In this article I am once again revisiting one of my favourite areas – the run style of horses, writes Dave Renham. As with the last article I am focusing on the run style profile of specific horses, using data from 2021 initially and then looking at 2022 results up to the time of writing. Initially, though, we will look at a bigger data set in order to set the scene.
For new readers let me explain run style. Essentially, run style is the position a horse takes up very early on in the race, normally within the first furlong or so. Run styles on this website are split into four categories, as follows:
Led (4) – front runners; a horse that takes an early lead (occasionally more than one horse disputes the lead, in which case we can have more than one ‘front runner’); Prominent (3) – horses that race directly behind the early leader(s); Mid Division (2) – horses that race mid pack or 'in touch' with the leaders; Held Up (1) – horses that race at, or near the back of the field early.
The number in brackets is the run style score that is assigned to each category. These numbers are extremely useful and I will again be using them later to create what I call ‘horse run style averages’.
Each race within the Geegeez racecards has a PACE ‘tab’ from where we can view some past run style data for the race in question. The word ‘pace’ is often used as an alternative to run style as the ‘early pace’ shown by each horse determines the position they take up soon after the start of the race. A maximum of the last four runs is shown as we can see below in this race taken from Epsom in July:
The latest run (LR) shows the run style in the most recent race, 2LR in the second most recent race, and so on. The last four runs are totalled for members (Total), and an average figure (Ave) is calculated also. In this race the most likely early leaders according to totals/averages were Toussarok and Pablo Del Pueblo. As it turned out, the opening furlong saw Pablo Del Pueblo lead with Toussarok racing in second place. Concierge, who had by far the lowest run style/pace total, conformed to past run style data also by racing at the back early.
And so it is that these pace/run style figures are useful in trying to determine beforehand how a race is likely to be run early on. They are of course not fool proof, but then, what is? Horses are animals, after all – they cannot tell us how they are going to run - and trainers and jockeys will potentially have an influence also (they, like us, are animals, too!)
If you have read any of my previous run style articles, you will know that early leaders generally have a solid edge in races; and particularly in races over shorter distances. For that reason, I am always looking to find means of improving my ability to predict a race's front runner / early leader.
It's time to give you some stats: as you probably know, we tend to call five- and six-furlong races ‘sprints’ due to the fact they are the shortest two distances in flat racing. Looking at all 5f and 6f races in the UK from 2017 to 2020 gives us the following run style strike rates:
These raw strike rates show a significant run style bias. One in every five wins was for the race early leader, while hold up horses only won around 1 in 16 of such sprint races. In essence, leaders in 5f or 6f races are over three times more likely to win than any individual horse up horse.
If we examine the A/E and IV values too we can see that these correlate strongly with the overall strike rates above:
The bias to front runners occurs in both handicap and non-handicaps over these sprints trips; and, while the win SR% is higher in non-handicaps, this is mainly due to the average field size being smaller. A/E and IV values are strong regardless of race type for these front runners.
So there is the background to the article and hopefully the graph above backs up what I mentioned earlier in terms of always looking for ways to improve our chances of predicting the front runner, more especially in sprint races.
Now to the meat and bones of my latest research where the focus is, not surprisingly, going to be UK races over 5 and 6 furlongs. There are also strong links to my previous article in terms of comparing two recent time frames, as you will see.
My initial dataset looked at all such races in 2021. To begin with I focused on all horses aged three or older that had raced at least four times in sprints. I used 3yos and older horses because I wanted to look at more experienced runners who, I contend, would be more likely to have developed a run style preference.
Horse run style averages
My first port of call was to produce run style averages for each horse, in exactly the same way that I have created run style averages in the past. To achieve this, I added up the Geegeez pace / run style points for each horse over the 2021 season and divided it by the number of races. The higher the average the more prominent the horse tends to race. The averages ranged from 4.00 (horses that led in every race they contested in 2021) to 1.00 (horses that were held up in every race they contested in 2021). Remember, I was exclusively using races over 5 and 6 furlongs to create my averages, so some of these horses may have run over further at some point in the season; those longer races have been ignored in order to allow for a 'pure' sprint dataset.
Once these figures were recorded, I then looked at the 2022 run style data (up to July 8th 2022) so as to create equivalent run style averages for 2022. From there, I wanted to compare the two averages – my hope being that the 2021 run style averages correlated with their 2022 counterparts.
To qualify, each horse needed to have run at least four times in each season. Of course, many of the horses would have run considerably more times, especially in the full 2021 season. In theory, the more the horses run, the more ‘accurate’ their run style average should be in terms of predicting future run style.
