Seven years on . . . a new look at a slick idea
PAIR SHAPED HANDICAPPING

In March, 1998 I wrote my first article for PPM, called “Pick ‘Em In Pairs”, in which I detailed my Pairs Analysis Technique, or PAT for short.

This technique centered around matching TAB 1 with TAB 2, TAB 3 with TAB 4, TAB 5 with TAB 6 and so on in what basically ended up being match races between each pair.

Once I had determined the “winner” of each pair I had to evaluate what I considered would be the beaten margin between the two in a real match race. The example I presented in that article comprised of a field of eight past champion class horses one could only dream about seeing in one race.

In a field of eight runners there will be four pairs (I might add in a field of odd numbers the last pair will actually be a trio; in a field of 9 you would compare TAB numbers 7, 8 and 9 with margins between them as if the race was a match race between the three). So, after comparing one another there will be four A selections which are the horses we initially continue to consider.

Let’s provide a field to make this clearer. The field is:

TAB 1: Tulloch
TAB 2: Gunsynd
TAB 3: Phar Lap
TAB 4: Rising Fast
TAB 5: Kingston Town
TAB 6: Dulcify
TAB 7: Gloaming
TAB 8: Archer

In this example we assume the winners of each pair were TAB No’s 2, 4, 6 and 8 which were Gunsynd, Rising Fast, Dulcify and Archer while the “losers” were Tulloch, Phar Lap, Kingston Town and Gloaming.

Let’s say we calculate that:
… 2 Gunsynd beats 1 Tulloch by 0.5 length
… 4 Rising Fast beats 3 Phar Lap by 0.75 length
… 6 Dulcify beats 5 Kingston Town by 1.0 length
… 8 Archer beats 7 Gloaming by 1.25 lengths

The “losers” are no longer considered as they have been eliminated as contenders for the FINAL A selection. If you are a one horse per race punter you will need to create another set of pairs from the “winners”.

Thus, your new pairs are TAB No’s 2 & 4 and TAB No’s 6 & 8 which means you are comparing Gunsynd against Rising Fast and Dulcify against Archer and again you must provide a beaten margin for each pair.

This time we calculate that:
… 2 Gunsynd defeats 4 Rising Fast by 0.75 lengths
… 6 Dulcify beats 8 Archer by 0.25 lengths

The next step is to apply the same process to Gunsynd and Dulcify with Gunsynd beating Dulcify by 1.0 length. There should be no after the race excuses as your final A selection has been calculated by thorough form analysis. But, what if the punter wishes to have more selections per race and wants to calculate a priceline?

Well, the hard work has been done along the way and it is a simple matter of organising the margins into a format ready for the required calculations. Starting from Gunsynd we find the following order:

By using the Don Scott tables from his books we can assess a priceline for each runner based on a market of 100 per cent.

In order to set your priceline you will need a simple calculator. Firstly, add up all the figures in the Deci column which in the example on page 10 equals 4.86.

Next multiply the Deci for a horse by the percentage you are setting your market to (100 per cent) and divide the result of that calculation by the total of the Deci column (4.86) which will give you the figures representing the percentage chance the selection has in this field.

For example, using Tulloch as an example, we see the calculations as 0.80 times 100 divided by 4.86 equals 16.46 per cent which is about 11/2 ($6.50). If, for example, you set your market to a tighter 80 per cent the calculations for Tulloch would be 0.80 times 80 = 64.00 divided by 4.86 equals 13.16 per cent which is about 13/2 ($7.50).

The setting of the percentage market you bet to is an individual decision. Most punters I have known over the years have used Don Scott’s recommendation of 80 per cent which gives you a 20 per cent margin against the bookmakers/public opinion.

Don Scott provided what he called a “Table of Advantages To Chances” which enabled you to work out each horse’s price provided you have established an estimated margin in kilos between each runner remembering that 1 length equals 1.5kg. Scott’s table is as follows:

I have now added another feature to tighten up the pricing process I have just shown you. If we assume I have calculated the prices for all my A selections for an eight race program in the same way, my A selections for a whole program might read like this:

Race 1 – Gunsynd  (the A selection of the example race above)
Race 2 – Maybe Mahal
Race 3 – Sunline
Race 4 – Super Impose
Race 5 – Northerly
Race 6 – Makybe Diva
Race 7 – Schillaci
Race 8 – Better Loosen Up

Although I priced Gunsynd at 15/4 ($4.75) I might ask the question, “What sort of value is Gunsynd when I look at the program”?

The way to determine this a fraction better is to apply the PAT method to Races 1 & 2 (Gunsynd versus Maybe Mahal), Races 3 & 4 (Sunline versus Super Impose) , Races 5 & 6 (Northerly versus Makybe Diva) and Races 7 & 8 (Schillaci versus Better Loosen Up).

Your calculations show that Gunsynd, Sunline, Northerly and Better Loosen Up are the “winners” of your initial PAT and, of course, you then compare Gunsynd against Sunline and Northerly against Better Loosen Up.

By following things through you will, eventually, have your Best Bet of the program with a price beside it as well as your second best bet and so on for all your A selections.

You could try another method where you divide the program into halves by comparing a quadrella of races instead of pairs of races. For example, you could compare Gunsynd, Maybe Mahal, Sunline and Super Impose as if they were all in one race and Northerly, Makybe Diva, Schillaci and Better Loosen Up as if they were in one race and apply the beaten margins process to determine your prices for each half of the program.

The reason I mention dividing the programme into two evolved from a study I did for a series of articles for PPM where I recorded the selections of the Paper Poll Favourites in The Australian newspaper and found that Paper Poll Favourites won about 50 per cent more races in the first four races of an eight or nine race program than in the second half.

I haven’t revisited that territory since then but I would be quite surprised if similar figures were still not the norm.

It makes sense that race clubs want the tougher later races as part of quadrella legs, hence the races with small fields or few genuine chances are generally in the first half of the program. There is nothing stopping the punter from operating only in those easier races and applying PAT accordingly.

In all my years of punting I haven’t seen a method of pricing that is as easy to apply as the PAT method that actually still has you, the punter, doing the form in a traditional manner.

There are a multitude of “points allotted” types of analysis where you give X points for days last raced, Y points for certain jockeys, Z points for last start winners or Don Scott type of ratings which are methodical BUT do not take into account your overall feeling about the horse’s courage.

How would you feel if your points system had a horse you considered a cat as your top selection but your heart, and an aspect of form you value, told you another horse should have been the selection and it wins? You and I both know what the answer is to that!

With the PAT methodology you can still have your “heart” selection come out on top unless there is serious compelling evidence to the contrary.

Over the years I have found it is quite surprising how quickly you can arrive at an A selection. When you seek further selections in the race the process becomes a fraction more involved and even more complex when you want prices against each chance but overall the most you will need is a simple calculator.

It is my suggestion, if you wish to become more professional in your punting approach, you give PAT a try; I am confident it will serve you well.

By Roman Kozlovski

PRACTICAL PUNTING - OCTOBER 2005