← Blog · 📝 Article · 1 November 2026

Trainer Stats for Lays: What to Check and What to Ignore

Trainer Stats for Lays: What to Check and What to Ignore

Trainer stats help you lay a horse only when you read them against the price, the race type and the sample size. A trainer’s headline strike rate is a baseline, not a verdict. The useful questions are: how is this yard doing right now, how does it do in this type of race, and is the market already pricing that in?

This guide shows which trainer figures to check before a lay, how to read them, and where they mislead.

Donkeyradar
Add Data to Your Lay Decisions
DonkeyRadar uses historical strike rates and live market prices to identify lay signals before races across the UK, Australia and the US.
Explore DonkeyRadar

TL;DR

Table of Contents

What a trainer’s record can and cannot tell you in a lay

A lay bet wins when your selection loses. That sounds like the mirror image of backing, but the maths is different. When you lay at short odds, your liability is small and your potential profit is the backer’s stake, so you need your selections to lose very often. Our guide on why 70% at 1.40 still loses explains that arithmetic in full.

That is why trainer stats matter to a layer. Trainers control preparation, placing and the race types a horse is aimed at. A yard in poor form, or a trainer whose runners are habitually overbet in a certain type of race, can point you towards a horse that is priced too short for its chance.

But a trainer’s record describes a group of horses, not the one in front of you. It is one input among many: the horse’s own form, fitness, the going, the field and, above all, the price. Treat a trainer stat as a reason to look closer, never as a reason to bet on its own.

The trainer figures worth checking before a lay

Most racing sites show a trainer table with runs, wins, strike rate and profit. These are the columns that earn their place on a lay checklist.

Recent form: 14, 21 and 42 days

Short-term form shows whether the yard is firing. Many data providers publish trainer strike rates over windows such as 14, 21 and 42 days, and comparing a short window with a longer one shows an upturn or a dip. A yard whose 14-day figure sits well below its season average may have a bug in the stable, a run of unsuitable ground, or horses simply not ready.

Two cautions apply:

Strike rate by race type

A trainer who wins 15% overall may be far better with two-year-olds in the spring and ordinary with older handicappers. Split the figures by:

The more granular the stat, the more useful it becomes, as long as the sample does not shrink too far. Once you are looking at six runners at one course over one distance, you have an anecdote.

Runners after a layoff

Trainers handle returning horses differently. Some have their string ready first time out, others need a run. If the horse you are considering laying is returning from a break, check that trainer’s first-time-out record specifically. A yard that rarely wins first time out is a useful flag. A yard that is excellent first time out is an equally useful warning not to lay.

Profit or loss to level stakes

Strike rate tells you how often a trainer wins. Level-stake profit tells you whether the market paid you fairly for it. From a layer’s perspective the logic reverses: a trainer whose runners lose money to back at starting price, across a large sample, is one whose runners are routinely overbet. That is the territory where lays tend to be found.

How to read A/E and impact value

A/E stands for actual versus expected. It compares the number of winners a group actually produced with the number you would expect given the prices they ran at. Because it is built from the market’s own odds, it adjusts for the fact that a yard with lots of short-priced favourites will have a high strike rate without being special.

The usual reading is:

A/E value What it suggests
Below 1.00 Runners win less often than their prices imply
1.00 Priced fairly
1.10 and above Profitable to back, on a large enough sample
1.20 and above A strong edge for backers

These bands come from a guide to reading trainer statistics beyond win percentage, which treats A/E as the first metric to read. For a layer the useful end of the table is the bottom: a trainer or trainer-and-race-type combination with a consistently low A/E, over a big sample, is one where the market tends to expect more than the horses deliver.

Impact value is a close cousin. The British Horseracing Authority describes it as a comparison of actual winners with the winners expected by chance: 1.00 is par, 1.50 means 50% more winners than expected, and 0.50 means 50% fewer. A trainer-and-race-type figure well below 1.00 over many runs is a candidate flag for a lay. A figure well above it is a reason to step away.

Impact value compared with expected winners

How many runs are enough?

This is where most trainer-stat mistakes happen. A figure like 0 from 12 looks damning and means almost nothing.

A practical scale:

That scale is from the same trainer statistics guide. Other analysts use lower credibility thresholds for narrow trainer-and-course angles, but they all make the same point: read the volume of runners before the percentage, and treat a historical angle as evidence, not proof.

In practice, start with a trainer’s broad record over several seasons, where the sample is large, then narrow by race type or course only as far as the numbers hold up. If you have to go down to 15 runs to find an interesting pattern, the pattern is probably not there.

How many runs are enough?

Why trainer stats alone can lose you money

Price and market position matter more

The most common error is laying a horse because its trainer is cold, without looking at the price. One racing data publisher shows that a trainer with no winner in the last 14 days can still win about 30% of the time when the runner is favourite, against a 34% chance for favourites from trainers above 10% over the same period. The detail is in Inform Racing’s trainer form analysis. A cold yard does not stop a good favourite.

