0ver the past few months I've had the pleasure of browsing through some old racing publications. I could remember some of the articles even though they dated back to the early 1970s.

Buried amongst this treasure trove was a most interesting target betting staking plan in a book called Al Illich's How To Pick Winners, which I decided to tinker with, on paper, in order to test the author's concepts.

The original plan was based on backing two to five horses in a race with the objective of winning $2 per race. However, regular PPM readers know I basically believe that the punter should start with 10 cents per race as the objective. There's always the option of increasing objectives when the punter develops confidence in the inevitable bet size increases and if the bank increases. (As it turned out, this was very good advice.)

It's not shown by the author how a punter will decide the number of runners to back but I am assuming the figure is derived from the number of chances on form. The issue of form selections is a MOST important part of the plan as this article will show.

To explain his plan, the author begins by stating: "The idea is to play to win a given amount per race. In this workout it will be $2. To determine the amount to wager, divide the objective by 8 and wager the resultant sum on two horses in the same race. Both our horses lose. We add the loss of $4 to be won on the next race, the debit, plus $2 to be won for the next race, making the objective $8 for race 2."

When a winner occurs, the profit from the race is deducted, the objective of $2 is added to the objective for the next race and the objective is still divided by 8. The overall premise of this plan is that once you back a winner at 9/1 backing two horses, 10/1 backing three horses, 11/1 backing four horses and 12/1 backing five horses, you will have cleared the objective or be close to it.

According to the author, this is worked out by adding the number of selections less 1 to the 8/1 divisor, e.g. 5 chances minus 1 = 4 + 8/1 = 12/1.

This does not always work out, however, due to the fact that often we should be betting slightly higher amounts than we actually do because of fractions, but eventually you will totally clear the objective column. The most apparent problem with dividing by 8 every time is that seldom will you be backing 9 / 1 winners in races where you believe there are only two chances and, to a lesser degree, in races with 3, 4 or 5 chances as well, so a change needs to be made.

The author acknowledges this by stating: "When contemplating a play of this type the idea is to find out which figures to use to divide into the objective. If your selections only produce very occasional 9/1 winners, then go to a lower figure so as to cancel the objective more often."

This makes sense, so anyone contemplating this plan needs to think deeply about their past strike rates and winning dividends in order to select an appropriate divisor.

What I needed to formulate was a realistic divisor for the number of horses I was prepared to select in a race and what races to bet on.

I decided to test a real wildcard, one I have often thought about investigating more thoroughly in the past ... the top TAB numbers.

For over three decades in nearly every State of Australia TAB No. 1 has won more races than TAB No. 2 and TAB No. 2 has won more that TAB No. 3 and so on for about the top 5 TAB numbers. Further down the order, changes occur more regularly due to many horses being on the same weights or being on the Limit.

For the three-month period covering May, June and July, 2003, I logged all Saturday meetings in SA, Victoria, NSW and Queensland by listing all HANDICAP races and the number of starters.

I decided to concentrate on field sizes of 9, 10, 11, 12 and 13 and made up a chart as to how many of the top TAB numbers would be selections. I did this by dividing the field size in half, dropping any fractions and subtracting 1, thus providing this simple chart.

9 starters = 3 horses
10 starters = 4 horses
11 starters = 4 horses
12 starters = 5 horses
13 starters = 5 horses

What this meant was that, after scratchings, I was backing the top 3 TAB numbers in fields of 9, the top 4 TAB numbers in fields of 10 and 11, and in fields of 12 and 13 the top 5 TAB numbers.

The issue of what divisor to use became my next problem (you will remember the author also considered this a major worry) and eventually I settled on 10/1. This covered a broader spectrum of odds than Illich would have contemplated in his original article as he was most likely considering much more fancied runners, but I figured 10/1 was a good starting point for this experiment.

The broad statistics showed there were 218 races for 119 winners (54.58 per cent) covering 929 starters, for a flat-stakes return of $433.15, based on 50-cent units, and a loss of $31.35, a 6.74 per cent loss on turnover.

A simple target betting plan of chasing 10 cents per race and a divisor of 10 yielded a profit of $301.35, based on turnover of $5097.50, for a 5.91 per cent profit on turnover. Yahoo! The plan works.

Yes, it does BUT: the profit came with some heart-stopping moments. At one stage the staking plan called for an outlay of $484 (4 horses at $121 each) with the Gods smiling at last with an 8/1 winner, thus reducing the next race outlay to $171.

Another two winners (6/4 and 9 / 2) followed with two more losing races and, eventually, at Bet No. 173, a juicy 14/1 winner cleared the ledger. By this time, suicide would have seemed an option!

It became clear that dividing by 10/1 was not high enough so I applied 15/1 as the divisor to Bets 129 to 196 (the area of problem). The Mt Vesuvius-like monetary explosion was halted but still the maximum outlay was $120 (5 chances by $24) and profits were blown out of all proportion when a 40/1 winner was landed, thus exaggerating the profit on turnover.

I am sure that dividing by 20/1 would be safer to one extent in that the increase of bet size will rise more gradually but, by the same token, you will need a higher priced winner to clear the ledger. It's a bit like being between a rock and a hard place!

If I had actually bet on the selections I would have applied "brakes" such as halving the objective or adding an extra 10 to the divisor to slow down the rising of bet size while hoping for the right-priced winner or winners to arrive.

As I mentioned earlier, this was an "on paper" experiment (just as well or I may have died of a heart attack by actually operating it) to test the author's overall concept of backing several runners in a race and using the one divisor to cover all runners.

I had never seen this concept in print and felt it worth a try on some sort of "sensible" approach. I could backtrack without bias of knowing the results; thus the experiment on the TAB numbers. The problem, of course, with just selecting TAB numbers is that some of the horses would be out of form or some could be better-than-average stayers resuming in sprints, so defying what I believe the author's original concept meant; that is, betting on
form horses.

It's my intention to do another study of this concept but to stick with the original idea of backing inform horses, as opposed to the "wildcard" TAB approach, by testing the published selections of PPM contributor E.J. Minnis. He selects five, sometimes six, horses in races where the formlines are more obvious and where there is a possibility of value.

In the meantime, can I advise you to tinker with the concept on your own past selections by using different divisors and safety brakes or whatever. Although my TAB numbers experiment "worked" in one sense, the reality is you would have been terrified as it progressed. However, overall the concept has promise for those who use target betting methods.

20031018

By Roman Koz

PRACTICAL PUNTING - OCTOBER 2003