The leading US racing media commentator, Andy Beyer, is rightly known as the 'guru' of speed handicapping. He says that he quickly realised that calculating 'speed figures' was the most powerful tool in the game.

He went on to become the racing world's leading advocate for time handicapping, and yet when he came to Australia and applied his speed ratings to our racing scene he ended up with egg on his face. They didn't work. When he did land winners they were mostly favourites; everyone else had chosen them as well!


So can speed handicapping work in Australia? Well, for many punters - serious ones - it is working. And it can be made to work for any punter, provided they are prepared to put in the 'hard yards' and use 'speed' as the basis of form analysis.

There is an argument that speed handicapping is not conducive to Australian racing because of the wide diversity of our racetracks. Critics point to the uniformity of American racing as underpinning the success of speed handicapping.

Over there, a punter can bet at the same track, day after day for two or three months, while in Australia the punter will be betting at Moonee Valley one day, Flemington the next and so on.

Nevertheless, there are ways to determine speed rating figures which can be used as an accurate guide to form analysis and final selection. The Rating Bureau's Dennis Walker has perfected the use of speed ratings in his ratings on the WinLine computer service.

Says Dennis: "I started developing a more than casual interest in time handicapping some time back, around 1978. I had read an article in a racing magazine which included a list of the major Sydney tracks and regular racing distances. The table published was incremented in 10ths of seconds and allowed you to compare the run of one horse in a given time over 1100m at, say, Randwick with the run of another horse over 1100m at Rosehill.

"I found that my task was to determine a standard (par) time for every track and distance, not only at metropolitan tracks but all tracks in Australia. This involved developing and testing a list of standard times for every distance, every track.

"My first and only reference point was the existing track records. I found that many were established by mediocre horses so I had to go a lot further creating more of a list of possible track records; some were unrealistic in terms of expecting the time to be repeated. The second task was to establish a sensible and sound way to compensate for track conditions. I then had no computer and no access to a database. The track condition as given in the
Sportsman was all I had to go on, so I had to rely on the official track condition.

"My first naive attempt involved using a factor for each official condition; fast. good, dead, slow and heavy. For example, let's say a horse ran 1m 14s on a slow track. I would divide the winner's time by a factor of 1.015, which would 'normalise' the time back to 1m 12.9s which was what I would have expected this horse to run on a good track.

"This presented many frustrations, as there was no way of determining the extent to which the track was slow. It was sometimes referred to as on the better side of slow, or almost heavy, etc.

"I no longer need to use these approximated methods to determine track condition; our fabulous Rating Bureau database has seen to that problem! The third task was to determine when we are given wrong information on race distances or winning times usually resulting in over-ratings.

"This was the most difficult hurdle for me to overcome, as much of the pre-race analysis was subjective. I have now manually adjusted approximately 20,000 races involving long hours staring at computer screens and tinkering with the keyboard.

"The tools I now use have been designed to create more accuracy in results from pre-race ratings."

Dennis says there are still frustrations to contend with when making use of the speed ratings. He needs to be only 3/10ths of a second (about 2 lengths) under-assessed from one race and there can be three or four future ratings affected by horses from the same race.

He adds: "But in my opinion, time or speed handicapping has been much maligned over the years. For as long as it has been written about in Australia, experts have negatively commented on times as a reliable tool."

As anyone who has read the late Don Scott's writings will know, he tended to dismiss and even ridicule time handicapping as a basis for rating horses. Scott offered five reasons:

  • Times at different tracks vary, partly because no two tracks are identical in shape and layout.
  • Track conditions are never exactly the same from day to day and are somewhat different on the same day; winds are always springing up, headwinds, tailwinds, crosswinds, etc., and even on good tracks the surface can vary according to the length of the grass.
  • Distances may not always be exact.
  • The official times may be quite inaccurate (reference the recent controversy over the time registered by Chief De Beers at Doomben).
  • If the pace is muddling, the final time will be much slower than average, especially for races beyond 1600m.

The supporters of speed handicapping acknowledge all the difficulties but maintain that, with careful application, race times can be incorporated successfully into selection processes.

Dennis Walker says: "The bottom line is a comparison of recent runs of each horse and extending this to possibilities, assuming the base information is correct to begin with. For simplicity, I suggest PPM readers try the following; it will give you a bit of fun and I promise it will find some good winners. It is how I started doing simplistic time ratings some 20 years ago.

"Let's assume that we have assessed our standard (par) time for the 2000m at Rosehill as being 2.00.0 - two minutes - or 120 seconds.

"We then treat that standard time as being worth 100 points and (even though it may not seem mathematically correct, don't worry about that) every one of those points as being equal to one-tenth of a second.

"Also assume, for this exercise that 6.5 lengths equals one second, or 10 points - i.e. I length equals 1.54 points.


"Now let's compare a couple of horses based on our speed formula - say, Intergaze and Octagonal.

  • Intergaze 15 / 3 / 97 at Rosehill over 2000m is beaten 0.2 lengths in 2.00.6 (120.6) which means the winner failed to reach our standard (par) by 0.6 seconds. "This, according to our time/lengths formula means that we must deduct 3.9 points from the 100 POINTS standard that we have set, plus a further one-fifth of 1.54 points (see above as equalling a length) to allow for the margin Intergaze was beaten. That's an additional 0.31 points.

