It's often been said that the art of betting is not really about picking winners. It's all about how you can pick enough winners at sufficiently good odds to allow you to make a profit in the long term.

Anyone can pick a winner. A little old lady with a pin could probably nail a winner every six or seven stabs! But, in the end, such an approach would be doomed. What a punter needs is the skill to find enough winners at the right prices.

The little old lady with the pin, even if she could stab one winner in six, would need to average odds greater than 5/1, but actually she would be lucky indeed to expect an average of more than, say, 7/2. She'd be in for a long-term loss.

Recently, a PPM reader sent me the results of some English research into three methods of selecting greyhounds. Now, of course, the figures relate to English greyhound racing (six runners per race) but they might equally apply to any form of betting. The research involved 1000 races and looked at (a) the favourite; (b) a Press expert's selections; and (c) a random selection made by a computer.

The favourite won 33 per cent of the 1000 races (329 wins) and there was a loss of 19 per cent (each return, by the way, calculated on the basis of settlement at an off-course licensed betting shop at SP, tax un-prepaid, i.e. with a 10 per cent deduction from total return for tax). The Press expert struck 287 winners (29 per cent) and lost 21 per cent, while the computer's random approach got 166 winners for a loss of 17 per cent.

As you can see, selecting the most winners doesn't mean you are better off. In this study, the computer got far fewer winners than the SP favourite and the Press expert-yet emerged with less of a loss at level stakes!

Always bear this in mind when you are ploughing into the favourites and the tips of your favourite newspaper expert: The chances of making a long-term profit by following them is not very good at all. So stop and think before rushing in to back that 6/4 chance and look at it in terms of your long-term financial prospects.

You simply must come to terms with your ability to pick a certain number of winners per certain number of selections. If you can pick only two winners every 10 bets then you have to realise that you need to average 4/1 each winner just to break even. There is no point in backing two winners at 6/4, or even 2/1, or 3/1, because you'll still end up losing.

If you expect to make money from 6/4 chances, you have to strike with four winners from every 10 bets to break even, and you'll need a fifth winner in the 10 bets in order to make a profit (five winners at 6/4 would return you 12.5 units, for a profit of 2.5 units.

Now, in answer to questions from some RPM readers, a couple of staking plans that should prove of value to those among you who are good selectors, and who can ensure, that 'the price is right' when the money goes on.

Plan A is for backing two horses per race. Some readers may be aware of it already. It's virtually a safety-first approach for the small-bet punter. You start with a bank of 100 units and you bet 1/20th. That would be, to begin with, five units. You bet $3 to win on your top selection and have $1 each-way on your second pick.

Increase your bets when the bank has reached 120 points. Thus, you would then be betting 5 per cent of that total (six units). When the bank reaches 140 units you increase the stake again (seven units). Always bet in the same ratio-3/5ths of the total on the top selection, the rest each-way on the second pick.

The second method is a really interesting one, and it assumes that you never bet unless you have two or three chances of a collect. Your starting bank is, say, $50 and your initial bet is $10. If it didn't win, your bank is $40 and the next bet $8.

You can approach the betting in a number of ways. Firstly, if you have a race in which you are confident of only one selection then you back it for a place. That gives you three chances of a collect. In another race, you might have narrowed the field down to two chances, or maybe three. In the case of two chances, you can make level stakes bets on both, or back them both to return the same amount of profit. Another approach is to save on one and put the rest of the bet on the other one. You can do the same with three chances in a race.

Example A: You fancy only one horse. It is priced at 4/1. You put your $10 on for a place. The horse runs 3rd. You get back $20 (evens for the place), a profit of $10.

Example B: You like two horses in the race, and you cannot split them. One is at 5/2, the other 3/1. You put $5 on each. The 5/2 chance wins. Your return is $17.50, a profit of $7.50.

Example C: You like two horses in a race, but you fancy one of them more than the other. One is 3/1, and the more fancied one is 2/1. You place $3 on the 3/1 chance to 'save' the stake and put the rest of the money, $7, on the more fancied runner. If that fancied runner wins, at 2/1, your return is $21 for a profit of $11.

Example D: You like 3 chances, but one of them appeals more than the other two. This fancied runner is 5/2. The other two are 6/1 and 7/1. You 'save' on them, putting, say, $2 on each, and then put the remaining $6 on the 5/2 chance. If it wins, you get back $21 for an $11 profit. If the 6/1 horse wins you get back $14 for a $4 profit, and if the 7/1 wins you get back $16 for a profit of $6.

These are just a couple of staking plans that can provide some succulent returns for you-as long as the prices are right. Remember the rules and you can't go wrong.

The multiple approach is an ideal one for the cautious punters among you. It ensures that any losing runs can be kept to an absolute minimum. The returns can be very good, no matter which approach you adopt.

By Jon Hudson