A Statistical Review of UK & Ireland Flat Racing 2023: Part 2 Trainers

9 min readDec 20, 2023

There were 685 active trainers across racing in the UK & Ireland in 2023, with Tim Easterby the most active with 1295 runners, closely followed by Charlie Johnston with 1243 runners. Joseph O’Brien was the most active trainer of flat racehorses in Ireland, sending out horses for 771 runs, followed by Jessica Harrington with 597 runners.

Andrew Balding had the highest amount of active runners trained out of his yard across all trainers in 2023, training 248 individual runners. Joseph O’Brien was a close second and had the largest string in Ireland with 243 runners.

Note the remarkable rise of George Boughey in the last couple of years, now ranked 12th overall in terms of string-size of all trainers across UK & Ireland.

In terms of average yearling sales value of horses within a trainers’ string, it was Charlie Appleby who unsurprisingly topped the table. On behalf of Godolphin, Appleby trained horses who were on average sold for £568,058 at yearling sales. Note that i am not including homebreds in the calculation here, whom may be even more valuable. Aidan O’Brien trained the most valuable string in Ireland, with commercial yearlings within his yard having an average sales price of £424,022.

Yearling Sales Price correlates incredibly well with Win Strike Rate, which we will see later on. (Throughout this piece i used commercial value of yearlings purchased as a proxy value for the entire string)

Trainer performance is an interesting aspect to qualify. I enjoy the challenge of it, particularly adjusting for the quality of stock a trainer is presented to train and thus the expected performance profile of that stock. In Part 2 of our 2023 Statistical Review series, we’ll look across Trainer Performance in 2023 and see whom the performance metrics point towards as having a good year.

Due to the nature of the handicapping system in racing, annual performance evaluations of trainers can be somewhat challenging. A trainer that had a bumper of a year in 2022 likely ended up with a stable of poorly handicapped horses for 2023. 3 year or 5 year time horizons are actually better indicators of training ability that a single year view. That said, the ability to be able to refresh the portfolio of horses in the yard could also be considered a key skill across the training ranks these days.

Win Performance Metrics

Let’s start by evaluating performance through the lense of Win metrics, before going deeper into some more contextual analysis.

Paddy Twomey topped the table on Trainer Win Strike rates in 2023, returning 29.47%, a continued reinforcement of his strong performance metrics for the last couple of years. This is a remarkable achievement given that the larger field sizes in Ireland compared to the UK make it more difficult for trainers to achieve high strike rates. Through various sources i’ve heard that Twomey is very firm on the horses that he accepts into his yard, with a firm focus on quality the priority. Horses that aren’t deemed good enough aren’t accepted, and those in the yard that aren’t making the grade are swiftly moved on. It’s a training strategy that is paying dividends.

Much was written in the press around the yard form of Charlie Appleby’s Godolphin string this year. Whilst not enjoying the higher grade successes of recent seasons, Appleby delivered the highest win strike rate of trainers in the UK, 28.46%.

Note the close correlation between the Trainer Win Strike Rates table and the Yearling value table referenced earlier in the article.

Looking at the main protagonists we can see that the ratings distribution of their horses is significantly higher than the general racehorse population. The orange line represents the % OR distribution across the entire racehorse population, Blue representing Appleby’s string, Green Twomey, Maroon O’Brien. Elite trainers train elite horses seems like a very obvious statement to make, but i’d pose that a training strategy orientated around a trainer being able to break out of the confines of the handicapping system (making you stop winning) is the best way to build a profile. Easier said than done, and sometimes horses can be reliable types that bring in revenue to pay the bills.

Ratings distribution of the flat racehorse population vs. C Appleby, A O’Brien, P Twomey

Shoutouts to a number of young trainers in the list who maintained high strike rates through the year, with a strength of string perhaps not as valuable as other more considered peers. Ed Bethell and Harriett Bethell had great years; Ed delivering 43 wins from 22 runners for a 19.37% strike rate (Average Yearling Value of 46.37k) and Harriett 13 wins from 75 runners for a 17.33% strike rate (Average Yearling Value of 31.49k).

Jack Jones & Ollie Sangster look like rising additions to the training ranks. Jones ranking 12th on the strike rate table, 18.03% from an average yearling value of 17.38k. Sangster proved a very shrewd purchaser of yearlings at the 2022 sales (more of that later), and delivered a strike rate of 16.46% from an average yearling value of 45.67k.

