It should take no-one long to realise that results in horseracing (or other sports for that matter) do not always reflect what went on or the relative merits of the protagonists.
An extreme example is the Kauto Star Novices’ Chase at Kempton in 2016. The Form Book has Royal Vacation as a 12-length winner, but Might Bite was nearly 20 lengths clear when falling at the last, which is just 152 yards from the finish.
Despite this, and to my personal bemusement, some commentators refused to speculate on what might have happened otherwise and rated Royal Vacation ahead of his rival.
A slightly less obvious one is a lightly-raced and promising horse finishing in eye-catching style into sixth in a field of handicappers, for which it subsequently gets dropped 2 lb by the official handicapper.
An even less obvious one might be a horse winning impressively on its debut but in a steadily-run race in which even the wide margin of its victory fails to reflect how superior to its rivals it was.
Results are a starting point for analysis, then, not the end of the matter.
That last scenario, of a horse winning impressively on its debut but still in danger of being sold short, is the one that transpired with WALDKONIG at Wolverhampton on Saturday evening.
The Kingman half-brother to Arc winner Waldgeist, a 600,000-guinea yearling purchase, came home clear, but in an overall time equivalent to a figure of just 63 by my reckoning.
If you use that time, or even the bare facts of his beating modest rivals by nine lengths, as a starting and ending point then you will be short-changing this exciting talent.
Fortunately, there are alternatives, including using your eyes and experience to estimate how much better than the result Waldkonig was – the “visuals” approach – and using sectionals to try to quantify such things more precisely.
Simply gathering sectionals, let alone making sense of them, has been a laborious process for much of racing history. But things have changed of late thanks to Total Performance Data, whose figures for striding as well as for sectionals are displayed in the Results Section on this site.
Thus it is that we know Waldkonig ran the last two splits in 11.24s then 11.27s, while most of his rivals were struggling to break 12.0s, and that his supercharged finish came after an even enough opening but a notably slow mid-race.
A deeper dive into the TPD archive reveals that Waldkonig’s final 1f ranks 39th of over 15,000 efforts at the course, of all ages and at all distances, and 55th for the last 2f (top 0.25% and 0.35% respectively).
It is much easier to run fast when you have been conserving energy earlier in the race, of course. Sectional analysis needs to address that and to try to quantify it. Sectional upgrading does just that.
By establishing a horse’s finishing speed as a % of its average race speed, and comparing that to the par finishing speed % for the course and distance in question, it is possible to estimate how much more quickly a horse could have run had its energy been distributed evenly (a more detailed explanation is to be found in the free-to-download “Sectional Timing: An Introduction by Timeform”, though I have made some tweaks to the methodology since).
This is what I have come up with for the Wolverhampton race.
Sectional upgrading is deliberately conservative when horses run slow overall times but fast closing sectionals (informed by a branch of statistics known as Extreme Value Analysis), but it still has Waldkonig a 106 horse, rather than a 63 horse, as overall time analysis might have it, or a 90-ish horse, as race and prior-rating standardisation might have it.
That is the figure of a listed winner or a Group 3-placed performer. Talk of Waldkonig as a leading Derby contender may be premature, but Derby contenders have to come from somewhere, and very few horses manage to run a 106 sectional rating first time up, while looking green, and with minimal opposition.
That sectional upgrading does indeed suggest that the horses Waldkonig beat were modest. It will be fascinating to see what they do, and especially what HE does, from here.
Time analysis over jumps, including sectional analysis, requires accurate information about both times and distances.
The BHA finally addressed the latter when re-measuring race distances, and obliging courses to issue details of amendments to those race distances on account of rail movements, in 2015.
The good news is that they also aim to improve the accuracy of jumps times by having the electronic timing system triggered at the start, and not just at the finish as at present. This should be with us later this jumps season or early next.
The intention is to begin timing as the lead horse passes the starter, something which has been industry standard for manual systems for a long time now.
Timing from when the tape goes up can artificially add many seconds to a race due to horses being a long way back from the tape and/or failing to break into a gallop immediately.
A good example of the current deficiencies in this area was provided at Sandown last week, when most time sources had GOSHEN’s impressive juvenile hurdle win significantly slower than the 4m 11.7s from passing the starter that Timeform and I independently arrived at.
The alternative version has Goshen running an ordinary time. The “true” time suggests something a lot better than that. He leads my own time-based figures for juveniles by 8 lb from Allmankind.
Something as basic as the time it takes for a horse to complete a race should not be open to so much interpretation and misinterpretation. In future, all being well, things will be different.