What Factors Make The Most Impact On Finding Horse Racing Winners?
This is arguably the most complex question anyone can ask when it comes to horse racing. There are dozens of potentially relevant factors and hundreds of irrelevant ones when it come to horse racing.
Commonly used factors include:
- Form
- OR
- Course & Distance
- Going
- Speed
- Weight
- Trainer/Jockey
- Class
- Days Since Last Ran
- Odds
And many more! In fact, Iāve written extensively on most of these factors in āHow to Choose the Right Factor Every Time in Horse racingā.
Today, Iām going to focus on Going, Class, OR and Days Since Last Ran (DSLR) and use Pool Impact Value (PIV) calculations to discover if these factors are worth considering when trying to find winners. Before reading on, I would recommend reading Michaelās articles on calculating Impact Value and Pool Impact Value. They are relatively short pieces but give you a clear insight into why you should use these calculations in the first place.
In a nutshell, youāre looking for a PIV figure of above 1 as this means horses are winning more than what you would expect on average.
For the record, all data comes from 2012-2016.
Going
Going is often believed to be one of the most crucial factors in horse racing. To make things interesting, I looked at the performance of horses that ran their last race under the same official āGoingā description. So under āFirmā for example, this relates to horses that also ran on Firm ground in their previous race.
Going | Bets | Wins | Strike Rate | PIV |
Heavy | 8879 | 1100 | 12.39% | 0.85 |
Good | 41047 | 4537 | 11.05% | 0.86 |
Good to Firm | 20508 | 2424 | 11.82% | 0.84 |
Good to Soft | 11862 | 1348 | 11.36% | 0.87 |
Soft | 17219 | 2064 | 11.99% | 0.86 |
Standard | 63833 | 7502 | 11.75% | 0.86 |
Youāll notice that quite a few conditions are not represented here; thatās because there was insufficient data to draw a conclusion. What we can see above is that betting on horses which ran on the same type of ground the previous race is not a particularly good betting factor.
I then decided to see how favourites fared using the same criteria and the PIVās were as follows:
- Heavy: 0.94
- Good: 0.91
- Good to Firm: 0.91
- Good to Soft: 0.97
- Soft: 0.93
- Standard: 0.94
As you can see, there is an improvement across the board yet none of the PIV figures even reach 1.00.
Finally, I took a look at the performances of favourites on the same Going as their last win. For example, if Horse X is running on Heavy ground and his last win was also on Heavy ground.
- Heavy: 0.96
- Good: 0.91
- Good to Firm: 0.9
- Good to Soft: 0.89
- Soft: 0.92
- Standard: 0.97
While there are a myriad of other criteria you could use, the data above suggests that Going is not necessarily as important as you might think when choosing winners. None of the above provides you with a PIV value above 1 which means under performance.
Class
Once again, Iāll start with the performance of favourites but this time weāll divide them into Class.
Class | Bets | Wins | Strike Rate | PIV |
1 | 1961 | 699 | 35.65% | 0.95 |
2 | 2290 | 640 | 27.95% | 0.90 |
3 | 4368 | 1402 | 32.1% | 0.92 |
4 | 11466 | 4129 | 36.01% | 0.93 |
5 | 12768 | 4629 | 36.25% | 0.95 |
6 | 7918 | 2722 | 34.38% | 0.94 |
While favourites manage to achieve a PIV rating of at least 0.9, none of them make the magical 1.00 barrier.
The next step is to look at the performance of horses that are in the first four positions in the betting. Here are the PIV figures by Class:
- Class 1: 0.91
- Class 2: 0.89
- Class 3: 0.9
- Class 4: 0.91
- Class 5: 0.89
- Class 6: 0.89
The PIV figures here actually go down! Finally, I looked at the performance of horses that were either dropping one or two classes. For example, horses in Class 4 races that ran in Class 2 or Class 3 races last time out.
- Class 2: 0.88
- Class 3: 0.87
- Class 4: 0.91
- Class 5: 0.88
- Class 6: 0.85
While Iām certainly not suggesting that Class isnāt an important factor, the data above at least shows it may not necessarily be a profitable one by itself.
Official Rating (OR)
OR is used in handicapping to ensure each horse is theoretically equal. Of course, the handicappers donāt always get it right and savvy punters can cash in heavily when the experts make a rare error. But can you see profits from using OR alone?
It was tricky to come up with a set of criteria so I ultimately decided to see how horses within certain OR ranges performed in a certain Class of race. The following is a fairly small sample of OR ranges but will hopefully be enough to paint a somewhat clearer picture.
