For a good few years now I’ve heard many statements along the lines of ‘Polytrack form works out particularly well at Ascot’. There are variations of it, Tom Segal said “He’s shown good form on Polytrack, a surface whose form works out well at Ascot” while Matt Williams who used to work for the Racing Post had this to offer “Much has changed at Ascot in recent years and since the straight track has relaid, a lot of the emphasis has moved away from stamina, with the speed horses doing especially well. It’s an established policy now, but all-weather specialists (Polytrack) go well at Ascot”. The horse, Pyman’s Theory was named after James Pyman also of the Racing Post, based on his theory that his sire, Exceed and Excel, offspring do very well at Ascot, as they do on Polytrack.
Only last week Jason Weaver on ATR said ” All Weather horses do well at Ascot” when talking about Animal Kingdom’s chances. Angus McNae is quoted with saying “horses with form on a synthetic surface tend to go well at this track, particularly on the straight course”, when talking about Ascot. Surely these experts can’t all be wrong? In my bid to find out, I decided to take a novel approach. I did some research!
Exceed and Excel offspring do do quite well at Ascot, but the sample is not big enough to make some of the wilder claims I’ve read about him. Since the start of 2008 they have a win strike rate of 14.36% on the Polytrack, 11.37% on Turf not including Ascot, and 11.94% at Ascot. There’s nothing in that at all , although it should be noted you would be up money backing his offspring to return a set stake at Ascot, and the in the frame percentage is slightly better at Ascot than other turf courses.
To ascertain if Polytrack specialists do particularly well at Ascot, using data from the start of 2009, I came up with a simple rating system to split horses into either, better on polytrack, or better on turf. I gave horses a score for each of their races based on the percentage of horses they beat, so a winner will get 100%, a last placed finisher 0%, anywhere in between will depend on the position, and amount of runners. I then averaged each horses score for all their Polytrack, and all their Turf runs, prior to a start at Ascot. To make it fair I only used Ascot handicaps, as using group races would have put all weather horses at a disadvantage. I recorded data for all Ascot handicaps, and also split it into two parts, one for the round course and one for the straight.
|Win %||Place %||% Rivals Beaten||Win A/E||Place A/E||Runs|
Poly > Turf means the horse had a higher Polytrack score going into the race than Turf score, vice versa for Turf > Poly. The A/E figures are for Average/Expected, and are arrived at using the Betfair win and place SP’s to get the expected amount of wins and places, and comparing them to the actual. Better than 1.0 means you’d have made money backing them, less than 1 and you would have lost. The first conclusion you might draw from the data above, is that you would have done better backing horses who had previously shown their best from on Polytrack. In the win market this would indeed be the case, although it should be noted that place data will nearly always be more robust, and no advantage was gained there. I did a chi square test on the win data, and with an actual winners of 88, with expected at 78.1, with a sample of 1062, the test showed there was a 24.45% chance the profit was solely due to chance.
The profit/loss figures are good to add some context to the data, but its the % of wins/places/rivals beaten, that will prove or disprove the theory that Ascot suits Polytrack horses more than other turf horses. The win and place strike rates are lower on the straight course, bigger fields is the obvious explanation for this. Looking at the win rates, the Polytrack horses do slightly better than the Turf horses, but for the places the figures are almost identical. For the % of rivals beaten, the turf horses come out fractionally on top, beaten 50.29% of their rivals as opposed to 49.97% for the Polytrack horses. Again there’s nothing in it though. Another conclusion is that the Polytrack horses don’t fare any better versus the turf horses on the straight course, as opposed to the round. This goes against the quote I read saying the straight course is as close as you can get to an all weather course.
There will be some who will look at the above win strike rates, and conclude that Polytrack horses do indeed outperform their turf counterparts at Ascot, the same people probably use win strike rates as proof or otherwise of a trainers form, or a draws suitability. In doing so they are ignoring far to much of the data. With most stats to do with horse racing we need more data, not less. Using places and % of rivals beaten gives our conclusions a much more solid look to them.
As another means of investigating the issue at hand, I also had a look at how Polytrack horses do when running at Turf courses apart from Ascot. The theory being that if Ascot suits them so much, they should do far better there than elsewhere. As before I only used handicaps.
|Turf Not Ascot||Win %||Place %||% Rivals Beaten||Win A/E||Place A/E||Runs|
This time the turf horses had a better win strike rate, as well as place, and % of rivals beaten. The margins are very small however and the only conclusion you could draw from them is that maybe, just maybe Polytrack horses are very slightly better off at Ascot than other turf courses. A more reasonable assumption would be that it’s not a disadvantage, but not enough to claim any significant advantage. As a final bit of research, I decided to see how horses who were placed in a polytrack handicap last time out, did when running in an Ascot handicap, and compared them to horses who achieved the same in a turf handicap last time.
|Ascot Hcp||Win %||Place %||% Rivals Beaten||Win A/E||Place A/E||Runs|
|Poly LTO Str||7.24||26.0||51.78||0.95||1.02||69|
|Poly LTO Rnd||31.57||42.1||55.05||3.63||1.48||19|
|Turf LTO Str||8.97||27.74||55.49||0.98||0.96||602|
|Turf LTO Rnd||12.42||39.39||59.03||1.06||1.13||330|
You might notice the 31.57% win strike rate for horses placed last time on polytrack, now running on the round course, as well as the win A/E figure of 3.63. In reality its based on just 19 runs, with 6 wins, and highlights the dangers of just using win strike rates regardless of sample size. The % of rivals beaten score is a much better guide at 55.05%. Indeed the above table further discredits the assumption that the straight course is the all weather in disguise. The win/place, and % rivals beaten scores, are all lower for the last time out polytrack placed horses, compared to the equivalent turf achievement. The most significant finding is that horses who were placed in a turf handicap last time, do well on the round course.
Given the amount of times I’ve heard variations of the theory, and the degree of certainty with which its been claimed, I’d expect far more concrete proof, and a bigger discrepancy in results than what I’ve found. Based on that, I’d have to say the assertion that ‘Polytrack form works out particularly well at Ascot’ is a myth, or at the very least a gross exaggeration.
If you found this article interesting, please share a link to it, via the social media buttons, your own blog, or forums.