, with real confidence, who isn’t.
This Is Not Theoretical
I want to be careful here, because this story has required precision at every turn and this section requires more of it than most. I am not naming a specific barn, a specific assistant trainer, or a specific operation, because I do not have a confirmed, documented instance to put in front of you. What I have is direct personal experience, and I think it’s important to say so honestly: I have been approached, myself, at various points, for trip observations and hidden trouble lines, the kind of on-the-ground, in-person knowledge that doesn’t show up in any database and never will. That tells me the mechanism described above isn’t speculation. It’s a real, active part of how information moves in this business, and I’ve seen the edge of it firsthand.
Pat Cummings, who runs the National Thoroughbred Alliance and has spent years studying the economics of rebates and computer-assisted wagering, made a related point publicly this week that’s worth putting in front of you in his own words. Cummings noted that a bettor wagering a billion dollars a year in this sport receives roughly $120 million back in rebates, and asked, pointedly, what kind of operating expense that bettor is running against that revenue, the clear implication being that an operation at that scale has every financial incentive to acquire any edge available, “however they get it.” When pressed publicly on whether regulated information should be available to everyone, Cummings did not soften his position. He called the current state of affairs entirely unfair, and said directly that this story “has the potential to be the biggest scandal in modern American racing history,” adding that, in his words, “the emperor’s been naked for years.”
I want to be honest about what that is and isn’t. It is the stated belief of a highly credible, knowledgeable industry voice, given on the record and in his own name. It is not, on its own, proof of a specific transaction. But when I put his economic reasoning next to my own direct experience of being approached for exactly this kind of information, and next to everything this investigation has already uncovered about how easily a sophisticated custom data product can be built, the picture gets harder to dismiss as coincidence.
What We Actually Know Now
History tells us, plainly and repeatedly, that where there is significant money, there is eventually corruption. Add legalized gambling at the scale this sport now operates, billions wagered annually, a rebate structure that rewards volume above almost everything else and the path to that conclusion gets considerably clearer, even without a single name attached to it yet.
Here is what this investigation has actually established, stated as precisely as I can state it. We know it is technically straightforward to compile a sophisticated custom past performance product, because we have seen one. We know it is equally easy to fabricate a convincing fake of the same general style, because we have now seen that too, and the difference between the two is not always obvious at first glance. We have learned, the hard way and in public, that having information is not the same as understanding what it means or how to use it responsibly, a lesson this publication had to apply to its own reporting along the way. We know there is some form of information pipeline operating in this business, separate from and more sophisticated than anything sitting in a regulatory database, because people with direct knowledge of the industry, myself included, have encountered it directly. And we know, from basic logic about what is actually valuable, that the sharpest operators are likely not paying for who a barn likes. They are probably paying, or trading on relationships built over years, for who a barn does not.
That is the iceberg. The leaked vet data was the visible tip, concrete enough to investigate, specific enough to fact-check, small enough to eventually resolve one way or another. What sits beneath it is older, less documented, and considerably harder to ever fully prove. I don’t know its exact size. I don’t think anyone outside of the people directly involved in it does. But I know enough now, from this investigation and from my own history in this game, to say with confidence that it’s there, and that pretending otherwise would be its own kind of dishonesty.