For a layer this has a direct consequence. You are not trying to find losers. You are trying to find horses whose price is too short for their chance. A trainer in poor form is a reason the price might be too short, not a reason the horse will lose.

The market learns

When a trainer pattern becomes well known, bettors adjust and the edge shrinks. A profitable historical angle can fade as the market becomes aware of it, and no historical statistic guarantees future results. Refresh your trainer figures regularly, and do not assume last season’s table describes this one.

Stable size and race code distort the headline

A large yard with plenty of well-fancied horses will post a strike rate that a small yard cannot match, and the two are not comparable. Jumps and Flat trainers also run on different cycles. Compare a trainer with its own history and with similar yards, not with the leaderboard.

A simple trainer check for each potential lay

Use this as one step in a wider selection process, not as a system on its own.

  1. Start with the price. Is the horse in the range you lay? Very short prices carry heavy liability against small profit, and very long ones lose too rarely to lay profitably.
  2. Check the trainer’s last 14 to 42 days. Note both strike rate and the number of runners. Ignore the figure if the yard has had only a handful of runners.
  3. Check the race type. Look at the trainer’s record in this type of race, at this course or distance, and confirm the sample is large enough.
  4. Check A/E or level-stake profit. A low A/E over a large sample is the flag you want. A high one is a warning.
  5. Check the horse itself. Recent form, days since last run, going preference and class.
  6. Work out liability. Decide your stake before you lay, and know the break-even point for the odds. Our break-even lay odds guide and lay betting calculator cover the numbers.
  7. Record the outcome. Without a log you cannot tell whether the trainer filter is helping.

For the wider process around this, see our data-driven checklist for picking horses to lay and our piece on horse racing data analysis for lay bettors.

Worked example: two yards, same price

Take two runners, both priced at 3.0 on the exchange in a UK handicap.

Horse A has two things in its favour as a lay candidate: a sizeable recent sample and a long-run figure that suggests the market overrates this yard in these races. Horse B has nothing, because both figures are too thin to count. You would look at Horse A harder, and you would still need to check its own form and the field before laying it.

Note that even for Horse A, a lay at 3.0 means a liability of twice your stake. A single loss takes several wins to recover, which is why the trainer filter is only worth using alongside disciplined staking.

Our perspective

We built DonkeyRadar on the idea that a single statistic, however neat, is a poor basis for a lay. Trainer form is one input. Our algorithm combines historical strike rates with live market prices, so the question is never only “is this yard cold?” but “is this horse priced too short for what the data says?”

That is also why we publish every signal before the race and keep the results permanently available. A trainer angle that looks good in a spreadsheet only means something when it survives live pricing, and you can only judge that from a record nobody can edit afterwards. Losing signals happen. Transparency is the whole point.

If you like checking trainer stats yourself, keep doing it. Use our signals as a second opinion built on tested statistical parameters rather than opinion.

Donkey

Try the signals alongside your own trainer checks

DonkeyRadar publishes lay signals for UK, Australian and US horse racing before each race, graded by staking tier, with direct links to the relevant Betfair Exchange market. The free tier gives you daily lay signals. DonkeyRadar Pro costs £29 per month after a seven-day trial and adds real-time alerts by email and Telegram, the full results history and API access for trading software.

Donkeyradar

You can compare the signals with your own trainer research on DonkeyRadar, and read how we approach selection in our Betfair lay betting strategy guide. You must be over 18 to bet.

Combining trainer stats with the rest of your lay process

Trainer data works best as a filter, and filters work best when they are few and tested. A sensible way to add trainer information:

If you want a broader view of how statistical models are put together, our guide to horse racing data models covers it, and 10 rules for profitable lay betting covers the discipline around staking and liability. If you want to see verified lay results in practice, our horse racing lay tips page shows how we report them.

FAQ

Is a trainer’s strike rate enough to decide a lay?

No. A trainer’s win percentage is a baseline check, not a decision tool. It ignores the horse’s form, the race type and, most importantly, the price. Use it alongside A/E or level-stake profit and the market position of the runner.

How many runs do I need before trusting a trainer stat?

Under 20 runs is noise. Around 100 runs is the minimum before leaning on a figure, and 500 or more is where you can trust it strongly. For narrow trainer-and-course angles, treat anything thin as a prompt for further checking, not a rule.

Does a trainer in poor form make a horse a good lay?

Not by itself. Poor recent form can be a useful warning, but a favourite from a quiet yard still wins a large share of the time. Check that the price is too short for the horse’s real chance before you lay.

What is A/E and why does it matter for lays?

A/E compares a group’s actual winners with the winners expected from the odds they ran at. A value of 1.00 means fair pricing. For a layer, a consistently low A/E over a large sample suggests the market overrates those runners, which is where lay opportunities tend to sit.

Do trainer angles stop working?

They can. Once a pattern is widely known, the market prices it in and the edge shrinks. Review your trainer filters regularly and judge them on logged results, not on how good they looked historically.

Sources