Does Studying the Form Actually Change Anything?

Start with the finding that annoys people, including us.

Backing the highest officially-rated horse in every race returns -15.37%, and it wins 19%. Backing the favourite returns -9.10%. So the horse the official handicapper rates best is a worse bet than the horse the market makes favourite.

In handicaps specifically it is -15.91%, worse again.

Our own model tells the same story from the other direction. It reads far more than the official rating: recent form, course record, distance record, the lot. Across 10,478 races its top pick wins 24.0% of the time, and it has returned −14.4% on 10,297 bets, which is −£14,825.

That is a lot of form study arriving at a loss.

The form is already in the price

The reason is not that form is meaningless. It is that everybody else has read it too.

By the time a race goes off, the market has absorbed the form book, the ratings, the trainer's recent record, the ground, the draw and whatever was said on the television. The price is the summary of all of it. So being right about the form gets you to roughly where the price already is, which is the point at which the bookmaker's charge decides the outcome.

You can see how good that summary is in our accuracy scoring. Across 10,478 races our model reads 0.098 against the market's 0.090, where lower is better. The market is the sharper forecaster, and it is not particularly close.

That is not a criticism of anyone's model, ours included. It is a statement about what a market with a lot of money and a lot of opinion in it is good at.

The corollary is important though: to profit from form study you do not need to read the form well. You need to read it better than the collective market, on this race, today. That is a much higher bar than most tipping copy implies.

What form study is actually for

It buys you three things, and none of them is an edge over the market.

It buys understanding. A race makes more sense when you know why the second favourite is where it is, and racing is a great deal more enjoyable when you can read what is happening.

It buys discrimination between bad bets and worse ones. Somebody who reads form will not back the top-rated horse in a twenty-runner handicap thinking it is a solid bet, and will not be surprised when a soft-ground specialist runs badly on quick ground.

And it buys the ability to notice when a price is out of line, which happens, and which is the one route to a genuine edge that we have seen any evidence for. That route, taking a price before the market corrects, is set out in how some people make a living at this.

What it does not buy is a way round the charge. 24 approaches, 30,082 races, 0 of them in profit, and several of the losers were built on more form data than any person reads in an afternoon.

A word on all of this

None of these pages is a tip, and none describes a way to win. They describe what betting costs, which is a different and more reliable subject. If your betting has stopped being fun, BeGambleAware has free, confidential help, and the National Gambling Helpline is on 0808 8020 133.

Common questions

Does backing the top-rated horse make money?

No. It returns -15.37% across 26,684 races, worse than simply backing the favourite.

Why is the favourite better than the official ratings?

Because the market includes the ratings and everything else besides. The favourite is the crowd's summary of all available information, and it is a sharper forecast than any single input.

Is your model better than the market?

No. On accuracy scoring it reads 0.098 against the market's 0.090, where lower is better.

So is reading form a waste of time?

Only if the point is profit. It makes racing far more interesting, and it stops you making the worst bets, which is worth something even though it is not an edge.

What would it take to beat the market on form?

Being more accurate than the collective judgement of everyone else who has read the same form, by enough to cover the charge. We have not managed it and we publish the figures either way.