"16", "name" => "Horse Racing", "category" => "Simon Rowlands", "path" => "/var/www/vhosts/betting.betfair.com/httpdocs/horse-racing/", "url" => "https://betting.betfair.com/horse-racing/", "title" => "Simon Rowlands: A different angle on the 'Whip Debate' : Simon Rowlands : Horse Racing", "desc" => "The BHA's new man, Paul Bittar, has walked into a veritable firestorm over the whip, new rules regarding which were introduced in October. Simon Rowlands gives a different angle on the debate......", "keywords" => "", "robots" => "index,follow" ); ?>

Simon Rowlands: A different angle on the 'Whip Debate'

Simon Rowlands RSS / / 23 January 2012 / 3

" class="free_bet_btn" rel="external" onclick="javascript: pageTracker._trackPageview('/G4/inline-freebet');" target="_blank">
Has the whip had an effect on how races are run?

Has the whip had an effect on how races are run?

"One interesting thing that came out of the study was a change in predictability of winners from before to after the implementation of the new whip rules..."

The BHA's new man, Paul Bittar, has walked into a veritable firestorm over the whip, new rules regarding which were introduced in October. Simon Rowlands gives a different angle on the debate...

I stated at the time of their introduction that I was "relatively agnostic" about the new whip rules. I did not feel well informed about the reasons for them, or about the reasons why people might legitimately oppose them. I waited to become better informed.

Three months on, and I feel scarcely any wiser. There has been a chronic lack of clear communication from the BHA on the one hand and from many opposed to the BHA's new rules on the other. The media - with the Racing Post a creditable exception in this instance - has been particularly guilty of being light on fact and logic and heavy on fiction and emotion.

In among all this, some off-the-cuff theories have cropped up, such as that the changes to the whip rules have led to a fundamental change in the way in which races tend to be run. It is more difficult to make up ground now that one means by which jockeys might galvanise a horse (the whip) has been so severely limited, or so the theory goes.

I may not have much to contribute on "The Whip Debate" itself - other than getting irate that some journalists seem incapable of holding more than one side to account - but I hope I can contribute where proving or disproving theories go.

What I did was look at a comparable one-month period before and after the new whip rules were introduced and gauged how horses that raced prominently or were held up fared under each regime.

In order to do this, I used the in-play symbols that appear on Timeform Race Passes and measured how they performed in terms of percentage of rivals beaten. As I have stated before (and will doubtless state many times hence), too many statistics commit the cardinal sins of ignoring the effect of field size and of regarding racehorse performance as a simple binomial: "did it win - yes or no?"

Percentage of rivals beaten allows for both field size and the fact that there is a world of difference between finishing second in a field of 20 and tenth in a field of 10. To some, the horses in both outcomes are both losers, and that is all that matters.

For all-weather racing in Britain in December 2010, horses with in-play symbols of "m" (raced prominently) or "M" (led) beat 55.9% of their rivals; in December 2011, the figure had increased to 59.3%. At the same time, horses with in-play symbols of "h" (held up) or "H" (held up, well behind) went from beating 50.8% of their rivals to 50.0% of their rivals.

There could be something in the theory on all-weather, in other words.

However, for jumps racing in November 2010 (just before the cold snap which caused many cancellations), prominent racers accounted for 66.3% of their rivals, whereas this figure had dropped to 64.0% for November 2011. Held-up horses again remained fairly steady, down from 58.0% to 57.4%. For the purposes of this study, non-completers were ignored.

There is not a lot to support the theory over jumps, judged by these means anyway, in other words.

One interesting thing that came out of the study was a change in predictability of winners from before to after the implementation of the new whip rules. After normalising Betfair SPs for field size, the degree of predictability could be seen to have increased for both Flat and jumps, especially the latter.

Winners have, on average, returned Betfair SPs lower (compared to random for the field sizes under consideration) after the whip rules than before. The average Betfair SP for a "normal" 10-runner race on the Flat decreased from [5.8] to [4.9], while over jumps it went from [7.6] to [4.8], in the periods under review.

There may be good reasons to oppose the new whip rules, but that they have led to more unpredictable results is not one of them, while their effect on how races tend to be run can at least be disputed in some respects.
Get unsurpassed levels of analysis and statistics for any race in Britain and Ireland with Timeform Race Passes. Click HERE to find out more.


'; } } ?>


  1. Jonathan da Silva | 23 January 2012

    Poorest part for me is the tendentious stuff from many columnists. Jockeys who get banned have explanations the Stewards did not accept explained as fact. Last week a horse given no chance to respond and hit in the wrong place was defended without explanation of why the bans and why it was wrong.

    Watching jumpers be whipped less has actually been enjoyable.

    I take people's points about breeding and Group 1 races and maybe Group 2. I can see how 8 strokes may be insufficient especially if they are using a UK whip and administered without excessive force in the right place. It's the pinnacle of the sport and very different to a class 2 handicap or a horse being asked to jump fences further stressed.

  2. Simon Rowlands | 24 January 2012

    Thanks for the feedback, Jon.

    I deliberately avoided getting involved in the whip debate per se in the above, but I would agree with you that the performance of some sections of the media (the broadcast media in particular) on the subject has been lamentable.

    While we could do with hearing MORE from the BHA regarding why it is necessary to restrict the use of a whip which others say is incapable of inflicting harm, we could do with hearing a great deal LESS from those who think the way to change rules with which the jockeys initially agreed is to flout them.

    The "logic" employed to defend this latter stance beggars belief in some instances. It is not helped by commentators insulting those who might disagree with them then denying a right of reply.

    Like you, I find racing more aesthetically pleasing now than before. I also don't find it any more unfathomable than before, and the above would seem to support that. But mine is just one personal take on events where there is too much subjectivity on display already.


  3. Simon Rowlands | 24 January 2012

    Thanks also to the correspondent - who wishes to remain anonymous - who contacted me to query my means for judging predictability in the above:

    "Interesting article. I don't see why you are using the measure you're using, however. It's more complex than the basic mean (per runner) and has lots of issues associated with it. Primary among these is that different races have vastly different expected values and therefore aren't really comparable.

    Races with short-priced horses will return much bigger values on average, almost regardless of results, which makes it much more a measure of the race programme and general competitiveness of races than the psychic abilities of the market.

    You also run into reader interpretation issues when you convert back to SP. If you're going to analyse impact values, you should refer to average impact values when reporting results IMHO rather than backward engineering them back to SPs.

    I'd just use average SP. To control for runners use SP/runners as the measure for each race rather than the inverse that you used. This has a lot of useful properties - all races with 100% books have an expected value of 1, so it's easy to interpret, display (you can multiply by 10 to get average SP for a 10-runner race), aggregate, compare etc."

    I think he is right and stand corrected.

    I am glad to say the correspondent also stated that he agreed with the gist of the article and that his alternative method for measuring predictability came up with similar results, albeit by less convoluted means.