A common racing myth is that in-form horses win more races. Another common myth is that horses making a quick return to the track following a good performance are more likely to win. Having done a quick trawl of the internet, I was dismayed to find little evidence backing up these alleged facts. Therefore, I thought I would do some digging in the search of a betting edge.
To conduct this analysis we need to classify and measure a horse's last performance in a race and measure any influence the time since the horse last ran may have.
Defining A Horses Performance
How do we classify a horse's performance in a race? Having checked Google and Wikipedia I was surprised that no standard definition existed for a horse's race performance. Therefore, I have established the defacto standard using the time tested classification system of Good, Okay and Poor:
- Good Performance - A horse that has won or has come close to winning the race. Typically placed or beaten less than two lengths on the flat and less than 5 lengths over the jumps.
- Okay Performance - A horse that has beaten half the field or been beaten less than 5 lengths on the flat or 15 lengths over the jumps.
- Poor Performance - None of the above.
Whilst I am sure there are many other variations I could have used, these provide a reasonable starting point for an analysis.
Influence of Time Since Last Race
Whilst measuring the performance level of a horse's last time out performance is useful, this needs to be considered alongside when that performance actually happened to determine if any significent relationship exists. To do this I have grouped the time since the horse last ran into the following categories:
- 1-7 Days
- 8-21 Days
- 22-56 Days
- 57-100 Days
- 100+ Days
- 200+ Days
Analysing Last Time Out Performance Versus Days Since Last Run
With our Performance and Time classifications in place, I have produced the following graph using 2016 racing data covering both the flat and the jumps.
The key learnings we can take from this graph are as follows:
1) Horses that had a 'good' performance last time out and re-appear within 7 days win almost 20% of the time. Almost 3 times more often than horses that re-appear within 7 days having registered a 'poor' performance last time.
2) A 'good' performance last time out, no matter how many days ago the performance happened, is significantly better than horses classed as 'okay' or 'poor' last time out.
3) All the classifications win races, but the statistics show that horses classed as 'good' win up to 1.5 - 2 times as often as those classed as 'okay' and those classed as 'okay' win between 1.5 and 2 times more often than those classed as 'poor'. Therefore, the price you take when betting a horse should consider these facts.
Analysing Last Time Out Performance Across Race Types
Another interesting angle is to see how last time performance translates across the different categories of race types using the classification I used in my previous post Gain A Betting Edge Part 1: Horses for Races
The key learnings we can take from this graph are:
1) 'Good' performances significantly outweigh 'okay' and 'poor' performances in the majority of cases.
2) Horses who had a 'good' last time out performance in claimers, maidens, sellers and the other category of races have significantly high win strike rates.
3) Horses running within 7 days of running to a 'good' level of performance in a handicap race have almost a 20% win strike rate.
4) In seller and maiden races, the 'good' and 'okay' categories have similar win strike rates.
Where Is This Information Available?
You can use any good racing paper or website to check for the days since a horse last ran and can apply the formula I have used to categorise the horses last performance into 'good', 'poor' and 'okay. It would take several hours to do that for every horse in every race.
A better way is to use a website I have developed (www.betturtle.com) where all this information is readily available.
By using the earlier graphs and the information contained in the BetTurtle racecards (click the 'Stats' button and select the 'Form' tab), you can quickly see the horse's last time our performance category and just above it, the number of days since the horse last ran.
So in the example race card above, I would use the following thought process:
1) I can see that this is a nursery handicap for unexposed 2 year olds and that 6 horses had 'good' performances last time and these will carry my focus.
2) John T Chance is the only one of these that has won a race.
3) Katebird and John T Chance have run most recently of the 6 and warrant close inspection in this handicap race.
4) I would then use the rest of the information available on BetTurtle and assess the chances of the 6 runners and decide which ones have a higher chance of winning than today's odds suggest (value price).
*** We now have a page on the website with all of today's runners rated having a 'good' performance last time out. You can filter and sort this list to help you with finding today's winner. View it here. ***