Simon Rowlands: Official going, winning distances and longest SP

Ante-post / Simon Rowlands / 07 March 2011 / 4

" class="free_bet_btn" rel="external" onclick="javascript: pageTracker._trackPageview('/G4/inline-freebet');" target="_blank">

What will the going be described as on day one?

Our blogger takes a look at three of the Cheltenham Festival Specials markets available on Betfair...

With the Cheltenham Festival approaching fast, various specials markets have started springing up all over the betting industry. Some of these represent little more than 'hit and hope' jobs at this stage, but it is surprising to discover just how far a modest amount of research can take you with the others.

I decided to look at the markets for official going, winning distances and longest SP.

The following is the historical data since 2001, the year in which the meeting was cancelled due to foot and mouth:

The official going market is settled according to the official description of the majority of the ground on the hurdle course prior to the first race of the meeting. In this, it is subject to human interference in both the imperative to achieve a certain outcome ("good jumping ground") and in the possible inclination to believe that that outcome has been achieved, come what may (what might be termed the "mark your own exam paper" phenomenon).

Simon Claisse has been Clerk of The Course at Cheltenham since 2000 and Timeform, using race times as well as other information, has disagreed with his description in four of the last five years.

In other words, it matters little what the ground is, if Mr Claisse states it is "good", or more commonly "good to soft", that is what you will get paid out on. He even changed it from "good to soft" to a more credible "good" halfway through the first day of last year's meeting.

Claisse consults with famed weather forecaster John Kettley in the weeks leading up to the meeting and waters as he sees fit. Current odds of [1.42] on Betfair suggest that Claisse will achieve his goal of describing the ground for the first as "good to soft".

The state of the going - real, rather than imagined - has an impact on many other markets, not least that of winning distances. Wider modelling of margins confirms that they increase as the ground becomes softer, though more gradually than might be supposed.

Gauging the real significance of winning distances at the Cheltenham Festival has been akin to trying to herd cats over the years. Not only has the going varied, but the numbers and types of races have varied, too, as has the means by which the margins themselves are determined.

The meeting went from being a three-day one to a four-day one after the 2004 Festival, and there will be another race - making 27 in all - this year. That race is a two-and-a-half mile novice handicap chase, while the Jewson Chase becomes a non-handicap.

The final column in the spreadsheet takes into account the different means of returning winning margins prior to 2009 and adjusts those margins up to 27 races on a pro-rata basis. It is not that simple, of course, but it is some sort of a guide.

The average of this column is 102, which is close to the midpoint of both the winning distance odds market and the winning distance line market, but the median is just 92.

It seems to me as if a couple of wide-margin years recently (one of which was run on genuinely testing ground) might have skewed people's perception. There is also a tendency for people to bet "high" - to want something to happen, rather than not happen - with these markets.

A sell of the winning distance line at 100 or more and a back of winning distance odds of less than 90 lengths makes appeal unless there seems a good likelihood of conditions turning testing. It should be noted that any abandoned or voided races will be made up at 10 lengths (see rules tab), even though the average winning margin has been just under 4 lengths.

The longest SP winner market is divided into "shorter than 33/1", "longer than 50/1" and in between. Bear in mind that just one instance of a rank outsider will trump everything else, and that, in the absence of that, just one instance of an in-between winner will trump the shorter-priced category. 187 of 209 races (89.5%) have been won by "unders", 21 races (10%) by "in between" and just 1 race (0.5%) by "overs".

If you convert that into probabilities, the implication is that the true odds should, respectively, be something like: [15.0]; [1.23]; and [8.2]. In the last nine years, the in-between category has prevailed eight times and the rank outsiders once.

I am, however, very sceptical that this is representative. A better measure - one which shows the middle category to have been overperforming greatly (and probably randomly) against expectation - is likely to be the average amount of the book taken up by each category in races in which a horse's true chance has been established from the Betfair SP.

