How to Bet With Ratings: From Raw Numbers to Rated Prices

What a Rating Actually Is — and Which Run to Rate the Horse Off

A time or speed rating is nothing mysterious. It's simply a number describing how well a horse ran on a given day — how fast it went, adjusted for the conditions it faced. That's all. The number itself isn't the skill. As Kingsley puts it, the big question is always the same: what do we think this horse is going to run today? The rating tells you what it has done; your job is to work out what it will do.

That's where most punters go wrong. A horse doesn't have one rating — it has a history of them. Do you rate it off its last start? Its best run of the last two? Its peak figure for the year? Its previous first-up runs, or its runs at today's distance? Each of those choices gives you a different number, and each captures a different slice of the truth. Kingsley has spent years testing these selection rules, and the lesson he draws from all that work is simple: the punter who thinks about which run applies to today's race is miles in front of the punter who just reads the biggest number in the column.

A useful way to think about it is building an expected rating. Look for the horse's most comparable past setups — same stage of preparation, similar distance, similar racing pattern — and ask what it produced in those. If a horse has consistently improved sharply at its second run back in previous campaigns, that history tells you more about today than last start's flat figure does. Project the number forward, don't just copy it forward.

Context First: Class, Weight, Jurisdiction and the Ten-Start Rule

A raw figure means nothing until you know the company it was earned in. Kingsley's approach anchors every speed figure to race strength — the true class of the race — and adjusts it for the weight the horse carried then versus what it carries today. A flashy figure run against weak opposition gets marked down. A solid figure under a big weight against strong company gets marked up. And when a horse is stepping up sharply in class, treat its old numbers with suspicion: the trick is predicting what it will do at the new level, not assuming the old figure travels with it. Kingsley notes the market often over-pays for the assumed improvement of young horses making big class jumps.

Some ratings actively lie to you. Horses shipping between jurisdictions are a classic trap — a horse that's been running midfield in stronger company interstate can carry inflated figures into a weaker local scene, start short, and get rolled, particularly if it's changed stables or is carrying a hidden problem. Odd-conditions runs are another: a figure recorded on a bog track doesn't automatically translate to a good surface. A lesson Kingsley teaches repeatedly is that context filters beat raw figures — sometimes the smartest thing you can do with a rating is throw it out.

Finally, respect sample size. Kingsley's rule of thumb is that you can't get a real read on a horse until it's had around ten starts. Lightly raced types are still climbing through the grades, trying new distances, and haven't shown their wheelhouse yet. Their ratings are a sketch, not a portrait — lean on them lightly and expect surprises in both directions.

Ratings as a Filter: Eliminate the Field, Then Hunt for Value

The first job of any set of ratings is not to find the winner — it's to get rid of the losers. In most races, the winner comes from the top handful of rated runners, and horses in the bottom half of the ratings rarely trouble the judge. So use your numbers the way Kingsley does: separate the field into live chances and dead wood, and stop wasting mental energy on horses that would need a miracle.

Once you're down to three or four key chances, the real work starts: which of the survivors is over the odds? This is where the margin of the ratings advantage matters. A horse rated clearly on top of the field is a much stronger signal than one that scrapes into first place by a whisker. And value doesn't only mean backing something — if your work says the favourite's figures were earned in weak company or on an unsuitable surface, the play might be to bet around it, or against it, rather than with it.

Kingsley also warns against betting the whole shortlist. In today's tight, low-margin markets, saver bets and second and third selections tend to drag a winning punter back to break-even. Filter with the ratings, dig for the value among the survivors, then have the discipline to bet only where the price is genuinely wrong — which some days means no bet at all.

Why You Can't Price a Race Off Ratings Alone — the Benter Blend

It's tempting, once you've rated a field, to convert your numbers straight into prices and declare you'll back anything over your quote and lay anything under it. Kingsley — echoing his own mentor — calls that a dangerous belief. An assessment is a stack of assumptions made under uncertainty: about the tempo, the track, the barrier, the jockey, the horse's fitness. Marking a horse at a price doesn't make it a correct assessment of its winning chance, and if your assumption about one runner is wrong, the price on every other runner in your market is wrong too. Plenty of things that decide races — pace, luck in running, stage of preparation — simply aren't in a raw time figure.

The answer, used by some of the most successful betting operations in the world, is to blend. The Bill Benter approach treats the public market as information, not as the enemy. Take your rated price and the market's price and weight them together — split the difference in whatever proportion your confidence justifies — for every runner, then re-frame the whole market so it adds back up properly. The market knows things your model doesn't: stable confidence, money moves, whispers from the track. Folding that in produces a sharper final price than either source alone.

This is also why Kingsley pays so much attention to market moves. A horse firming hard against your assessment isn't an insult to your ratings — it's a data point. Sometimes it exposes a hole in your work; sometimes it confirms the market has over-reacted to the obvious and your price is the right one. Either way, the punters who last are the ones who let the market inform their numbers instead of arguing with it.

Proving Your Ratings Have Signal: The Back/Lay Spread Test

How do you know your ratings are actually worth betting on? Kingsley's test works from both ends. Don't just track how your top-rated horses perform — track how the horses your ratings hate perform as well. If your fancied runners outperform the market and your unfancied runners underperform it, the gap between the two proves your numbers genuinely discriminate. If both groups perform the same, you don't have a rating system, you have a random number generator.

This spread test is especially valuable during a losing run. A lesson Kingsley teaches from his own toughest stretches is that headline results can be bleeding while the underlying model is still healthy: the top-rated runners holding up while the bottom-rated ones keep losing tells you the signal is intact and the problem lies in selection, staking or price — not in the ratings themselves. That distinction stops you tearing up years of good work over a normal patch of variance.

The broader principle: test everything, record everything, and be honest about the results. Every rule you bet off should have survived proper back-testing before it gets a dollar, and even then, temper what the back-test promises before assuming you'll match it in the real world. Ratings can be the foundation of an edge — but only if you keep proving that edge exists, one honest review at a time. And whatever the numbers say, never bet more than you can comfortably afford to lose.

Common questions