The Grand National a couple of weeks ago has brought into sharp focus the subject of starting prices (SP), and the potential ease with which they can be manipulated. Whilst it is not the purpose of this scribe to conjecture about the ramifications – after all, if proven there could be a criminal case to be answered – it may be instructive to review some data and draw some conclusions.
Before that though, a spot of context and background gen.
What is an overround?
The race that brought the spotlight onto starting prices was not in fact the worst case of a “bookies’ book” in the past year. It was nevertheless a bad case: the returned SP’s had an overround of 165% (for the uninitiated, a definition of overround can be found here). In a nutshell, the suggestion is that if a bookie could have laid all horses in the National to lose an equal amount, he/she would have made profits of somewhere in the region of 39% of turnover.
For lots of reasons that’s a simplistic interpretation of the bottom line figure. (For instance, no bookie on the planet will have laid all horses in the race to lose the same figure, sweepstakes burglars aside). What is not in question is that that Aintree book leaned too far towards the seller and away from the buyer.
This has furrowed many a brow in the racing industry, and has led to questions being asked about the independence and, perhaps more pertinently, the very point of the Starting Price Regulatory Commission. More on them in due course…
A history of price fixing?
My own starting point, apart from the general noise in the various media (and social media) cubbies in which I loiter, was some excellent research by Timeform’s Head of R&D, Simon Rowlands. The first thing to say about Rowlands’ research is that it pre-dates the National by some months. Indeed, his original two-part investigation was penned last October.
Those two articles (part one here and part two here) offer a lot of context for what will follow and are well worth a read. Whilst there is much danger in paraphrasing, I’ll attempt to bring the click averse up to speed pronto. Here goes:
– There have been some alarming disparities between SP and Betfair SP (BFSP) in recent times which could be argued to suit off course bookmakers looking to lay liabilities into the on course markets
– This has been most marked at low liquidity on-course meetings, primarily on the all weather.
– Starting prices are supposed to be returned based on the prices available with on-course bookmakers who are NOT taking hedging money from off-course bookmakers (e.g. high street firms). Thus, it should be impermissible for high street firms to influence the starting price market. (More on this can be found at the SPRC’s website here).
– With regards to the mechanics of overrounds, generally speaking the bigger the field the smaller per runner the overround. Thus it is never appropriate to discuss overround without mentioning field size.
– Rowlands’ baseline research covers all flat races in Britain from 2011 to 2013, and when the 2014 data was compared against those benchmark figures, it appeared that ‘eight runners’ and ‘all weather’ were recurring themes. A conclusion drawn from the number of runners observation was that these races often presented unfavourable each way propositions for the off-course layers. Though of course we know this could not be true because off-course layers have no means of influencing the on-course market. [*Cough*].
– It is the job of the SPRC to define policy for the recording of starting prices; and it would appear that they are failing to effectively police their policies.
On that last point, I question whether it is actually the role of SPRC to monitor and control the return of SP’s, as opposed to ‘merely’ define how those SP’s should be returned. More on that anon.
Meanwhile, for some time now, I’ve anecdotally felt that there was a propensity for the top end of big Saturday handicap markets to truncate – and therefore increase the overround, both in total and per runner. This seemed as good a reason as I’ve had to drill down into my perception and establish if there’s any truth to it.
The Governance of Starting Prices
I’ve been party to the business of starting price returning for over 20 years. Back in the day, 1996 to be precise, I was employed by SIS in their text room. At that time, before racing had dedicated satellite channels on your Sky box, SIS was the sole broadcaster into betting shops.
My role included, amongst other things, relaying prices from the course to the betting shop via a teletext update service. At the start of a day’s racing, I’d be told which bookmakers’ boards were going to be used for the purposes of returning the SP’s.
The Press Association was employed then, and is employed now, to validate the Starting Price returns. The rules governing this process are defined by a small group calling themselves the Starting Price Regulatory Commission (SPRC). They have a board of four members: Lord Donoughue, a former minister (chair); Professor David Metcalf (steward and top don at the London School of Economics); Lord Lipsey, former tote director; and Jim Donnelley of the Press Association acts as company secretary.