This eliminates the so-called favourite-longshot bias and produces expected figures of 93.5% unders, 5.5% in between and 1% overs. The resulting probabilities produce (by my reckoning) estimated likely odds of [1.84] in between; [4.2] overs; and [4.6] unders.

It may seem counter-intuitive, but, with the market as it is, unders looks the best bet. It could be expected to come off only about 22% of the time, and has done so not once recently, but [8.0] is too big if you believe theory and past market expectation rather than potentially misleading evidence.

'.$sign_up['title'].'

'; } } ?>

(4)

Paul | 15 March 2011

Interesting stuff Simon.

Two questions about winning margins.

1 - You say that the measure of winning margins change in 2009 how so?

2 - Has the number of runners remained constant over the last 10 years and do you think that this has any significant effect on the winning distances?

Thanks
Paul
Simon Rowlands | 15 March 2011

Good questions, Paul.

The means for returning official margins changed prior to the 2009 Cheltenham Festival from 4 lengths per second - regardless of circumstances - to 4, 4.5 or 5 lps according to the main official description of the going (further details on my blogs about time analysis). This applies to all jumps racing.

I nearly put in average field size but did not want the spreadsheet to get too big and was also mindful of the fact that this year's average field size could only be guessed at at the time.

The averages are as follows: 2002, 20.6; 2003, 19.0; 2004, 17.3; 2005, 19.5; 2006, 20.2; 2007, 19.6; 2008, 17.6; 2009, 19.5; and 2010, 18.8.

My experience with field-size adjustments, both in winning-margin betting and in standardised form handicapping, is that they make plenty of difference at smaller values and when there are sizeable changes but little when the field sizes are as big and the variation is as small as has been the case at the Cheltenham Festival.

That said, it is not negligible. A back-of-the-fag-packet calculation (I would need to factor in precise distances of all races run in each separate year) suggests a 17-runner race and a 20-runner race would cumulate to in the region of 8 lengths difference over 27 races.

I am hoping for big fields, close finishes (under 90 lengths cumulatively) and no great surprises (longest SP under 33/1) to give my own personal Cheltenham Festival a profitable basis!

Simon
Paul | 15 March 2011

Simon,

Thanks for coming back to me so quickly.

A few things of interest there, particularly the means of returning official margins as it seems to promote a double whammy on softer going. To explain, I would expect horses to finish further apart on heavier going so to then multiply that amount by a bigger factor would seem to be double accounting?? So lets hope for drying conditions as well as big fields leading to close finishes!!

The other thing I noted in your main article was this "There is also a tendency for people to bet "high" - to want something to happen, rather than not happen - with these markets.". An interesting theory, that I happen to believe in quite strongly, and I think it has lead to the winning distance being skewed a little bit high which might help us out too. I have a little betting system that utilises this 'human trait' but that's not for public consumption...
Simon Rowlands | 15 March 2011

It is not a double whammy, as the lengths-per-second is LESS the softer the ground, in an attempt to allow for the slower speed at which horses are usually travelling in such conditions. It means well, but it fails.

Nonetheless, what a margin once was and what the identical margin would be now are not the same, so an adjustment to my figures seemed necessary.

On the second point, I have to own up to not having researched this in any great detail. It was a bit of a throw-away line, though, like you, I strongly suspect it is correct. I am hoping so, anyway!

After day one, 10/1 has been the longest-priced winner, with outsiders not even getting much of a sniff, but "unders" in the Betfair market can still be backed at 5.3.

I have a horrible feeling this is going to be one of those value bets that comes unstuck, but those odds are too big if you have not already got involved.

The market for the cumulative winning distance odds is relatively unformed, but "unders" was last matched at 2.14, which suggests we have got a bit of value at 3.0 or a bit bigger.

The current make-up is 20.5 lengths after 7/27 races. The ground has little or no "good to soft" in it and may even be tending towards "good to firm" if it does not rain. So far, so reasonably good.

Simon