Time Handicapping and speed ratings

Why did we bother with Par Times?

Speed ratings are the most popular form of rating races in the U.S.A. but have always been a “second fiddle” approach in Australia because of the intrinsic difficulties in establishing reliable speed ratings figures.

Anyone who has “had a go” at speed ratings will tell you that one of the major hurdles to overcome is to establish par times for distances at Australian race tracks. The way many people have approached this problem has been to use the track record times for individual distances at each major track. That didn't work.

Now the problem with that approach has been that these track “records” are not easily sourced and the circumstances in which they have been set may not be generally available for any other horses to emulate. It may be, for example, that on the day the 1200 metre record for Caulfield was set, the runners may have had a 60 kilometre an hour tail wind for the entire length of the straight. Did it? I don’t know and unless you were there and had an amazing memory, you probably don’t know either.

There are some amazingly fast track record times in Brisbane, for example, set by horses that most would regard as “dead set freaks” and “ordinary” horses do not go anywhere near equaling these times in normal day to day racing. It therefore follows that any horse that races consistently in events at these distances on these tracks will have a significantly poorer speed rating than those that, by chance, race at other tracks and other distances.

It means that if you are trying to be different from the “herd” and use speed ratings effectively, you need to be able to establish just what a good performance over 1200 metres on a dead track at Ballarat is as compared to a 1000 metre run at Wagga on a good track or 1100 metres on a heavy track at Benalla is with little room for error.

The other problem was in allocating class records - champion horses have to start somewhere so what if you're looking at a class record for a provincial track set by a soon to be champion before they hit the big time?

We took a rather simplistic approach to this when we put together our Par Times book a few years ago in examining examined many tens of thousands of races from the last ten years, established the ten best times for every distance, every track, every condition, averaged the result and come up with a “par time” that would equate to a good performance for every distance, every track, every track condition - where else were we going to start? So that was fine by us and at least gave us a starting point.

A recent correspondent who had got our Par Times wrote and said, hey, that's fine, but where do you go THEN? It seems there are many ways to use the par times as what there are ways to utilise data with traditional weight handicapping so if you are new to the speed caper here's an approach that may kick you off and get you started - a way to use past performances as an indication of what the horses may do in a current race - but - as we always go to great lengths to point out - YOUR approach and your thoughts will almost always be better than ours and provide you with more satisfaction when you're right.

We try to get an average of the last four performances for each horse we are looking at.

We have set up a little programme in Excel that you may like to copy and “play” with:
On an Excel spreadsheet, set up 6 columns titled, for example, race time, par time, btn margin, barrier, wt>55, rating.

Note: this is just a demo of an idea on how you may proceed - it is not intended to be the be all and end all of ideas - a starting point for consideration - your ideas are probably way better than mine!

Race time is the actual time of the race you are looking at expressed in seconds. In other words, if it was a 1200m race on a good track at Albany and they ran it in 1.11.1 secs, you would enter 71.1 in the race time column and 69.6 in the par time column (as indicated in this book as the par time for a good track for Albany)

If the horse was beaten 3 lengths, enter 3 in the btn margin column. Enter the barrier it started from in the barrier column. We also enter any weight it may have carried OVER 55kgs. In other words, if it carried 58kg, you would enter 3.

Ah, you say, but how does it calculate the speed rating? The formula to enter in the rating column is the key to it. You may need assistance with this if you are unfamiliar with how Excel works. However, the formula is as follows: (this would go in your column F - the rating column - as the calculating formula)

=((100-(((A2+(C2*0.16))-B2)*10))+(D2/5))-E2

To roughly explain it, if running a par time equates to 100, (I guess like a 100% effort) we deduct 1 point for each one tenth of a second more than par time that the actual race was run in, after allowing an extra .16 sec for each length beaten, make an allowance for the barrier it started from (regardless of track), and also make an allowance for every kilo carried over 55kg, which is generally accepted as the level at which weight counts against performance. Generally, this column is only used in races over 1150 metres.

Now you may prefer to use your own formula – this is merely a thought starter to get you going. Another consideration is the way that some race distances are varied to even out wear and tear on the track so while the race distance should have been 1000m they run some really weird distance like 1011. This seems to be more of a problem in Victoria. The scale we use to arrive at a modified race time is as follows:

3 metres .1 sec
4 metres .2 sec
5 metres .2 sec
6 metres .3 sec
7 metres .3 sec
8 metres .4 sec
9 metres .45 sec
10 metres .5 sec
12 metres .6 sec
15 metres .7 sec

These times are loosely formulated - no - they are not EXACT but formulas with 6 and 7 decimal laces are really messy to work with (and we believe unnecessary).

The generally accepted belief is that 6 lengths equals 1 second or 1 length equals .16 of a second. So you table of margins would look something like this:

2 lengths equals .32 secs
1.5 lengths equals .24 secs
1 length equals .16 secs
half length equals .08 secs
.3 length equals .05 secs
long neck equals .04 secs
neck equals .03 secs
head equals .02 secs
anything less equals .01 secs

Use that as a rough guide and apply it CONSISTENTLY and you will be pretty close to the mark. Note the two tables are NOT in correlation to the actual physical length of a horse BUT the key is to apply them consistently as is the case, I guess, with all data relating to horse racing. You can't beat consistency of application

The formula can contain ANY number of variations you may like to try and combine your speed assessment with more traditional weight handicapping factors provided you bear in mind that you are working towards a 100% assessment of a horse's chances.

A lot of people will argue till the cows come home that speed assessments are dependent on too many other factors (like race pace and jockey tactics) to be relevant. They well may be right. And we agree that using just one race to base your speed assessment on is totally inadequate. The argument over speed versus weight handicapping will never end.

It may be worthwhile remembering that while all those charged with bravado and certainty were yelling out "go west young man!" the smart ones went east - and got all the good looking women!

© RaceRate.com If you wish to use this article on your web site, feel free as long as it is complete including this copyright notice