Bob Wilkins Q&A

[cormack15]

I’ve posted Bob Wilkins answers to the recent Q&A session. (Bob is the author of the excellent).

https://theracingforum.co.uk/horse-racing-forum/celebrity/bob-wilkins-author-bioenergetics-and-racehorse-ratings-t81347.html

[Prufrock]

I wouldn’t mind asking a supplementary question, but will understand if it does not get answered as I have missed my opportunity, viz:

You would expect theory about pounds per second (from which pounds per length can be deduced) to be borne out by results, but the evidence is far from convincing.

Average cumulative margins beaten per rival for horses in handicaps on the Flat in UK under normalised conditions (before the BHA introduced variable lengths-per-second) were much more in line with conventional wisdom of what pounds per length should be.

As an example, dividing by the values for average cumulative lengths beaten per rival would point to: 3.64 lb/length at 5f; 2.98 lb/length at 6f; 2.68 lb/length at 7f; and 2.50 lb/length at 8f (or some multiple of all those values) on turf.

Why are theory and empiricism so far apart, and why should we trust the theory to the exclusion of the evidence?

[cormack15]

I’ll send that on to Bob Pru – I’m sure he won’t mind answering it.

[Prufrock]

Thanks.

Incidentally, on the subject of sectionals, we may not have electronic ones now, but we did do for some time, and I imagine something could be learned from them from a bioenergetic point of view.

For instance, the average speeds in yd/sec for all horses running at 1m at Kempton March 2006 to May 2008 by-furlong (start to end) were:
16.58
17.78
17.92
17.49
17.66
18.44
18.71
14.78

Whereas the average speeds for horses running handicaps in times that were good compared to their abilities were markedly different when normalised to the same final time, notably in the final 1f (as you would expect) but also in the first half of the race.

Would gladly share the data for all four all-weather tracks if it was of interest.

SDR

[indocine]

As an example, dividing by the values for average cumulative lengths beaten per rival would point to: 3.64 lb/length at 5f; 2.98 lb/length at 6f; 2.68 lb/length at 7f; and 2.50 lb/length at 8f (or some multiple of all those values) on turf.

These were derived from the old 0.2s/L system, pre feb 08?
What do they become if you normalise them for his fixed chosen length = 2.7m? Bigger than his even?

[Prufrock]

A horse’s actual physical length is a red herring in this instance, as a length was a fixed unit of time (0.20 sec, as you say) then.

The actual figures (converted into time, just so there is no confusion on that score) were: 0.0840 sec per rival at 5f; 0.1006 sec per rival at 6f; 0.1120 sec per rival at 7f; and 0.1220 sec per rival at 8f.

Having inverted those figures, it does not matter what constant you multiply them by, pounds against unit time comes out as a flatter curve than the one proposed in the book.

For instance, horses at 8f get strung out fractionally less than 1.5 times as much – in terms of time – as those at 5f under normalised conditions. Theory suggests it should be a bit over 1.75.

[Author]

Posts