To try to develop your own selection method can be quite an interesting, stimulating, mind-testing challenge. Whether your strategy is simple, or more complicated, actually developing a system of your own is a worthwhile and satisfying experience.

In the August issue of PPM, I outlined a trifecta system which revolved around horses going down in prizemoney the most (or up the least). I have found this to be quite an interesting selection method and one that regularly finds winners at longer prices in the first couple of selections.


I have noticed that if you take the first two horses in the pre-post betting market with the first two selections in the 'prizemoney drop' method, you will land many winners from these four horses, including good-priced ones. Boxed in quinellas for $3 per race (50c units) they can provide you with a cheap and interesting day's betting.

Achieving the elusive trifectas is a more difficult task. For instance, with trifectas, you may decide to take these four horses to win, the same four to run 2nd, with maybe two additional runners to fill 3rd spot (4x4x6) at a cost of $24 (50c units).

Let's say we decide to bet on fields of 10 or more runners. As an example, we will pretend there are 13 runners in the race on which we decide to bet. Then assume we are fated to get both 1st and 2nd placegetters. Statistically, only four of your original six selections remain (out of the 11 runners still 'live') to fill that 3rd position.

This illustrates the task you face, even in a relatively small field, despite the fact that you have bet $24 on 48 combinations! But if you want to win you have to plough on.

There is no right or wrong way to develop a system. Any system can fall prey to long losing runs. What you have to do, I believe, is quietly consider any system strategy in general terms and adapt it to your own way of thinking and investing.

Every month in PPM, the writers generate many system ideas for you to consider. They have been doing so for years. If selecting winners was easy, someone, somewhere in the world, would have 'invented' the perfect system well before now!

But I keep nibbling away at the problem. I will show you what I mean by system development, as we consider the previously unpublished rules of the 'prizemoney drop' plan.

RULE 1: Work on races of 10 or more runners. I selected 10-or-more starters to put value into the trifecta divvies. You may consider eight-or-nine runners and more are OK, but that fields of 13 and 14 and more make the task that much harder.

RULE 2: No hurdles, steeplechases or races run on straight tracks. You may also wish to exclude races for 2yo's and fillies and mares. Many favs win 2yo races and the form in mares' races can be topsy-turvy. Another strategy to reduce your number of bets is to invest only on races 1,100m and further. Another is to restrict them to between 1400m and 2100m, where the form generally is more open and predictable.

RULE 3: Delete all runners which have not raced in the past 27 days. You may prefer 14 days or 21 days as the cut-off.

RULE 4: Delete all runners which did not finish within five lengths of the winner at last start. I have selected five lengths, which I think will not be too significant for a horse dropping substantially in class. You may consider 4 or 4.5 lengths will eliminate many no chancers.

RULE 5: Where horses are equal in their prizemoney-change from the last start, sort into pre-post market order. I use the pre-post market because it is generally a reasonably accurate assessment of the ultimate starting price for city meetings. You may prefer barrier order, jockeys in order of their standing on the premiership ladder, highest average prizemoney per run, etc. There is no hard and fast rule.

Now let's see how the 'prizemoney drop' method could look with some additional considerations:

  1. Get a formguide which gives you the prizemoney of the previous race each runner contested. The Sportsman is ideal.
  2. Work on races of 9 to 15 runners.
  3. No jumps races, no straight track races, no 2yo races, no mares.
  4. Bet only on races between 1400m and 2100m.
  5. Delete all runners which have not raced within the past 20 days.
  6. Delete all runners which did not finish within four lengths of the winner last start.
  7. For each remaining runner, determine the difference in race prizemoney between the current race and the last start.
  8. Sort the horses into order of those going down the most in race prizemoney from their last start and/ or up the least in race prizemoney. Where runners are equal, sort in pre-post market order. If equal still, sort into saddlecloth (TAB No) order.
  9. To find your four trifecta selections for the win slot, take the first two favs in the pre-post betting market with the first two horses from the prizemoney drop list.

These final four selections should be the anchor for your trifecta bets. You may like to box the four for $12 (50c units). Or you can add more for 3rd so you have a 4x4x6 or 4x4x7, etc.

Let's look now at other angles for deciding on your trifecta combinations. The first deals with the impact of favourites.

