Front runners over jumps
In the last article, I returned to my first book, Flat Racing For Profit, with the intention of reviewing some of the systems I proposed back in 1996. However, I became side-tracked by the Tote vs SP analysis which resulted in a system free article. For this piece, I have revisited my second book of 1996, Jumps Racing For Profit, and promise not to again get distracted by price analyses, and instead I will focus on a system, specifically the one covering front runners in jumps racing.
Front runners sprang to prominence, at least they did in my race analysis, in the 1980s when Martin Pipe began to dominate jumps racing first with Richard Dunwoody and then with Peter Scudamore driving horses home from the front. The fact that this style of running means that horses have a good view of every obstacle, and can pinch a few lengths at the start, makes it an appealing way to race.
However, it does depend on the pace set, and that depends on the jockey. In Dunwoody and Scudamore, Pipe had the ideal jockeys and could rest assured that the pace set would be suitable for the horse. On the flat, Lester Piggott and Steve Cauthen were the equivalent, and all four jockeys seemed to have the most accurate internal clocks.
In 1996, I classified a “front running winner” as a horse for which the comment-in-running contained “made all” or “made virtually all”. Naturally this was simplistic, but it was easy to apply and completely unambiguous which is important for this type of research.
I found that approximately 12% of winners were front runners. The going did not seem to make a significant difference to this figure and whilst Firm or Heavy ground did produce a higher proportion of front running winners, this was ascribed to the fact that races on these surfaces attract smaller than average fields. An analysis by race distance did show a slight advantage to the shorter races, though, as would be expected.
Bringing these figures up to date we find that approximately 14% of jumps races are won by front runners, using our definition, and that the distribution of winners by going is very similar to that found twenty years ago with the proportion of winners on Firm and Heavy going slightly higher than the other ground classifications.
However, the results for race distance are not in line with previous findings with the proportions having a distinctly uniform appearance.
Whilst some may find these statistics interesting, they do not form a system since they relate to race performances, not future race performances. To make more use of them it is necessary to identify probable future front runners. Naturally, a runner’s previous performances are the best guide to this and in 1996 I decided to use the following rule to identify a probable future front runner: “A horse is identified as a probable front runner if it won its most recent outing and the terms “made all” or “made virtually all” appeared in its comment-in-running”.
Applying this rule to the data, and analysing by racecourse, I found that certain tracks seemed to be more profitable for potential front runners, and the following rule was identified: “consider backing any horse classified as a probable front runner at Carlisle, Catterick, Doncaster, Fontwell, Hereford, Hexham, Kelso, Lingfield, Musselburgh, Perth or Plumpton”.
So, how does this system rate now, twenty years on, and is there any potential for highlighting probable front runners nowadays?
Taking a sample of around 25,000 race performances by horses which had won their previous start in any style, I found that the average loss to Bookmakers’ starting price for their next race was 15p/£ and 5p/£ on the exchanges before commission. However, for horses defined as “probable front runners” using the previous rule, these losses reduced to 9p/£ for bets placed at industry starting price, and break-even for the exchanges. These better returns are also reflected in the strike rate, with probable front runners winning at 23% whilst the others win at 21%.
Interestingly, restricting the analysis to the shorter races, those up to 16 furlongs, improves the returns to -4p/£ and +6p/£ with the success rate improving to one in four. Heavy ground is also an intriguing condition to examine. The 340 probable front runners in the sample which were raced on heavy going produced a win rate of 31% and returned a profit at exchange prices of 19p/£ before commission. Furthermore, soft ground qualifiers also made a decent profit (8p/£ from 780 performances), but had the same win rate (22%) as other last time out winners.
However, the original rule related to race tracks, so highlighting those tracks mentioned in the 1996 rule, and running the analysis free from any other condition, produced the following results: 169 winners from 611 races returning a profit of 2p/£ at Bookmakers’ starting price and 12p/£ on the exchanges before commission.
It is always surprising to find a system that remains profitable after such a long break, but it should be noted that whilst it was profitable at Bookmakers’ starting price, at 2p/£ this is only just the case. But of course, all bets placed at Bookmakers’ starting prices are bad bets, so taking the exchange figures is an acceptable method of evaluation for 21st century punters and on that basis the system would appear to be viable.
Originally, the courses selected were based on a strike rate threshold of 35%, applying the same condition today would add in Kempton, Southwell and Wincanton and it will be interesting to see if these three courses continue to return profits in the future.
Obviously, I won’t be analysing races in 20 years’ time, but maybe someone will keep an eye on this method and hopefully find that it has remained profitable in 2037.