As I write this, so far this calendar year there have been 128 races of the types that I like to recommend and prefer to bet on. I'm talking about the Group races (1, 2 and 3), the Special Conditions races and the Listed races.

I'm still suspicious about the 2yo races that sneak in here but have included them for the sake of this article. The period in question covers 26 weeks with an average of five bets per week, an ideal situation for all the good races.

An analysis of all the winners showed the total return on NSW TAB dividends was $1228.30 for an outlay of $128 ($1 units). The profit was $1100.30 which means there was a lot of margin for error and you could still have come out way ahead.

There were no 100/1 winners and the actual breakup of the win prices bears looking at more closely.

Between $1 and $3 - 26 (20 per cent)

Between $3 and $10 - 57 (45 per cent)

Between $10 and $20 - 32 (25 per cent)

Between $20 and $30 - 6 (5 per cent)

Greater than $30 - 7 (5 per cent)

The longest-priced winner was $44.60 and there were only three of them in the $40 range.

The first two sets of winners are in a price range that I consider is pickable, obtainable, etc. That is, 65 per cent of the winners of these races are up for grabs. The trick is to know, or at least come up with, a plan to make some money on this information.

I'm going to make an assumption here that may not hold up, so don't say you haven't been warned.


The total win dividends above were $1228.30 and 65 per cent of this amount is $798.40. To be sure we get the winners I'm suggesting, you can bet on four horses in each race for a total outlay of ($4 multiplied by 128 races) $512.

Remember this is outlaid over 26 weeks so the betting is about $20 per week. The theoretical profit is ($798.40 minus $512) $286.40 or $11 per week. All unit bets here are, of course, $1.

Okay, that's the theory and the next step is to come up with a plan for picking the four horses you'll back. You may already have a plan for this. Keep in mind we're betting on the best races where the pools, and therefore the win dividends, are better than average. The excitement of having four good horses running in the classic races is also a big bonus and thoroughly enjoyable, especially when you win.

In the December 1999 PPM, I presented a method called the WARP system in which I determined four factors that will produce the required selections for this plan.

The method suggests that you identify the horse with the highest Win Ratio, Average Earnings, Rating (Zip or other) and Place Ratio. if there's a tie, go to the next one down so that you end up with four horses for the race.

I'm also pretty sure that the prepost betting market will allow you to identify the best four horses.

If you're unhappy with your progress, or want to add some genuine excitement to your betting, try this one out. As I've said, these really are the races to be involved in and having a plan that allows for four bets and a potential 65 per cent race collect ratio is going to keep you happy a lot of the time.

The other thing worth a mention is the 'repeaters'. You see, 16 horses came back to win a race (within 30 days mostly) again. In the case of Mr Innocent and Tie The Knot, they won four races each. Twenty-three of the winners actually came back for another bite of the cherry and, ignoring the first win dividend, the subsequent wins were a breakeven situation. Ken's Joy and Sunline were not disgraced with three wins each!

There's a lot of satisfaction to be gained by ignoring all but these types of races (Group, Special Conditions and Listed) and really concentrating on working out the best four horses to bet.

You learn a lot and in the case of the 2yo races I'd be ignoring them completely. If you prefer to be more restrictive, then leave out the Listed races.

I wouldn't do that myself because, most of the time, the good form is there. Certainly, though, when in doubt about a particular race, leave it out.

Doing the numbers you will require an average win dividend of $6 to break even and about 80 winners from the 128 races to obtain the 65 per cent race collect ratio. But you're betting on four horses in each race so I believe it's achievable. Let me know what you think.

by Keith Roth