What a rating tells you, and what it cannot
A rating is a claim about a horse. A price is a claim about a race. They look like the same kind of object because both arrive as numbers, and they are not the same kind of object at all.
Two different questions
A rating is an estimate of how good a horse is, on the assumption that it gets the chance to show it. It is built from what the horse has already done and it is designed to travel. That is its whole purpose. It has to mean roughly the same thing in April at Musselburgh as it does in August at Goodwood, or it is no use to anybody.
A price is a much narrower thing. It is a statement about one race. This field, this trip, this ground, this draw, whatever pace is likely to be on, on this particular afternoon, with these particular rivals.
One is a property of the horse. The other is a property of the situation. If you own a horse, you already know this in your bones, because you have watched a well handicapped animal run a stinker in a race that did not suit and come back a fortnight later and win off the same mark.
What a rating is genuinely good at
Ranking. That is the honest answer and it is not a small one.
A rating tells you where your horse sits against the ones it will be meeting. It tells you which races are realistically open and which are aspirational. It tells you roughly how much improvement would be needed to be competitive a grade higher, and whether the improvement you think you saw last time was real or was a soft race flattering everybody in it.
It is also stable, which is underrated. A number that changes slowly is one you can plan around, and entries and campaigns are rating work.
What no rating can contain
Everything that belongs to the race rather than the horse.
Whether your horse gets an uncontested lead or has two others beside it that want the same ground. Whether the pace goes off too quick and comes back to the closers, or dawdles and turns the whole thing into a sprint your stayer cannot win. Whether the ground has genuinely changed since the entries were made. Whether the draw matters at this track, in these conditions, on this day.
None of that is exotic. Every form student in the country reads for it. But it cannot live inside a number that is meant to describe the horse in general, because it is not about the horse in general. It is about Thursday.
A rating answers who is best. It does not answer what is going to be asked.
The two mistakes
The first is treating a rating gap as though it were a price gap. Your horse is clearly the best in the race on the figures, so it should be short, and if it is not short then the market has missed something. Sometimes true. Usually it means the market has noticed something about the race that the figure was never built to carry.
Best is not the same as likely, and the numbers are blunt about this. Across 60,671 races with one clear favourite, the favourite won 35% of the time and did not win 65%. The favourite is very often the highest rated horse in the field. Being the best one in the race and getting beaten are not in tension. They are the normal state of affairs.
The second mistake is quieter and it is the one that catches researchers rather than owners. It is letting the price leak into the rating. If any part of your figure has been informed, even indirectly, by what the market thought, then you no longer have two opinions to compare. You have one opinion wearing two coats. Test it against the market and you are testing the market against itself, and you will get a flattering answer that means nothing.
Keeping them apart is most of the work
This sounds like a technicality. It is closer to the entire job.
In practice it means deciding in advance what goes into the horse side and what goes into the race side, and then being disciplined enough not to move the line when the results are disappointing. It means recording our own timestamped market snapshots before the off, so that when we hold the price constant we know exactly what price we held and when we read it. A price recovered after the race has already been coloured by the result, and once that has happened the separation you thought you were maintaining has quietly collapsed.
And it means accepting the answer when the separation is clean and the finding is dull. We have had several of those. They are more useful than the exciting ones, because a dull finding you can trust tells you where not to spend the next six months.
What the figure is actually for
If you own a horse, a rating is a planning instrument, not a forecast. It tells you what your horse is, and it tells you where it fits. It is the right tool for deciding what to enter and when to be patient.
It is the wrong tool for deciding whether Saturday is the day. That is a question about the race, and the race is a different object with different information in it. The trainer who says the figures look great but there is too much pace in it is not contradicting the rating. They are answering a question the rating never claimed to answer.
This is a general account of what ratings can and cannot carry. It is not advice on any individual horse or race. Elite Pass has no validated betting edge and does not claim one.
Related research
- The horse has not changed, the conditions haveA horse that ran badly last time usually did not run badly. It was asked a different question, and it answered that one honestly.
- Reading the race, not the horseIf the best horse loses two times in three, which one is best is rarely the useful question. Here is the one we ask instead, and how we ask it.
- The bookies aren't afraid of AI. They should be rubbing their hands together.AI is becoming astonishingly good at explaining why a horse might win. That may be exactly why bookmakers have very little to fear.
Elite Pass publishes research, not guaranteed outcomes. Findings are measured against the market and remain subject to replication and sample size. The evidence room and the public ledger carry the full record.