Chases, Jumping, Falling: An Analysis

In this article I am taking a one-off look at chases and, in particular, looking at a key factor for any horse that runs in such races, namely jumping, writes Dave Renham. I am looking at data from 2015 to 2022 which covers eight full seasons and I am looking at both UK and Irish results. The main aim for this article is to try and have a better appreciation of what factors impinge on a horses’ jumping. That might be racecourse related, it might be going related, it might be a combination of these and other factors. Any profit / loss figures have been calculated by using Betfair Starting Price (BSP).

In 2015, both Tony Keenan and the editor Matt Bisogno produced some research in this area - which you can read here and here - and this article will both build on the earlier work and bring it up to date.

I think most racegoers will agree that being able to jump fences is a key requisite to be successful in chase races. Imagine a 3 mile chase where two horses of similar ability race against each other. If horse A jumps cleanly all the way and horse B makes numerous jumping errors, one would expect horse A to win a very high percentage of the time.

The difficulty of jumping fences varies: some courses have more challenging obstacles – the Grand National course at Aintree springs to mind. However, the ‘big’ tracks of Ascot, Cheltenham, Kempton, Haydock, Newbury and Sandown also have more challenging obstacles than the majority of smaller tracks. Clearly it should be more difficult to jump ‘stiffer’ fences, but there are other considerations to take into account, for example, the state of the grass either side of the fence. If the turf is slippery or slightly worn, then regardless of how difficult the fence is to jump, landing cleanly can become more problematic.

We also need to consider that some tracks have more fences than others at certain distances. Did you know, for instance, that the minimum number of fences per two miles is lower at Irish courses than it is at UK ones? As an example of this let us look at two racecourses, one in Ireland (Naas) and one in the UK (Wetherby); and observe the difference when it comes to a two mile chase. Naas first:

 

 

You can see that, for two mile races, horses are required to jump two fences in the home straight on the first circuit and then eight fences on the second circuit, giving a total of 10.

Let’s compare to Wetherby now:

 

 

Wetherby’s shortest chase distance is actually a furlong less at 1m 7f, but they have to negotiate 13 fences in total; four on the first circuit, nine on the second.

Also, in terms of number of fences, one could argue that courses with a higher concentration of fences closer to the finish might prove more problematic than courses with fewer fences at the business end. Another factor to consider is race type: some courses have more handicap chases than others; some have more races for novices etc. All these factors will play a role of some sort.

As will be becoming clear, there is plenty to consider here. My initial starting point from a stats perspective is to look at individual courses and which are hardest to jump round. To do this I am simply taking the number of chasers at each track and finding the percentage of runners that either fell or were unseated. I have ignored horses that were brought down; I am not saying this is the best or only method to test out which courses are the hardest to jump round, but it is logical, and horses that fall/unseat would have almost certainly produced a jumping error of some sort. Of course the fall may not necessarily be down to poor jumping, it could be down to jockey error. Alternatively it may be down to something I alluded to earlier - landing on a slippery or churned up section of ground. However, the majority of casualties will be down to a mistake / poor jump.

What is so important when doing this type of research is to appreciate that there is often no right or wrong way, no right or wrong answer. Also one needs to be aware that there are potential flaws in any idea/method and, where possible, one should try to address them when sharing the relevant data.

OK, back to the percentages of horses that fell / unseated by course. The UK first:

 

 

Aintree tops the list, as I am guessing most expected, and as it did in 2015. The Aintree stats include both chase courses, Mildmay and Grand National, but it will come as no surprise that for the National course the percentage increases to 23% in terms of fallers/unseated riders. For the record it is 24% for the Grand National itself (over the past eight years).

But would you have expected Fakenham to be second in the list? Or indeed Ludlow third? Now I have already mentioned other factors that are potentially in play here, but there are other points to consider as well as these. For example, class of race / horses differ from course to course – that could affect the percentages, as can the speed at which horses approach the fences. The going is another factor as generally it is harder to complete the course when the underfoot is soft or heavy as compared to firmer conditions. Hence courses that have a greater percentage of chase races on softer ground will see the percentages slightly skewed. Races that are run at a quicker pace will probably see more fallers, etc, etc. This is another example of what I meant earlier about being aware of potential flaws in any research of this type. The individual course plays a major role in how easy or hard it is for a horse to jump round, but there are more pieces to this complex puzzle.

Let us compare the above with Irish course data now:

 

 

It is not surprising to see Leopardstown near the top, but I wasn’t expecting Listowel to be at the head of the list. What is interesting is if we compare the percentage figures for all UK courses to all Irish courses.

 

 

As the bar chart shows, in percentage terms far more horses have either fallen or been unseated in Irish chases as compared to UK ones. One potential reason for this is that Irish races have bigger fields on average. Hence could there be more crowding / less space at certain fences, which could cause horses to make more bad errors that are bad enough to see the horse come down? That’s certain plausible.

Could this be partly down to the going? The reason I posit this is that in Ireland 40% of the chases were raced on ground described as soft or heavy, in the UK the figure was 34.8%. Hence, it seems now would be a good time to compare the fell/unseated percentages by going. The most obvious thing to compare is soft or heavy versus good to soft or firmer. Here are the figures:

 

 

Certainly there are more jockeys falling when the conditions are soft or heavy. It may look like a small difference, but due to these figures coming from thousands of races it is almost certainly a material factor. Further, when isolating races on good to firm or firmer only, the fall/unseat percentage drops to 6.2%. This is more evidence that supports the theory.

