How to Pick Horses to Lay: A Data-Driven Checklist

1. Start from the right question

Most people asking “which horses should I lay?” are really asking “which horse will lose?” That’s the wrong question — in a 10-runner race, nine horses lose, so that’s easy and worthless. The right question is: which horse is more likely to lose than its price implies?

A horse trading at lay odds of 6.0 is priced to win about 17% of the time. If the data says its true chance is closer to 10%, laying it has positive expected value. You’re not predicting a loser — you’re finding a mispriced one. Every item below is a filter for that.

2. Check the price is in the 5–10 band

Price decides everything: your liability, your profit, and the strike rate you need to break even. Below 5.0 the margins are too thin — one loss undoes four wins. Above 10.0 the liability spikes — one loss undoes nine. The 5–10 band is the sweet spot where horses lose often enough and the liability stays manageable. If the horse isn’t in that band, skip it — no matter how weak it looks.

3. Confirm the field is 6+ runners

Field size sets the base probability. In a 4-runner race every horse has ~25% just by turning up — too high for reliable laying. At 6+ runners, more horses can beat your target, which is exactly what you want. Below 6, strike rates drop below break-even. It’s a hard floor, not a preference.

4. Read the historical strike rate

This is the single strongest predictor. A horse that has lost 18 of its last 20 starts carries that record into the 21st — and the market doesn’t always price it in, especially after a break or a class change. Don’t look at one flashy run; look at the rate across the recent career. A low win rate over a meaningful sample is the backbone of a good lay.

5. Look for the weakening signals

On top of raw strike rate, these push a horse from “maybe” to “lay”:

• Stepping up in class. A horse beating weaker company doesn’t always transfer up. The market often over-rewards a recent win.
• Returning from a layoff. Fitness doubts the price rarely captures fully.
• Unproven conditions. No winning form at the trip, the going, or the track is a quiet red flag.
• Drifting in the market. If informed money is walking away from a horse, that’s a signal — not a bargain.

None of these are narratives about jockeys or paddock looks. They’re data points you can check before every race.

6. Time the market

Place your lay roughly 15–60 minutes before the off. Too early and the price is unreliable — not enough money in the market. Too late and you’re chasing a moving target with thinning liquidity. That window gives you a price that reflects genuine opinion with enough liquidity to get matched at the odds you want.

7. Exclude what you can’t model

Discipline is as much about what you don’t bet. Maiden races — first-time or non-winning runners — have no historical strike rate to work with, so there’s no edge. We exclude them entirely. If the data isn’t there, there’s no bet. A pass is a perfectly good outcome.

8. A worked example

Run that checklist across an entire day’s racing by hand and it’s hours of work. That’s the point of a model: it applies these exact filters to every qualifying race and surfaces only the horses that pass. See the method in our full strategy guide, or the daily output on today’s lay tips.