The problem, so far as I can tell, with this “independent body responsible for the integrity of the starting price” (their words), is that I cannot find any evidence to disabuse of the notion that they are entirely self-appointed.
Be that as it may (or may not), the fact remains that this entity prescribes how starting prices are arrived at. They explicitly state that the process is designed to deliberately ensure that off-course interests cannot manipulate the on-course prices used for SP returns to their own ends. That is to say, high street firms are perfectly within their rights to funnel liabilities back into the betting ring on-course, but they should be doing that through bookmakers excluded from the SP returning process.
It would appear that, with SPRC being an independent ‘process defining’ single cell organism, they are unable to monitor or control the everyday operational aspects of SP returning. It would further appear that there is no day-to-day accountability for those returning SP’s. [NB The fact that I’ve been unable to identify any form of ‘chain of command’ or feedback loop into BHA does not preclude there being one. However, it is fair to assume that if there was an official reporting line, it would form part of the governance structure published at SPRC’s website].
Overrounds on the Biggest Betting Races
So far, so what?, you might say. Well, the implication is that starting prices may be susceptible to manipulation without any officially sanctioned checks or balances; and, if that’s the case, then any punter that doesn’t take a price on their wagers may not be receiving the full stack on a winner.
To the recent hullabaloo around the Grand National, and the wider data outlined in Rowlands’ pieces, I decided to add my own tuppence. I looked at two datasets, with over-lapping but materially different market shapes.
The first was the top ten betting races, by turnover, in each quarter of 2014, as published on the Horserace Betting Levy Board’s website.
The data is not necessarily straightforward to consume, more so due to a key assumption I’ve used for baseline purposes. Specifically, I took Simon Rowlands’ flat data and have applied it across both flat and jumps. Whilst that’s not an unreasonable thing to do, it will mean that a tolerance should be applied, both logically and statistically, when looking at the National Hunt race returns.
This group of forty races includes both jumps and flat, handicaps and non-handicaps. Where the overround per runner is red, it was higher than the flat average for 2011-2013 for that number of runners; where it is amber, the overround per runner was marginally higher than the flat average for 2011-2013 for that number of runners; and where it is green, the overround per runner was lower than the flat average for 2011-2013 for that number of runners.
Here are a couple of observations:
– Although four of the five biggest overrounds per runner were in non-handicaps, 72% (18 of 25) of the highest betting turnover non-handicap races per quarter actually offered better than average value.
– Of the handicap races in the sample, two-thirds (ten out of fifteen) offered worse than average value based on overround per runner set against field size.
Overrounds on Saturday Handicaps
To assuage my own suspicions that the Grand National was far from an isolated incident, and rather merely the latest in a litany of poor value Saturday markets, I looked at Saturday handicaps in the UK worth more than £50,000 to the winner.
There were 37 such races, 23 flat and 14 jumps, during 2014; and the overround information is below. A quick key will help, as will a couple of parish notices:
– Ran: Number of runners
– Ovrnd: Overround on the race
– Exp O/R: The expected overround per runner based on the Timeform 2011-13 flat research referenced above
– Act O/R: The actual overround per runner in the race
– Diff: The difference between expected and actual (negative is worse than average)
– Flat / NH: Race code, flat or National Hunt (jumps)
– Type: Sprint = flat 7f or shorter; Non-sprint = longer flat races; Chase / Hurdle: as they are
As with the HBLB data inferences, the assumption that flat overrounds are equally appropriate for National Hunt races is moot. It serves as a ‘best guess’ barometer for the purposes of this post.
There are plenty of observations, and some conclusions, to be drawn from the above table.
First, it is noticeable that five of the six worst value ‘big pot’ Saturday handicaps were flat races at seven furlongs or shorter. As an aside, it does seem that the programme book is awash with high value sprint handicaps, such races accounting for 38% of all Saturday handicaps with a win prize of greater than £50,000. I must try to acquire a 90-rated seven furlong specialist..!!
The overall average overround per runner across the 37 race sample was roughly 0.06 per runner worse than the average of 2011-2013 flat races. Given the average field size of 20.5, that equates to roughly 1.23% above what might be expected per race.
There were 23 races where the overround was worse than the previous three year average, and 14 races where it was better.
