Wizard of odds

Whether you are buying red diamonds, a Bahamas holiday or half a dozen bananas, someone, somewhere has chosen and assigned a price at which you may buy.  You may consider their evaluation a bargain and good value, or extortionate and overpriced. I have sat on this side of the gambling industry fence for many years now, predominantly as a horse racing price maker and odds compiler. How do I compile odds and price races, I am constantly asked. Well, here goes.

With the world’s information so much more readily accessible, gone are the days when odds compilers emerged from a darkened room having been surrounded by desk calculators, VHS recorders and Timeform Black Books. Now we have statistical models and digital form analysis, all available at the click of a button, as well as numerous online bookmakers with ‘early shows’, exchanges, and tipsters all vying for our custom. They can all help a line maker, but the art of odds compiling hasn’t changed, and indeed many professional gamblers still choose to compile their own odds in order to identify market value 

A bettor and a compiler will break down a race the same way, but while one has an objective – to back the winner – the other is being subjective, looking to assign each horse a percentage chance of winning. Under that day’s conditions, if a race were run an infinite number of times, what percentage would each horse win? Effectively as a bookmaker, I am aiming to lay every horse at the right percentage. To get there, I liken it to piecing together a jigsaw puzzle. No two races are the same, yet every compilation starts the same way – with a blank sheet of paper.

It would be remiss of me to suggest that a horse’s current form, trainer and jockey, particularly names like Aidan O’Brien and Frankie Dettori, are not of obvious significance when putting the four corner pieces down. What follows, however, may provide more insight to the nuances of the process.

Race Conditions – initial research
My second career may be as a weather forecaster! It is crucial to know the expected state of ground conditions for a race, so a large amount of time is spent obtaining the most accurate and specific weather forecasts. I would always start by familiarising myself with past history and trends of a race.

Ground, Distance, Track & Draw analysis
In my opinion the most important variable in any horse race is the going. If a horse doesn’t act on the ground, it is much less likely to win. All horses stay, some just stay faster than others, but proven suitability of unique tracks like Brighton (where some horses are all at sea) is a big plus, as is obviously a low draw at Chester.

Video analysis
There is no substitute for watching a race and forming your own opinion – then you can only blame yourself! A desperately unlucky run or a horse drawn on the wrong part of the track may necessitate a form upgrade, or conversely a horse that hung left under pressure may not cope in a big field at Epsom. A small filly may not want to be burdened with carrying heavy weights in handicaps. Trust your own eyes.

Ranking and Rating
I would initially assign a personal rating for each horse, accompanied with any salient notes, based on ability, the aforementioned variables, and their suitability for that day’s race. This effectively gives me a relative rating, and a ranked order of the most likely winners down to the outsiders.

Shaping the Race
I try to predict how the race will be run by analysing individual running styles. We may discover half the field like to go forward early and a strong pace is guaranteed, in which case they may all compromise each other’s chance of winning. There may be a race with only one front runner, so an easy, controlled lead is expected – but if that horse is drawn in the car park there might be no pace at all. Of course, things very rarely pan out as expected, but races are often won by the smallest margins.

There are many good sources of speed ratings, and while I try to keep their reliance in pricing to a minimum, only good horses can run fast. It may highlight a speedy final furlong in a previous race that may give a horse a great chance in a race expected to be run slowly. Horses get assistance too, and any change of headgear may give them the extra focus they need, especially for that horse that you spotted  hanging left.

Other factors 
With a bookmaker’s hat on, it might be that commercially we need to be as competitive as possible for a particular horse on a certain day, which also needs to be considered. Any current liabilities, previous market moves, or even particularly popular horses can all affect the prices.

After adjusting my initial ratings to encompass any educated guesswork such as race speed and shape, now the pricing. The most important horse in any market is the favourite. A clear favourite is easier to price than a race with many fancied runners. In truth, there is no hard and fast way to price the jolly, only that with experience it becomes instinctive. All the other horses have a lesser percentage chance of winning related to the favourite and are simply priced accordingly. 

Continued Fluidity
So many factors can continue to influence prices until the race. A trainer or jockey may be blitzing their way through the card, or a big draw bias may become prevalent. A deluge of rain may have resulted in withdrawals, or a performance on the card may have boosted another horse’s form. I have always found it so important to be flexible and modifiable – the most successful and strongest opinions are very often when you have changed your mind late on.

Success or Failure?
I am always asked how odds compilers are measured. What makes a good one? The current on-course starting price return serves as one tangible measure of accuracy, but in such a subjective occupation is it better to price a losing 2/1 favourite at 6/1, or a winning 6/1 chance at 2/1? Who is right and who is wrong? In the gambling industry, sometimes it is better to be lucky, rather than clever.

Glyn Warne is Fitzdares’ head of antepost and international horseracing.

Please play responsibly