Let's now go back and dig deeper into the differences with the UK and the Irish data. Earlier we saw that horses in Irish races fell or unseated more often than in UK races, as a result of which I had a theory that any horse that fell or unseated last time out in a chase would perform better next time in the same race type if their mishap occurred in Ireland. The thinking was simple: for whatever reason jumping fences in Ireland appears harder, so horses that had their mishap that side of the Irish Sea may have had more of an excuse than their UK counterparts. Let us see if the stats back up this hypothesis:

 

 

OK, so this is the reverse of what I was hoping for! What do I know??! However, the differences at first glance do still look quite powerful.

We need to be aware that punters, as a rule, take last time out performance as a key factor when contemplating any bet. Horses that fall/unseat do not generally get positive attention from punters. Hence an obscure reason of which country the horse fell in a chase last time out will go under the radar of most people.

Not surprisingly, if you flip the idea slightly and just concentrate on which country the chase race was run (after falling/unseating), the stats are remarkable similar:

 

 

The stats are virtually the same because most horses do not hop back and forth racing in Ireland one week, then next time in the UK, or vice versa. A few of the better horses do switch occasionally especially for the big festivals or races. However, invariably not many of these horses would have fallen or unseated last time.

Now, of course, these stats may be skewed slightly by the fact that, as I mentioned earlier, Irish chases have bigger average field sizes. This certainly explains the difference in strike rates, but it does not necessarily account for the differences in profit/loss; or, more importantly, the significant difference between the A/E indices. However, once again, when analysing data we need to be aware of any extenuating circumstances that may affect our stats. Also, we need to appreciate that sometimes we simply cannot explain the results, or at least it may be difficult to offer a logical reason why they have turned out the way they have.

So what about distance: is it a case of the longer the distance the more fallers/unseated riders? That makes possible sense as a longer distance means more fences to jump. However, it seems to be that longer distances actually see slightly fewer horses falling or unseating their riders thank shorter trips in percentage terms.

 

 

So what could be happening here? A logical explanation is that shorter races are run at a quicker pace; a quicker pace potentially offers up scope for more, or more serious, jumping errors. However, whatever the reason, I must admit I had hypothesised beforehand that longer races would have had the slightly higher percentage.

I mentioned earlier that race type will have a potential effect on the figures so let's review this. One might expect less experienced chasers (i.e. novices) to make more jumping errors than seasoned handicappers, and therefore fall / unseat more often. Here is a look at the fell/unseated percentages across different chase types:

 

 

The stats definitely back up the theory; novice and beginner chases have the highest figures; handicaps have the lowest.

Something I did notice was that in 2022 Beginner Chases were restricted to Irish courses only. No Beginner Chases in the UK that year. I have since noted that at the end of Jan this year (2023) there was a Beginners Chase at Lingfield. I am not quite sure what happened in 2022, maybe Matt knows [I don't, but the fact that there has only been one double-digit field for a UK beginners' chase since 2015 ought to be a part of the reason- Matt]

What about jockeys? They must play some sort of role in all this. I am guessing some jockeys are slightly more adept at getting their horse to jump fences; likewise some jockeys may have a better ability to stay in the saddle when others might tip over. Here are a list of jockeys with their fell/unseated percentages – it gives their overall figure, their figure in handicaps (exc. Novice handicaps) and their figure in ‘all other’ races. To qualify they must have ridden in the last year and had at least 300 rides in handicaps, as well as 300 + rides in other chases combined. They are ordered alphabetical.

 

 

For the record, the average figure for fallers/unseated for all jockeys in all races (and indeed all horses) is 7.1%. Several jockeys in the list have an overall percentage of under 5% which I count as a positive - I have highlighted them in green. Just one jockey has a figure over 10% - that is Luke Dempsey who not surprisingly rides in Ireland. Indeed three of the four jockeys with the highest figures primarily race in Ireland (Blackmore, Dempsey and Kennedy). Brendan Powell has the worst record as far as a UK jockey is concerned, so you might be wary in the future if he in on board a suspect jumper.

Another interesting finding is that the more experienced jockeys in chases fell or unseated less often. Jockeys that had ridden in 1000 chases or more (2015-2022) had a percentage fall/unseat rate of 5.7%; jockeys who had ridden 500 to 999 chases had a percentage of 7.1; jockeys who had ridden in less than 500 had a percentage of 7.9%.

 

MAIN TAKEWAYS

 - A bigger percentage of horses racing in chases fall or unseat in Ireland

- A bigger percentage of horses racing in chases fall or unseat on soft or heavy ground

- A bigger percentage of horses racing in chases fall or unseat in Novice chases/Beginner chases

-  A bigger percentage of horses racing in chases fall or unseat when racing on the National course at Aintree

- A bigger percentage of horses racing in chases fall or unseat in races of under 3 miles compared to chases of 3 miles or more

- Horses that fell LTO in the UK have appeared to perform better next time than their Irish counterparts

- Sean Bowen, Paddy Brennan, Brian Hughes, Denis O’Regan, Sean Quinlan and Sam Twiston-Davies are the six jockeys with the lowest fall/unseat rate of the more experienced jockeys

- More experienced jockeys tend to have a lower fell/unseated percentage than less experienced ones

 

To conclude, I hope you have found this an interesting read. It is certainly a different aspect of racing to what I normally write about. I have only touched on a small part of this whole jumping idea, and if enough people comment they would like to find out more, then I’ll happily pen another piece sometime in the future.

Other Recent Posts by This Author:

• 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
• Early Odds to Opening Odds to Starting Price