Cheltenham Festival: The 15 year View

It is almost that time! For many, the Cheltenham Festival is the highlight of not just the National Hunt season, but the whole racing year, writes Dave Renham.

In this article I will attempt to break down the facts and figures going back as far as 2008. This gives us 15 years’ worth of data to crunch, which is plenty to get our teeth into. Fifteen is also a neat number as we can easily compare 5–year periods (2008–2012; 2013–2017; 2018–2022) to see what, if anything has changed.

My main focus will be looking at the data as a whole – market factors, last time out (LTO) factors, etc. At the end I will delve briefly into Grade 1 contests only. In terms of profit and loss, I am going to use Betfair Starting Price, and take into account commission on potential profits.

Cheltenham Festival Stats for All Races

Since 2016 there have been 28 races in total at each year's Cheltenham Festival and that will be the same in 2023: four days with seven races on each day, and a rich variety of different race types and distances.

 

Market Factors

Let's first examine the results by market price. Although I am quoting profits/losses to BSP, the market price bands I am examining are based on Industry SPs. This is simply because we have more defined prices for this group:

 

The Evens to 9/4 bracket has proved the most profitable in ROI terms and, taking shorter priced runners as a whole, the market has been a pretty good guide. Combining all runners priced 6/1 or shorter we have seen 182 winners from 807 (SR 22.6%) for a small BSP loss of £7.42 (ROI –0.9%).

Despite these relatively positive figures, there are strong fluctuations year on year as the graph below shows.

 

 

As you can see, the win percentage / strike rate peaked in 2016 at 33.33%, whereas 2010, 2014 and 2017 saw percentages dip under half that figure. Eight of the years would have turned a profit, seven a loss. Hence one needs to be aware that results for runners priced 6/1 or shorter are difficult to predict for a one–off Festival, 28 races always being a small sample size. However, having said that, taking the overall data into account, one could do worse than focusing attention on this price band.

 

Performance of Favourites at the Cheltenham Festival (2008-2022)

A look at favourites next. Taking all favourites as a group (clear and joint favourites), they have secured 113 wins from 443 races (SR 25.5%). However, backing all of them would have returned £43.12 less than staked, equating to a loss of nearly 10 pence in the £. Before ‘binning’ favourites as a betting option though, let me share the stats comparing clear favourites with joint favourites:

 

 

There is quite a difference here! Of course, it is sometimes difficult to predict who the favourite will be pre-race which can be an issue for trying to exploit ‘market data’. However, as a general rule, the stronger the favourite the better. What I mean by that is, horses who are a much shorter price than the second horse in the betting tend to do best here at the festival. AND of course this type of favourite can be confidently predicted before the off.

For favourite fans here are a few profitable angles in relation to clear market leaders:

 

 

The best figures come from horses aged 7 to 9: we do need to be careful bracketing runners by age, in case there is back-fitting in play. However, this age bracket of runner is around the optimum age for high level jump racing and much is known about such runners. By that I mean we usually have detailed form lines for runners within this age bracket. Of course there are races at the festival where 7 to 9yos do not take part, but in the races they do, if any clear favourite is in this age band, I believe it demands close scrutiny.

We will examine Irish trainers versus UK trainers in more detail later, but Irish-trained clear favourites have done well. If we combine the clear favourite records of Willie Mullins, Gordon Elliott and Henry De Bromhead, 39.1% of them won (54 wins from 138) for a profit of £27.25 (ROI +19.7%). The ‘rise’ of Irish runners will be a theme of this piece, and this can be seen when we look at a year by year breakdown of clear favourites that were trained in Ireland.

 

 

The graph illustrates a clear upwards trajectory with the last four years averaging out at just under 20 per meeting (19.5 to be precise). Essentially this means that around 70% of all races in the past four years have had an Irish-trained favourite. Compare this to the first five years (2008 to 2012) where the average was 7.4.

 

Last Time Out (LTO) performance

Cheltenham Festival Performance by LTO finishing position (2008-2022)

Onto last time out factors now with my initial focus being on where a horse finished in its most recent race:

 

 

Although horses that either finished 3^rd LTO or 5^th or worse have made a profit, this is down to big prices skewing the figures. As we can see, strike rates are low across the board, but if there is an area to concentrate on, it does seem to be last day winners. This is because they are the biggest group, have by far the best record win wise, and they have just about broken even.

Earlier I noted that LTO winners that went on to head the market at the festival have proved profitable. What about other areas?