Flat
OR | Class | Strike Rate | PIV |
51-59 | 6 | 10.31% | 0.87 |
60-65 | 6 | 12.66% | 0.86 |
71-79 | 4 | 10.98% | 0.85 |
80-85 | 4 | 13.28% | 0.88 |
91-99 | 2 | 8.09% | 0.81 |
100-105 | 2 | 8.31% | 0.86 |
I went through All-Weather races using the above criteria and the PIV remained below 1.00.
National Hunt
OR | Class | Strike Rate | PIV |
75-85 | 5 | 10.25% | 0.88 |
86-95 | 5 | 10.88% | 0.85 |
115-125 | 3 | 12.09% | 0.85 |
126-135 | 3 | 14.66% | 0.9 |
While it is again an admittedly small sample of data, we can surmise that Official Rating isnāt a particularly profitable factor on its own.
Days Since Last Ran (DSLR)
This is an interesting metric as it can indicate if a horse performs better after a quick return or a lengthy lay off. Iāve divided it into the three codes.
National Hunt
DSLR | Bets | Wins | Strike Rate | PIV |
1-6 | 4640 | 709 | 15.28% | 0.96 |
7-14 | 30538 | 3221 | 10.55% | 0.84 |
15-30 | 73005 | 8293 | 11.36% | 0.86 |
31-60 | 45421 | 5119 | 11.27% | 0.87 |
61+ | 49890 | 4651 | 9.32% | 0.82 |
Flat
DSLR | Bets | Wins | Strike Rate | PIV |
1-6 | 8724 | 1257 | 14.41% | 0.94 |
7-14 | 46230 | 5027 | 10.87% | 0.84 |
15-30 | 68235 | 7709 | 11.3% | 0.87 |
31-60 | 27645 | 2863 | 10.36% | 0.86 |
61+ | 29045 | 2656 | 9.14% | 0.83 |
All Weather
DSLR | Bets | Wins | Strike Rate | PIV |
1-6 | 5146 | 848 | 16.48% | 0.97 |
7-14 | 26630 | 3095 | 11.62% | 0.84 |
15-30 | 36135 | 4096 | 11.34% | 0.85 |
31-60 | 16637 | 1749 | 10.51% | 0.85 |
61+ | 17574 | 1730 | 9.84% | 0.84 |
Unfortunately, there isnāt a single instance of a PIV of 1 or more. However, you can see a clear trend here. The PIV when betting on horses that had their last run in the 1-6 day range is by far the highest across the board.
I decided to analyse the performance of favourites who had run within the last 6 days (over all codes) and found the PIV to be 1.00. While National Hunt and Flat favourites had a PIV of 0.99, All-Weather favourites had a PIV figure of 1.02. All Weather favourites under these criteria also gave you a ROI of 3.28% at SP prices and a nice 8.24% ROI when using Betfair.
From this we can see that All-Weather favourites who last ran in the 1-6 day bracket perform slightly above expectations.
Conclusion
After lengthy research, it seems as if there is no single factor out of the four Iāve looked at thatās likely to lead to long-term profits. DSLR appears to be the most important factor on its own and it would be wise to focus on horses that have run within the last week.
This doesnāt mean you should ignore the factors above, it just means you need to combine them in order to find winners and more importantly, profit. It is further confirmation that there is no āmagicā factor in horse racing and if you believe anyone who says there is, I have a bridge Iād like to sell you!
Is there a flaw in the method of calculating PIV in that SPs are used, these are not an indication of winning success as the bookmakers margin is included in the figures? As an example in a two horse race where the horses have an identical chance the true odds of wining for each horse are 50%. However the bookies will offer let’s say 5/6 which on your calculations would give the odds of winning of 55%. Therefore using a PIV of 1 as a benchmark is not correct and possibly explains why all the PIVs are less than 1.
No response ?
Sorry for the delay in replying Michael. As you say there is also a margin to take into consideration when using SP odds, this will kick the figures out slightly. However we know that SP figures do (including bookmaker margins) give a very accurate probability on a horse’s chance of success. If you replace the SP with Betfair SP, this will be even more accurate as the over-rounds are tiny in comparison. Without doing a direct side-by-side comparison I wouldn’t be able to tell you the exact difference taking bookmaker margins into account would have, but would not expect it to be huge. However, I can be very confident it wouldn’t push any of the above stats to a PIV of above 1, primarily because this is already taken account of in the market heavily.