Open any UK racing site and you’re pointed at one number. RPR. Topspeed. The site’s own headline rating. The idea is that this one figure has done the thinking for you and your job is to nod along.
It’s a comforting story. It’s also a poor way to read a race.
So should you use more than one rating? Yes, I think so. In my data, the top-rated horse on a single rating wins about 32% of the time. When three of my ratings all pick the same horse, that rises to 41%. But the price already allows for it. Agreement tells you which horse to look at. It doesn’t, on its own, tell you the horse is good value.
The short version
- Each rating measures something different, so they disagree more often than you’d expect.
- When ratings agree on the same horse, the win rate climbs. The A/E ratio stays close to 1.00, which means the market has already priced it.
- Treat them as ideas to test, not systems.
Why isn’t one horse racing rating enough?
A single rating is built to compare horses, and that’s useful. It’s also incomplete. A speed rating cares how fast a horse covered the ground. A handicap mark cares what weight makes the field equal. A power rating tries to blend several measures into one chance of winning. Asking one of them to answer every question is like asking one witness to describe a whole crime scene from one corner.
Two terms first. “Rank 1” just means the horse a rating puts top in its race. PR is my own power rating, an estimate of each horse’s chance of winning. The A/E ratio compares how many winners a group of horses actually produced with how many the odds said to expect, so 1.00 means the price was fair.
Here’s what the top-ranked horse on each rating did in UK racing from July 2013 to March 2026, out of more than 1.5 million runners in the database:
| Rating | Win rate of Rank 1 | A/E |
|---|---|---|
| PR Rank 1 | 32.3% | 0.99 |
| VDW Rank 1 | 25.6% | 0.99 |
| Dynamic Speed Rank 1 | 17.8% | 1.00 |
| Official Rating Rank 1 (handicaps only) | 14.3% | 0.99 |
Every one of them sits around 1.00. The market can see the same numbers you can, so following any single rating gives you a respectable strike rate and no edge. These figures describe those 13 years, and the next 13 might look different.
What does each rating measure?
These four answer four different questions:
- PR is my own power rating. It’s an estimate of a horse’s chance of winning today, allowing for the field, the conditions and recent form. It’s the nearest thing to an all-in answer, which is why PR Rank 1 has the highest win rate of the four. I go into it in the PR rating stats.
- VDW is a long-running form-based rating that leans on consistency and class over peak figures. It tends to favour the sound, reliable horse over the talented but unpredictable one. There’s more in the VDW rating stats.
- Dynamic Speed is time-based, adjusted for going and class. It doesn’t care about consistency or whether the conditions suit. It just asks how fast the horse has run. Speed ratings explained covers the idea.
- Official Rating is the BHA handicapper’s mark, which sets the weight a horse carries. It’s built to make the runners equal, so the fact that the top-rated horse still wins 14.3% of handicaps tells you the handicapper is good at the job and the market has noticed. I look at it more closely in the Official Rating post.
Four ratings, four ways of looking at a race… and they disagree a lot. PR Rank 1 and VDW Rank 1 are the same horse in 68,300 of roughly 160,000 races, which is about 43%. So in around 57% of races they pick different horses.
What happens when two ratings agree?
The win rate moves a long way:
| When these agree on the same horse | Win rate | Cases |
|---|---|---|
| PR Rank 1 and VDW Rank 1 | 35.6% | 68,300 |
| PR Rank 1 when VDW disagrees | 29.8% | 91,470 |
| PR Rank 1 and Dynamic Speed Rank 1 | 38.5% | 35,963 |
| VDW Rank 1 and Dynamic Speed Rank 1 | 32.3% | 28,341 |
| Top Official Rating and PR Rank 1, in handicaps | 29.0% | 25,429 |
That’s nearly a six-point swing for PR Rank 1 depending only on whether VDW sees the same horse. And when the top Official Rating in a handicap is also PR Rank 1, the horse wins about twice as often as the top Official Rating does alone (29.0% against 14.3%).
The A/E across all of these combinations sits between 0.98 and 1.01. The market isn’t asleep. What improves is how accurate your pick is, not the price you get. Agreement gives you winners, and it doesn’t give you a better price.
What happens when three ratings agree?
When PR Rank 1, VDW Rank 1 and Dynamic Speed Rank 1 all point at the same horse, it wins 41.0% of the time across 16,239 cases. That’s roughly two winners in every five.
Splitting those cases by race type and class shows where the A/E moves above 1.00:
| Three-way agreement | Win rate | A/E | Cases |
|---|---|---|---|
| Hurdles on turf | 46.8% | 1.02 | 3,628 |
| Chases on turf | 44.1% | 1.02 | 2,627 |
| Class 2 | 37.5% | 1.04 | 827 |
| Class 3 | 42.2% | 1.02 | 1,601 |
| Non-handicaps | 49.8% | 0.98 | 8,101 |
In Flat turf races and Class 1 the A/E drops below 1.00, because the market prices these more sharply.
The non-handicap row is worth a look for a different reason. Half the three-way agreers win, and the A/E is 0.98, so the market knows. That tells you something about which races have a clear favourite on every measure and which don’t.
How do I use several ratings in practice?
The aim isn’t to find a magic agreement and bet it blind. This part is my judgement, not something I’ve tested, so take it that way.
I let the race decide which ratings lead. In an all-weather sprint handicap I’d start with the speed rating and recent form. In a staying chase on soft ground, raw speed matters less, so I’d lean on PR and VDW and check the Official Rating isn’t pitching anyone too high.
When the ratings agree, the next thing I check is the price. If the horse the ratings like is a short price, the market has spotted it too, and I’m asking whether the price still leaves room. That’s a race-reading check, not a system.
When the ratings disagree sharply, PR says one horse, VDW another and speed a third, I’d treat it as a race where I’m less sure, and bet smaller or pass. A single headline number on a free site won’t show you that disagreement. Layering ratings does.
What this doesn’t tell you
Agreement between ratings sharpens your shortlist. It doesn’t promise a profit, because the price has already moved to meet it, and every figure here is UK racing from July 2013 to March 2026 at Betfair SP.
If you’d like to see where the ratings agree and disagree on today’s runners, the racecards show them side by side. See today’s racecards.