 

Performance of LTO winners by Gender of Horse 

I want to share some gender data with you in terms of the gender of horse. Male LTO winners make up around 90% of all such runners, but female LTO winners have comfortably outperformed their male counterparts at the festival in terms of strike rate:

 

 

194 female LTO winners have run at the festival with 31 winning. Not only that, if you had backed all of them ‘blind’ they would have secured you a profit of £116.24 (ROI 59.2%). If sticking to solely mixed sex races (races open to both sexes) the stats, albeit from a small sample, are even more impressive: 13 wins from 69 (SR 18.8%) for a profit of £116.14 (ROI +168.3%). Indeed, looking at the last three festivals (2020, 2021 and 2022) LTO winning female horses running in mixed sex races have won 8 races from just 20 runners!

 

Performance of LTO winners by LTO Race Class

Time to examine whether the class of race that the horse won last time out makes a difference... and it certainly seems to!

 

 

Horses that won a Grade 1 contest LTO have scored close to one race in every four which is impressive. Backing all runners would have yielded a good profit also of over 22p in the £. Horses winning LTO in either Grade 2, 3 or Listed company have very similar strike rates, but it is Listed LTO winners who have created the best profit (£49.48 returning 41p in the £).

LTO winners outside Graded and Listed company have by far the poorest strike rate as you would expect. They have incurred losses of £116.97 (ROI -8.3%) over the period of study. LTO winners outside Graded and Listed company have not surprisingly struggled even more when the race at Cheltenham is a Graded one – in these races their record reads 50 wins from 957 (SR 5.2%) for a loss of £149.46 (ROI -15.6%). Losses have been steepest in Grade 1 contests with your £1 bet returning on average 79p (loss of 21p in the £).

 

Performance of LTO winners by LTO Course

The next question I will try and address is, does the track at which the horse ran LTO make a difference? One might expect that horses that ran at a top track last time would outperform those that didn’t. The table below looks at any course that has sent 75 or more runners next time out to the Cheltenham Festival. I have ordered it by win strike rate:

 

 

What immediately resonates is the record of Irish tracks: the top four in the list are all Irish tracks and runners from all four (Thurles, Leopardstown, Naas and Navan) have secured decent profits at the festival to BSP. Irish tracks also take positions 6 and 7, giving them six of the first seven spots in the list. Focusing on those top four courses, here is win strike rate breakdown by 5-year groups:

 

 

The last decade has seen a notable uptick in performance which mirrors the type of pattern we saw earlier in terms of the increasing number of Irish runners that have started clear favourite. In that favourite data, the years 2008 to 2012 saw the smallest market leader numbers by some margin. Of course, we know about the dominance of Irish winners at recent Festivals but there is still plenty on which to chew in relation to possible value edges.

Before moving on, any system punters out there may want to consider an angle based on last time out runners from these four Irish tracks. It combines some positives we have already noted and is as follows:

1. LTO run at Leopardstown, Naas, Navan or Thurles
2. LTO run in Graded / Listed Race
3. Finished in first three LTO

The results were:

 

Ten of the 15 years would have yielded a profit, and a very good one in nine of those ten positive renewals. Three years made small losses, two years quite big losses.

Sceptics will naturally be highlighting the fact that this system idea is back-fitted, and to a great extent they would be right. However, the rules are simple, logical, and there are not many of them, all of which is positive from a system building perspective.

I am definitely not advocating that this system is one that punters should use ‘blind’ at the 2023 festival, but it may offer a potential starting point, to at least give you a pool of runners to consider. Also, for readers with little time to study form, I am confident there are plenty of systems around that are less likely to produce a profit at the Festival than this one.

 

Irish runners versus UK runners

We have already noted some positives connected with Irish runners or those that raced in Ireland last time. It goes without saying that the vast majority of horses racing at Cheltenham that raced in Ireland last time out would have been from Irish stables; in fact 97% of them were. Hence there definitely has been a strong Irish bias.

Below is a breakdown of the records of all UK trainers versus all Irish trainers:

 

 

Looking at this, we can the Irish bias in all its glory. Irish-trained runners have more than twice the strike rate of their counterparts trained in UK. Moreover, they've enjoyed a 55p in the £ difference in their returns, and a clear differential between the A/E indices.

In recent years their stranglehold has got stronger and stronger. Below shows the number of Irish wins by 5-year groups:

 

 

These figures are skewed inasmuch as the last five years have seen a big increase in the number of Irish horses travelling across. However, the win strike rate for Irish runners in the five years from 2008 to 2012 was 6.8%, whereas in the past five years (2018 to 2022) it has been 9.7%. So the Irish are sending more runners than they did more than a decade ago, and are winning on average more often. That, obviously, is a potent combination.

Indeed, Irish runners have outperformed UK runners in terms of win strike rate in the last ten festivals starting from 2013 as the graph below neatly illustrates.

 

 

The UK runners did close the gap in 2022, after a dreadful 2021. Will they be able to get any closer this year? Only time will tell, but you have to expect the Irish to come out on top overall once more.