The first two or three favourites will win, or be placed, in so many races that it is tantamount to betting suicide not to put them in somewhere in your multiple trifectas. The first couple of fancies in the betting will pick up around 45 to 50 per cent of all races. Leave these out of your winning chances and you are way behind the trifecta eight-ball before you start!

The other side of the scenario is that the first two favs do NOT win the other 50 per cent of races. These are the ones we are after if we have 'value' in mind.

Stick rigidly to the favs and you will struggle to inject enough value into your trifectas to make it a worthwhile battle in the long run. Tri's are difficult. much more difficult than many might believe.

Usually, it's only the occasional big dividend which keeps trifecta bettors ahead of the ledger.

Between these big divvies may be lots of small ones and lots of near misses. Generally, the less you bet on trifectas the longer it takes for any divvie, big or small, to come around.

To indicate how difficult it can be to land trifectas, I checked out results from a recent Saturday. The average tri divvie over 31 races was $1818 for a $1 outlay. Now, tri's do not average this much because they are easy to get!

We need to find ways to accommodate the favourites as well as one or two longer-priced runners. And, of course, we like to do it with as little form study as possible!

I have already outlined one approach in this article. Now, another one, this one using the favourites for the basis of the angle.

Knowing that the first three horses in the pre-post betting account for a high percentage of winners, I decided to have a look at the market in relation to a 3x3x7 combination.

To do this, I randomly selected eight weeks of old Saturday formguides, plus results from the Sunday papers. I then determined a number of key rules:

  1. No hurdles, straight races at Flemington or Victoria Park, no 2yo races where several runners have not raced before.
  2. Operate only on fields of 10 or more runners, although smaller fields (perhaps eight runners) would be suitable for a 3x3x7 combination.
  3. Operate on Saturday or quality holiday meetings in Sydney, Melbourne, Brisbane and Adelaide.
  4. Place runners in pre-post market order. If runners are the same price (on the same betting line) use the order printed in the newspapers, or the TAB order.

Over the test period, there were 21 races per week. You could reduce this further by eliminating more race classes, or work on one or two States only. Overall, the combination of 3x3x7 won on five of the eight weeks with a total gained of $705, an average profit of $88 per raceday.

Trifecta combinations which called for larger stakes fared even better. The following listed combinations were checked for seven weeks. I deleted the other week because a single trifecta paid over $1800 and would have given a distorted, though welcome, outcome.

COM BET
(50c)		 WIN-WKS
(from 7)		 PROFIT AVG
3x5x73051049150
2x5x824566094
3x5x62451070153
3x5x83641340191
Quite good results - plus the fact that I haven't included that more than $1800 divvie. You might now be saying, 'How long has this been going on' but it's not simple. If it was so easy, everyone would be doing it.

Long-term, combination 3x3x7, or any of the combinations, might be a winner using the pre-post but it's something that needs to be tested vigorously over several years of results.

What I do know is that it is a quick and easy method to find a solid trifecta linkup and get regular returns. If you look at the idea further you may find that it works better if you restrict bets to certain meetings, or States, or races with a certain number of runners.

My tests certainly showed that in the test period it would have been more profitable to have not bet in Adelaide.

Finally, another interesting method for determining your top tri selections:

  1. Start from TAB number 1 and move to the first runner which has won its previous two starts this campaign (two wins only, not three or more).
  2. If this runner is 5/1 or under in the pre-post betting, it is the selection.
  3. If this runner is longer than 5/1, move on to the next horse which has won its previous two starts; the same 5/1 rule applies. Ignore any pre-race scratchings which are likely to alter the pre-post market. Use only the original formguide pre-post prices.
  4. If no dual-race winner qualifies, move on to the next race.
  5. Suggested investment is for a win only, or I unit a win and 4 the place. The selection can also be used as a standout tri banker.

The same theory was tested with a cut-off of 3/1, not 5/1. From 32 meetings, there were 30 selections with 16 outright winners. From 32 meetings studied with the 5/1 cut-off, there were 54 selections with 30 winners, 8 2nd placegetters and 6 3rd placegetters. Only 10 of the 54 failed to run a place.

These are, I believe, quite extraordinary figures and I am continuing to run tests on the theory. I hope you do the same.

By Barry Meyer

PRACTICAL PUNTING - NOVEMBER 1997