"Therefore Intergaze's speed rating becomes 100.0 - 3.9 - 0.31, which equals 100 - 4.21, or close enough to 96.

"Now let's do a similar workout for Octagonal using 2400m at Rosehill and 2000m at Flemington.

"My time standard for 2400m at Rosehill is 2.26.4 (146.4) and for 2000m at Flemington it is 1.59.8 (119.8).

  • Octagonal 22 / 3 / 97 at Rosehill over 2400m wins in 2.27.4 (147.4), which is exactly one second outside my standard rating.

"Using our formula we deduct 10 points (for one second) from our optimum points score and we get a rating of 90. Since Octagonal won there is no beaten margin to deduct, so the rating remains at 90.

  • Octagonal 10/3/97 at Flemington over 2000m wins in 2.01.0 (121.0), which puts him 1.2 seconds on the worse side of my standard rating.

"Using our formula, this equates to 12 points being deducted from our optimum 100. Since Octagonal also won this race, there are no extra points to be deducted for a beaten margin, so his rating for this race is 88.
 
"I hope this simple example gives you some basic insight into how to start using times. The secret in this particular example comes down to accuracy in your 'par times', which is quite subjective. To get more accurate results in the long term there are other variables you must consider, not the least of which is the track condition and the accuracy of the winner's time and the distance of the race."

The British expert Nick Mordin has written many articles and books on speed ratings. Obviously, he is a 'true believer' when it comes to handicapping on times.

In his most recent book, Nick writes: "The theory behind all speed ratings, including mine, is extremely simple: to tell how fast a horse can run today, you merely have to know how fast it has run before."

Like Dennis Walker, Nick concedes that to make it all work you need to arm yourself with a 'set of standard or average times that a hypothetical horse would achieve at every distance on every track'.

He says it's no easy task because you can't simply add up all the times taken for a particular trip at a certain track and then work out the average. That would produce a misleading result for two reasons:

  1. There is a huge difference in the quality of horses who race at each track.
  2. In many races, horses don't run fast enough in the early stages to achieve a representative final time, and this is increasingly so as the distances of the races lengthen.

Nick overcame the problem by careful study of races at every distance and noting the various Classes of race. He was finally able to come up with a set of 'standard' times that were as accurate as he could make them.

Ronald Firth, a US-based business consultant, has written extensively on speed handicapping in a small newsletter distributed among racing fans in California. Firth believes the best way for ordinary punters to 'cash in' on a form of time handicapping is to use recent times as the standard.

He explains: "Take any recent meeting and write down the BEST time for each distance. Use that as your standard for the next meeting. Using standard times from years past can be fraught with inefficiencies. Using recent times you keep up to date with what the horses are running NOW, and not what they were running 10 years ago.

"With the constant updating of standard times you will be able to keep pace with changing strengths in the Classes. For example, if there were five 1200m races on a programme and they were run in 70s, 70.4s, 71s, 71.2s and 70.3s, you take the 70s as the standard.

"Every horse in the 1200m races can then be given a speed rating using that as the base. It's so easy. It saves you having to have a huge par times base."

Firth, of course, worked out his approach on the basis that the horses would be racing at the same track at a long Californian meeting, such as those at Hollywood Park or Golden Gate. But his idea can still be effective in Australia, provided the user takes into account other factors.

The a roach, then, is to work out each horse's rating after a meeting has been run. You use Dennis Walker's 6.5 lengths to one second rule, etc. It is easy then to decide a speed rating for each runner.

Let's assume an 8-horse race at, say, a city track somewhere in Australia. The winner ran 71s for the 1200m, and this was the best of the card and so becomes the standard' time.

Obviously, the winner is going to get 100 points. If he won by 2 lengths, that means the 2nd horse has a rating of 97. The third horse may have been beaten 3 lengths, which means a rating of 95.4. And so on.

Remember that one length is the equivalent of 1.54 points. Thus, if the last horse, say, was beaten 9 lengths, his rating would be 100 minus 9xl.54 equalling an 86.1 rating.

All very simplistic, but certainly a starting point for the tyro speed handicapper trying to get a grip on what it's all about. If you stuck to sprint races up to, say, 1400m you should enjoy some good successes with such an approach, provided you didn't rely solely on the speed ratings as the selection arbiter.

Obviously, you would need to take other factors into account. But they would provide you with an important lead. Naturally, this is an approach that will suit not all punters. Many couldn't be bothered but there are many out there who are hungering for a fresh approach that might give them an 'edge' over their fellow punters in the battle for the gambling dollar.

If speed handicapping ratings can provide you with a set of pretty accurate figures, then you will have gone a long way to solving the form problem in races, no matter where they are run.

  • MORDIN ON TIME, by Nick Mordin (Aesculus Press)
  • BEYER ON SPEED, by Andy Beyer (Houghton Mifflin).
  • THE RATING BUREAU, contact Dennis Walker.

By Dennis Walker

PRACTICAL PUNTING - JUNE 1997