Winter is coming
About a week ago I was thinking about topics to turn my hand to, playing with possible angles and messing around with data, writes Jon Shenton. Given that winter is coming™ I was looking through general National Hunt stats and I stumbled across the below potentially revelatory information.
The data is sorted by ROI at SP and only includes trainers that had 100 or more runners over fences in the review period. Summarising all of the data, that’s 4890 runs in total with a decent £812 profit at a £1 level stake - a return of 17%. To be honest I pore over data constantly and I can’t remember seeing anything this blatant, immediate and maybe obvious before, in fact I’m kicking myself I hadn’t clocked it previously. My only partial defence is that I have a slightly odd penchant for a class 6 Southwell AW handicap as a betting medium, clearly I should widen my focus to the chasing game!
Just before we plough on, let me explain the 20/1 upper limit choice: in basic terms, when looking for angles, quite often you can find what appear to be mightily profitable paper systems, but they are skewed by one or two big winners. Now, one could reasonably argue that they are a component of the angle, however, I prefer to look for good solid reliability and hopefully sustainable profit, especially when trying to get a handle on a large data set such as ”all chase races” as we are reviewing in this article.
Back to the data, there are plenty of familiar names involved and also as you might expect, plenty of smaller yards potentially flying under the radar. In theory backing all runners from all of the yards systematically may yield a return but underneath this high-level view there should be interesting pockets of gold where a sharpened focus could improve the yield.
So how can we zoom in? Well we could just analyse each trainer individually and sequentially, hunting for micro angles. Maybe if we have unlimited time that would work but I have deadlines to meet, bills to pay, a family to talk to and a day job to think about.
I think in this case a quick look at consistency of performance over each calendar year may give a better view of the trainers to follow in the chase field. The table below gives that summary, blatantly and unashamedly ripping off an Instant Expert type view to assist in landing the individual nuances in the data.
First thing to note, 2018 as a whole so far is certainly disappointing in comparison to previous years. It may be seasonality, i.e. we are only part way through 2018 and perhaps a number of these trainers hit their straps in Q4 so there could be juicy profit to be had over the next few months. It may be that the market is now wising up to the yards’ chasing proficiency. It may be reversion to the mean. It may be that the dataset was happenstance in spite of the 100+ runners stipulation. Or It may be something else entirely!
The original reason for pulling that view of the data was to find trainers upon which to focus; to look for those with a high degree of consistency both in strike rate and profit. We’re not really interested in flash in the pan performance: for instance, we can see that Kerry Lee makes the list purely on the back of a Hollywood 2016.
All of the trainers in the table with a comment of either “Consistent” or “Improving” are worth prioritising in the first instance but, for the sake of brevity, I’m going to concentrate on three: Dai Burchell at the top of the list, Michael Scudamore and Nigel Hawke. Perhaps I’ll visit some of the others in a future edition!
I have a confession to make. Until starting this research I had genuinely never heard of, knowingly backed, or even looked at a Dai Burchell horse. I follow the flat more closely than the jumps – that’s my excuse – but even so! Anyway, for those in the same hitherto ignorant boat as me, the yard is based in Ebbw Vale, has around a dozen horses in training currently and has been operating since 1983. Dai and I have something in common too: apparently, we both enjoying walking around castles, perhaps our paths have crossed before unwittingly.
A keenness to learn more about Burchell reveals he is clearly a chase specialist, his hurdle strike rate being less than half as good. The yard’s sole bumper runner barely merits mention.
As I’ve mentioned in a previous article one of the first things I check where a trainer is concerned is how their fleet perform after a layoff. The table below shows Burchell’s chase runners split by days since last run.
If a Burchell horse hasn’t had a run in 76 days then it doesn’t win. Granted there are only 18 of them so not a reliable number but the fact that the place percentage is also lower than the rest of the sample I’m happy enough to ignore it, leaving us with the following.
A 24%-win rate and 78% ROI by backing all Dai Burchell chase runners if they’re fit (i.e. ran in the previous 75 days). Ordinarily I’d take that any day of the week, but there is a very substantial elephant in the room that needs addressing. This is a small yard and this seemingly bulletproof performance is down to a small number of horses winning multiple times.