Comparatively evaluating Trainer Win % and Average Yearling Price the below chart shows the over performers based on price expectation.

Personally i prefer to use %RivalsBeaten² for that type of analysis, so here’s the equivalent view there.

Win Expectancy vs. Field Size & Market

Aidan O’Briens runners topped the performance tables in terms of delivering wins vs. those expected based on field size of races competed in and win probability derived from market price. Returning +64.46 wins v expected due to field size (WOE) and even out performing the market +8.08 WAX. (Amazing for such a high profile trainer)

William Haggas delivered the highest out performance in the UK, +60.81 vs. field size, but the market had his runners found.

Amazingly, the trainer(s) who most out performed the expectations of the betting market was John & Thady Gosden. Returning +13.70 wins compared to the number expected by market price.

Winners to Runners & Stakes Winners to Runners

In the previous section we have mostly been evaluating cumulative wins and overall performance of the string. Now let’s look at the trainers that delivered the highest proportion of individual winners and individual stakes winners from their yard.

The average winners to runners ratio across all horses trained was 35.95% and average return of stakes winners to runners 2.34%.

Ruth Carr was top on Winners to Runners Ratio across her string in 2023, delviering 23 individual winners from 36 individual runners, a %WTR of 63.89%. Twomey also backs up his overall strike rate here, with a %WTR of 58.82%, combined with elite performance in stakes races with 17.65% of horses winning a stakes race. Remember to compare that to the population average of 2.34%…

Credit to Mick Appleby, who had a breakout year at the top level with Big Evs, but continued to deliver consistent performance out of his yard; 54.13% of runners winning a race for owners this year.

The Millman yard often score well on these metrics year on year, and they did so again; returning a winners to runners ratio of 53.49% — 23 individual winners from 43 runners. They do this year on year with inexpensive stock and would be a beneficial addition to any owners’ trainer portfolio.

The table below shows the Top 20 trainers by Winners to Runners % across 2023.

Taking the view using %Stakes Winners to Runners, Aidan O’Brien claims top spot with 23.64% of runners winning a stakes race. Then Twomey, then C Appleby.

Credit to Willie McCreery who had a very strong year at the upper echelons of the sport with 4 stakes winners from 51 individual runners; Insinuendo, Ocean Jewel, Self Belief and Vespertilio.

The Value Table

The Value Table is a league table that we use to display which trainers are delivering higher performance than the price of the stock in their yard would expect.

In our value table, William Haggas was the top performing trainer across the UK & Ireland. Our model expected his string to deliver 91.93 wins over the course of 2023 and they delivered 141 race victories, an over performance of +49.07 wins. Ralph Beckett enjoyed a very successful 2023 and delivereed +42.59 wins above expectation vs. our price model.

Adjusting per 100 runners tells the story with a standardised set of samples. Here, Twomey is the king again, delivering +18.84 wins per 100 runs vs expectation. Remember that he wins at nearly a 30% strike rate, whereas our model expects him to win at an 11% strike rate in Ireland. That out performance is over double that of the nearest ranked trainer, Peter Chapple-Hyam at +9.90 wins per 100. Ed Bethell shows up well again in third with +9.62 wins per 100 runs.

The Justice Table

A binary performance sport like horse racing; multi competitors and a single winner per race is a perfect case for hard lucky stories. Horses can run well without getting their just rewards over a season that often spans a small volume of run samples per individual. That’s why we use the Justice Table to evaluate the wins that underlying performances should have returned. This can help us to identify which trainers have been unlucky this year and should be sitting higher in the standings.

Here are the trainers that old man justice deems deserving of better statistics than the raw win figures suggest.

Across 1,243 runs, Charlie Johnston horses beat 4866 other rivals home. His runners return 173 wins vs. 156.64 expected based on the field sizes of their races. That said, our model feels that the Johnston yard should be further rewarded statistically based on the underlying qualify of these performances; actual Win% return was 13.92% and our model feels the horses beaten across races, combined with the field size of those races means that the Johnston horses deserved a strike rate of 17.01% and 211.44 wins returned. The luck gods under-rewarded the yard by -38.44 wins over the course of the season.

As mentioned, Ralph Beckett had a fantastic season. But it could have been even better; his horses were running well all year and should have returned a 25.96% strike rate based on overall performances, yet only netted 20.43%. He was -35.99 wins under-rewarded by the racing gods.