How to Track Horse Racing Bets the Right Way

Nine columns and four metrics turn a vague impression of your betting into numbers you can use, with 13 years of data on why sample size matters.

8 min read

Ask most bettors how they got on last month and you will hear a feeling. They remember the good Saturday at Newbury, the steamer that shortened, the trixie that nearly came in. What many do not have is a number.

Tracking gives you that number. When you record every bet, you can see which angles are paying their way, whether a losing run looks like ordinary variance (the natural ups and downs of results), and whether your method still suits the markets you bet in. It turns a vague impression into something you can work with.

How should you track horse racing bets?

Record the date, course, selection, bet type, odds taken, stake, return and outcome for every bet. Then work out four things: strike rate, ROI, Impact Value and the A/E ratio against Betfair SP (the starting price on the exchange). Each one is explained below.

With those numbers you can judge whether your results look stronger or weaker than the prices implied, and whether the sample is still too small to trust. 13 years of GB and Irish racing show how easily small samples can mislead.

Which columns does a bet tracker need?

Tracking is easier when each entry takes seconds. Nine columns are enough to answer most questions later.

For every bet, record:

  • the date
  • the time of the race
  • the course
  • the horse you backed
  • the bet type (win, each-way, place, lay or dutch)
  • the odds you actually took
  • the stake
  • the return
  • the profit or loss

Take the odds at the moment you placed the bet, not the SP that came in later. Take the return as the cash returned after the result, including any place portion of an each-way bet. Profit is return minus stake.

If a bet came from a strategy that needs context, add one note column. “Top-rated on my ratings, soft going” is a usable note. “Fancied him” is not. The notes are for you in six months’ time, when you want to know whether a particular angle held up.

Useful extra columns are going, class and Betfair SP at the off, because they let you test later whether your results lean on one condition.



Which metrics show how you are doing?

Tracking is the first step. The numbers you work out from the rows are what you act on. Four numbers and one chart are worth knowing.

Strike rate. The number of winners divided by the number of bets. It tells you how often your bets win, not how much you make. A 20% strike rate means very different things at 20/1 and at 1/4.

ROI. Return on investment: total profit divided by total stake, as a percentage. It tells you what you make per pound staked. ROI needs a decent sample behind it. A figure from 50 bets can swing a long way on one or two winners.

Impact Value. Your strike rate divided by the strike rate you would expect from the field at random. An Impact Value of 1.00 means your selections win as often as random choices. 1.10 means they win 10% more often. It is a good first filter, though it does not show profit on its own, because the market may already charge for the extra strike rate in shorter odds. Our guide to Impact Value goes into more detail.

A/E ratio. Actual winners divided by expected winners, where the expected figure comes from the Betfair SP your bets were struck at. Our guide to the A/E ratio explains it fully. Above 1.00 means your horses won more often than their prices implied. Around 1.00 means the prices were about right.

Bank curve. Stake one unit on every bet and plot cumulative profit over time. It is a picture more than a metric, and many people find it the quickest way to read their results. A steady slope across a thousand bets tells a different story from one big jump in week 27.

What do 13 years of racing show about sample size?

A tracker is only as useful as the sample behind it. The Betfair favourite is a good yardstick. In the verified database for GB and Ireland (1 July 2013 to 14 October 2026, Betfair SP, level stakes, 2% commission), the favourite won 34.4% of 167,257 races at A/E 1.00, with a return of -1.4%. The longest losing run was 28 races.

That is a very steady picture over thirteen years. It also shows that even the shortest-priced horse in each race can go through long quiet spells, which is worth knowing when you look at your own results.

Now look at shorter windows. In an earlier analysis, the favourite’s win rate moved between about 33.0% and 34.5% from one year to the next, with A/E between 0.98 and 1.00. Each year held more than 12,000 runners. A bettor who tracked favourites for one year would see a slightly different picture from one who tracked another year. The long-run figure is the steadier guide.

These figures describe those years of racing and will not repeat exactly. The point is the pattern: strike rate settles down slowly, A/E more slowly, and ROI after commission slowest of all, because a few big-priced winners can dominate it. A system that looks great after a small sample needs more bets before you rely on it, and so does one that looks poor. Tracking is there to stop a bad spell pushing you into a panic decision, and it does not promise that every angle will survive.

How do you work out your own A/E?

A/E is where tracking turns from admin into a tool. The calculation is short:

  1. For every bet, take the Betfair SP at the off and divide one by that price. That is the implied probability.
  2. Add those probabilities across your whole sample. The total is your expected wins.
  3. Count your actual wins in the same sample.
  4. Divide actual wins by expected wins.

An A/E of 1.04 would mean your horses won 4% more often than their prices implied. Whether that is meaningful depends on the size of the sample, so use it alongside the number of bets. An A/E of 0.96 would mean the reverse.

Use Betfair SP rather than bookmaker SP as the pricing basis. Bookmaker prices include a margin, so the implied probabilities add up to more than 100%, and that pulls A/E down for reasons unrelated to your selections. Betfair SP has a much smaller margin, so it is the cleaner yardstick. Apply 2% commission on net winning markets only. A market here means one race on the exchange: if your combined bets in that race finish in profit, deduct 2% of that profit.

When can tracking help you retire an angle?

Tracking also helps you let go of angles that are not earning their place. Many bettors add new ideas and never remove old ones, and the spreadsheet fills up. A tracker gives you the evidence to tidy it.

Two examples from an earlier analysis (GB and Irish runners, July 2013 to March 2026) show what to look for. Days since last run is easy to track, because you record the date of the race you backed. Across 72,243 runners returning within seven days, the strike rate was 14.1% against a field-wide baseline of 10.85%, an Impact Value of 1.30. The A/E was 1.02, so the market priced most of that recency signal. Across 247,284 runners returning after eight to fourteen days, the strike rate was 10.5% and the Impact Value 0.97.

Career starts show a flatter version of the same pattern. Horses on their sixth to fifteenth career run won 11.9% across 473,349 runners (Impact Value 1.09). Horses with more than thirty career starts won 8.6% across 155,520 runners (Impact Value 0.79). The A/E was 0.99 in both groups. The Impact Value shows where the winners are, and the A/E shows how much of that the price already contains.

Apply the same two checks to your own angles. An Impact Value that stays above 1.10 across a few hundred bets tells you your selections win more often than average. An A/E at or above 1.00 over Betfair SP, after commission, adds that the prices were not already taking it away. If neither shows up after a fair sample, that angle is a candidate to retire.

A reasonable starting checkpoint is 300 to 500 bets per angle. Angles with longer prices or lower strike rates need more bets than that, because a few winners can swing the result. Below the checkpoint, treat your result as a question to keep asking. A 100-bet sample of your own picks can swing a long way either side of the true figure.

Where to start

Pick one angle you already follow, set up the nine columns, and record every bet at the moment you place it. If you would like to see a full day’s runners with their ratings and Betfair prices, open today’s race cards and choose the races you want to follow.