 

Grade 1 Races

For the final segment of this article I want to have a brief look at Grade 1 races. These races comprise 50% of the 28 Festival contests and, in the last 15 years, they have accounted for roughly the same percentage of all the Festival contests (some of the newer races being upgraded during the review period).

The betting market comes under the spotlight first.

 

Market Factors in Cheltenham Festival Grade 1 Races

I have split the prices as I did earlier in the article and here are the Grade 1 only figures:

 

 

The data show a poor record for odds-on runners, but in general short- to mid-range prices do quite well. The cut off price looks to be at 14/1 – at this price and bigger Grade 1 runners have performed poorly. Strike rates are below what is the 14/1+ norm for all National Hunt races and losses have been significant.

If we look at market position data instead, clear favourites in Grade 1 races have just edged into profit, albeit by only £6.77 (ROI +3.6%); backing ALL runners in the top four in the betting would have yielded a profit of £55.24 (ROI +6.8%).

 

LTO performance in Cheltenham Festival Grade 1 Races

One group of runners to avoid in Grade 1 races seems to be those that ran relatively modestly or poorly last time out. Horses that finished 5^th or worse on their prep run have accounted for just eight winners from 282 runners (SR 2.8%) for a hefty BSP loss of £129.01 (ROI -45.8%). Meanwhile, last day winners have secured 141 wins from 1200 runners (SR 11.8%). They, too, made a loss but nowhere near as severe, at -£50.53 (ROI -4.2%).

 

LTO Race Class

A look next at race class on their previous start (all Cheltenham Grade 1 runners).

 

 

There is a sliding scale of strike rates as you would expect. Horses that raced outside Graded/Listed company have a poor record.

If we focus only on LTO winners, it is interesting that each LTO Graded category made a small individual profit to BSP, as did those who won a Listed contest.

 

LTO course

In terms of the course Grade 1 Cheltenham Festival entries ran at last time, Irish courses have again outperformed UK ones. This time around I have grouped all courses in each country for the comparison:

 

 

It is no surprise to see horses that ran in Ireland LTO coming out on top in terms of strike rate, returns and A/E indices. There is, however, one Irish course where caution might be advised, and that is Gowran Park. Just 2 winners from 90 runners in the last 15 years prepped there, with losses amounting to over 88p in the £.

 

Gender of Horse (LTO winners only)

We saw earlier that LTO winners that were female had a better strike rate than males, as well as proving profitable. Focusing on Grade 1 races only, this pattern is replicated once more:

 

 

A strike rate of close to 1 in 5 is excellent and female LTO winners have secured a profit in Grade 1 races of £66.94 (ROI +85.8%). Hence any female running this year at Cheltenham who won last time out might be a horse to consider as a betting opportunity.

 

Irish runners versus UK runners

It is abundantly clear from what we have seen to date that, in general, Irish-trained runners outperform those trained in the UK at the Cheltenham Festival. From the LTO course (by country) figures we can see that this is also the case in Grade 1 races (as most of the runners that ran in Ireland last time are Irish-based). What I would like to share is the number of Irish wins in Grade 1 races broken down by year:

 

 

The last ten years (from 2013 onwards) have seen Irish runners dominate these events more and more. Indeed, in the last two years we have witnessed double figure victories and, considering there were only 14 Grade 1 races in each of 2021 and 2022, this is mightily impressive (or concerning, if you're a British-based racing administrator, trainer or owner). To spell it out, in the most recent two Cheltenham Festivals, Irish runners have secured 22 wins compared with just six for the UK.

 

Key Takeaways

Before winding down, here are some of the key stats I suggest you keep in mind:

1. The betting market is a good guide. Clear favourites are reasonable value in all races including Grade 1 contests. Focusing attention on horses priced 6/1 or shorter should give a sporting chance of making a profit. In Grade 1 races avoiding horses priced 14/1 or bigger will usually save you some cash.
2. Irish runners are likely to outperform their UK counterparts. This is especially probable in Grade 1 races. The trainers Henry de Bromhead, Gordon Elliott and Willie Mullins have good records with favourites.
3. Female horses have a good record when following up a win last time. This is true even in Grade 1 contests.
4. A prep run at Leopardstown, Naas, Navan or Thurles has provided good profits over the past 15 years.
5. Last time out Grade 1 winners are generally decent value.
6. In Grade 1 races it looks best to avoid horses that finished 5^th or worse in their final prep.

 

Hopefully this article has offered some good general guidance to follow, with the hope that it will find a winner or two along the way. This is my last article before Cheltenham, so good luck, and I'll see you on the other side with some early thoughts for the 2023 flat campaign!

- Dave R

Other Recent Posts by This Author:

• Price Movement in NH Racing Markets
• More on Price Movement from Opening Show
• A Study of Owners in UK Flat Racing, Part 2
• A Study of Owners in UK Flat Racing, Part 1
• Seeking Value with Female Flat Jockeys