488 horses had enough runs in both seasons so it was a decent number of horses to examine. In order to compare the two averages I decided to create six run style/pace average ranges. I used the following groupings - 3.50 to 4.00, 3.00 to 3.49, 2.50 to 2.99, 2.00 to 2.49, 1.50 to 1.99 and 1.00 to 1.49 - then I assigned a letter to each creating six run style ‘categories’ as follows:
Doing it in this way made sense as I felt it was an easier way to compare the data and hopefully easier for you to understand my finding. To begin with I looked at category ‘A’ horses in 2021, those with run style average 3.50 or above, and compared them with their run style category in 2022. There were 30 category ‘A’ horses from 2021 and these are the categories in which they resided as of 8th July 2022:
As we can see 13 of the 30 horses have repeated their extreme front running style of racing in 2022. Eight others are still at the upper end of the run style bracket (category B) with figures of 3.00 to 3.49. Meanwhile, just one of the 30 has averaged under 2.00 this year (categories E / F) meaning just one horse has totally reversed his/her run style from the previous year.
For all that it would have been great to have seen all 30 horses in category ‘A’ for 2022, these remain very pleasing figures.
At the other end of the scale I wanted to look at the performance of genuine hold up horses from 2021 – those in category ‘F’ (average 1.00 to 1.49) to see how their 2022 run style splits compared:
The results are positive once more, with 19 of the 45 horses again landing in the lowest run style category (F). A further 18 are in category ‘E’, the second lowest grouping. These findings so far are the type of figures I was hoping for, but I had expected them also based on previous research in similar areas.
From here I wanted to look at ALL 2021 run style categories, not just ‘A’ and ‘F’, and make a comparison with 2022. For this I wanted to see what percentage of horses from each 2021 category landed either in the same category or the ‘next door’ category in 2022. To help make sense of what I mean by that sentence (as it is a bit ‘wordy’), my findings are in the following table:
This table shows very clearly that the run style of a high percentage of sprinters does not alter that much. It is fascinating to note that horses which primarily race nearer the back than the front (categories E and F – average 1.00 to 1.99) have very high consistency percentages (87.64 and 82.22 respectively).
It is also worth sharing that almost two-thirds (313 of the 488 horses, 64.1%) have 2021 and 2022 run style averages within 0.50, or half a run style grade, of each other; that's another indication of how useful and accurate run style averages can be.
I want to leave you with the horses whose run style averages from 2021 to 2022 (to date) have a difference of just 0.30 or less. We should be fairly confident that this group of horses have a definite run style preference:
|Horse||5-6f 2021||2021 RSA||5-6f 2022||2022 RSA||Diff|
|Grandads Best Girl||5||2.4||5||2.4||0|
|King Of Stars||10||3.8||5||3.8||0|
|Recall The Show||9||2.89||9||2.89||0|
|Our Man In Havana||13||2.31||4||2.25||0.06|
|Come On Girl||10||1.5||7||1.43||0.07|
|Lord Of The Glen||14||2.07||5||2||0.07|
|Elland Road Boy||5||2.4||4||2.5||0.1|
|La Roca Del Fuego||10||3.5||5||3.6||0.1|
|Many A Star||5||2.4||6||2.5||0.1|
|Dark Side Prince||11||3.09||5||3.2||0.11|
|Point Of Woods||8||1.38||4||1.5||0.13|
|Sound Of Iona||16||1.88||6||2||0.13|
|Look Out Louis||6||3.33||5||3.2||0.13|
|The Thin Blue Line||10||2.3||12||2.17||0.13|
|Ey Up It's Maggie||10||3||7||2.86||0.14|
|Miss Nay Never||4||3||7||2.86||0.14|
|Whittle Le Woods||5||3||7||2.86||0.14|
|Miss Bella Brand||11||1.91||4||1.75||0.16|
|Lost My Sock||15||1.87||6||1.67||0.2|
|Eeh Bah Gum||10||3.1||9||2.89||0.21|
|Due A Win||7||2||4||2.25||0.25|
|Isle Of Lismore||8||2||4||1.75||0.25|
|Prince Of Abington||8||1.75||5||2||0.25|
|Hey Ho Let's Go||9||2.67||13||2.92||0.26|
|Show Me A Sunset||15||2.27||5||2||0.27|
There are 201 horses in this list and I am confident they will show a similar running style for the rest of this season. When one of more of these horses contest a five- or six-furlong sprint, the easier it should be to predict how the early run style/pace is going to look. Whilst I have only applied these run style comparisons to 5-6f races, the concept could be easily expanded to horses that race primarily in other distance groups, too.