Flat handicaps (Saturday, £50k+ to the winner) overall were 0.08 per runner worse than average for field size (direct comparison with control data). On an average field size of 21.2, that equates to an extra 1.7% margin for the bookies per race.
NH handicaps overall were 0.02 per runner worse than average for field size. Given that the comparison data is for flat, it’s perfectly possible that National Hunt races in the sample actually performed better than average. And, naturally, it is also possible that they performed worse than shown. In other words, it may be reasonable (but no more than that) to assume that the biggest prize fund Saturday National Hunt handicaps offer a fair price to SP punters, more or less.
Ascot has quite a few big Saturday handicap prizes, and it averages out at 0.11 per runner worse than average overround. On their average field size of 19.9 that amounts to a bonus of 2.19% per race for the Berkshire bookies (Geoff Banks, take note when you read this 😉 )
When looking at field size, the ten races in the sample with 16 or fewer runners actually offered slightly better than average value (+0.03 per runner), from ten races. Those with between 17 and 20 starters (15 races) presented SP punters with 0.09 per runner less than average value; and those with 21 or more horses involved were roughly comparable with the 17-20 bracket (12 races, 0.08 per runner worse than average).
The best value in the big Saturday handicaps seems to be in flat races over a mile and more, with the likes of the Lincoln, Ebor, Cesarewitch and the Cambridgeshire all offering better than average value to SP punters, in 2014 at least.
Where does all this SP overrounding leave us?
There are two main areas of conclusion to draw together.
The first is in the data. The starting price overround on the most valuable Saturday handicaps seems to be higher per runner than expected. On £50k+ Saturday sprint handicaps, it is materially higher than expected.
As with the Timeform study’s assertion that the “bad each way” nature of eight runner races is leading to some anomalous SP returns, there is a case to suggest the same may be true of the bigger field Saturday handicaps. If one were to layer conjecture on top of speculation, one might suspect that off-course money was moving to manage the positions of horses laid in concessionary place markets. But, as we know, the returned SP’s come from the boards of on-course bookmakers independent of off-course liabilities. [*Cough*].
The second area of conclusion is around governance. Whether or not the data points to a suspicion of price fixing – something (my lawyers and) I am happy to leave for the reader to ponder – the fact remains that there is currently no effective means of policing the ring to ensure that such behaviour cannot happen.
The SPRC are doubtless well-meaning, but they’re headed up by career theorists (Professor Metcalf) and policy makers (Lord Donoughue), and operate seemingly without mandate. Further, they don’t appear to have any interest in policing the day-to-operations of the betting ring; nor do they have powers to the same end.
Indeed, given the BHA’s own Integrity Services department has a remit where, amongst other things, “all betting markets are monitored in real-time, seven days a week, for any suspicious betting activity”, there is a solid argument to suggest the SPRC is entirely superfluous.
There could be a secondary argument around whether or not the BHA’s team are either a) sufficiently well briefed to detect these SP nuances, or b) sufficiently well resourced to pursue investigations.
The most recent case – the 2015 Grand National – is also the most high profile, by some distance. I very much doubt it will result in a house of cards type investigation outing systematic malpractice by off-course firms, mainly because – of course – that has absolutely not been happening. [I must get this *cough* seen to].
But the enquiry should run more deeply than merely matters at Aintree a fortnight ago.
What can we do to protect ourselves from excessive SP overrounds?
At the end of the day, bookmakers and markets generally, give and take. It is our individual responsibility to get the best value from our betting pound. That means doing any or all of the below, where such options are available to us:
– Take the best price available on a Best Odds Guaranteed basis
– Learn how to beat SP, and take the best price available on a non-BOG basis [Hint: Geegeez Gold is built to isolate early value like no other form tool in Britain. It WILL help you consistently beat SP]
– If you can do neither of the above, back at BFSP, especially for highly liquid markets such as those in the samples above
Obviously, the easiest way to be immune to any SP anomalies is not to ever bet at SP. In this digital day and information age, there are very few reasons why one would back at Starting Price. The above fact and theory are simply a few more nails in the coffin of SP wagering, for this punter at least.
p.s. What are your thoughts on SP markets? Have you noticed anything you weren’t quite sure about? Leave a comment and share your thoughts.