In fact, if we look at the 34 winners we have only 11 unique horses accounting for them, with such luminaries as Rebecca’s Choice, Ratify and One For The Boss getting their noses in front a total of 15 times between them. There have been only 16 individual horses running for the yard since 2012 so to have 11 different winners is a strong performance, but even so it does make the data evidence slightly less compelling.
Having said that, what we know for certain is that Dai Burchell has a small, highly effective unit of chasers, who can win on a regular basis. They’ll certainly be going in my tracker or angles to flag each time they run for deadly serious consideration.
Let’s look at one of the more mainstream names, the Welsh National and Cheltenham Festival-winning trainer MJ Scudamore, who has horses such as Mysteree and Kingswell Theatre associated with the yard. My eye was drawn to this trainer on account of a profitable return in each of the last four years – significantly so in each of the last three years – and look to be on an upward curve, at least at first glance.
As with Dai Burchell, the record of Scudamore runners in chases is far superior to hurdles and NH Flat: his hurdle strike rate is 9%, less than half of the chase equivalent. The hurdlers also have an ROI of -60% compared to the 27% profit on chase runners.
Digging deeper into the record over fences there is a great deal of consistency across the usual factors I’d initially use to home in (days since last run, race class, horse sex, age etc). There is, however, one area which merits attention.
The above graph clearly shows a fallow period through the summer jumps period, which is perfectly understandable: the yard keeps ticking over, but the focus is seemingly to gear up for the main season over the winter months. The strike rate trend is mirrored by profitability, April to September runners collectively delivering a yield of -9%.
If we take Scudamore chase runners from October to March we’re left with a nice return of 64% at SP (78% Betfair Exchange). The yard certainly has a bigger scale than Burchell with 46 horses in training, according to scudamoreracing.com.
Again, the 42 winners include multiples, with 18 unique horses delivering them.
One final thought on this section relates to recency bias. Very often when churning through data (trainer data especially) one check I do is to differentiate between whether a horse placed last time out. For some trainers form of the horse can be less relevant, they are better at bouncing back from more moderate runs (maybe as they drop down the handicap). Below shows the differential for the Scudamore chasers through the winter months.
No discernible difference in strike rate but the P&L is much healthier for horses that didn’t hit the frame in their previous run. The average SP for a winning horse placing LTO is a shade over 10/3, whereas for those off the board on their most recent run, the winners return an average 15/2 SP. Recency bias in numbers? I hope to review this in more detail over the coming months.
Recency bias or not, don’t be put off by a recent moderate run by a Scudamore chaser.
The final trainer subject to a more detailed review is Nigel Hawke, who is based in Tiverton and has a team of nearly 30 horses in training. Whenever I hear or think about Nigel Hawke the first thing that springs to mind is geegeez.co.uk’s Trainer Snippets report, specifically with regard to last time out winners. Hawke’s name seems to be a permanent fixture on there! So, first port of call is to check that and, given his general strong record over fences, I’d expect to see a nice angle looking specifically at LTOW’s.
Sure enough, the theory holds water: a small number of runners but solid enough. As a slightly off-topic bonus, below is the record of Hawke’s LTO winners running over hurdles over the same period, a very similar story. Conclusion: Nigel Hawke last time out winners are worth backing irrespective of obstacle type
Back to Hawke in general chases, here is his record by class of race
As can be seen, Class 5 – the basement level in National Hunt – race performance is fairly wretched by comparison to all other classes. I’m always comfortable taking these factors into account when it’s either at the elite end or the opposite (let’s call it grassroots) levels.
By focusing on Class 1 to 4 races only we are left with 46 wins from 212 runs (21%) and a starting price ROI of 44%, the 46 wins delivered by 19 different animals.
There are a number of generally profitable trainers over fences in the review period and using a quick check of consistency to zoom in on particular yards, I would flag three angles worth putting in the notebook for the winter months:
- Dai Burchell chase runners if they’ve had a recent run (in the last 75 days)
- Michael Scudamore chase runners running from October through to March, focusing on those that had a moderate last run (didn’t place)
- Nigel Hawke’s chase runners in Class 1-4, and all last time out NH winners
- Jon Shenton