The battle between level stakes betting and some form of 'fancy' staking is one that rages on wherever racing and betting is discussed. Naturally enough, it's an issue that affects us all every time we have a bet. We can pick a winner, but do we back our selections in a way that maximises our wins and minimises our losses?

It's a vexed question. I have always believed it desirable to regulate stakes. Now this can take many avenues. You can regulate yourself to bet solely for level stakes, or you can map out another path for yourself, using sensible progression, or target betting (like the 6-Point Divisor outlined in the April PPM).

Most punters bet by whim, on impulse. They figure a horse's chance in their mind and, depending on their feeling of satisfaction with it, they nominate a sum. Sometimes this will pay off, sometimes it won't. My own experience has been that I usually bet too much on losers and not enough on winners (common complaint, I'm sure).

I think if you look back on your betting you will find that more often than not it corresponded to my own fatal pattern. What we can deduce, then, is that it's not the winners you back that matters in the long run, but how you back them.

So, let's assume that we all agree on one vital point: Proper staking is going to benefit you more than haphazard staking. Regulated staking has to be regarded as being a superior form of betting.

Progressive staking has much to admire about it - but only when it is used cautiously. You can never double up after losers for example; that's suicide. And a rapid increase in stakes after a loser is also a nono.

All staking plans which call for an increase in stakes after losers run a risk. It's all a matter of to what degree the risk is allowed to be. A formula calling for a gradual rise in your bets total against a losing run will usually improve the overall cash position in the bettor's favour when compared with level stakes, but there certainly will be times when the reverse applies.

This is when a lengthy losing run strikes! But let's look at things in a positive manner. How can we harpoon the prices, so to speak, and bet in an orderly fashion and give ourselves a fighting chance of emerging a long-term winner?

I think we have to split our bets into price brackets and bet accordingly. The shorter the price the more confidently we can frame our stakes progression.

My idea is for a series of three brackets, which assume your bets are going to be from odds-on through to around 10/1 and no higher than that. We can look forward to a high strike rate if we stick to very short-priced horses.

If we say our bets will be from, say, 4/5 through to about 5/2, we can hope to crack a winner every two bets, with a maximum losing run of around 5 or 6 bets. This is my estimation, anyway.

On this basis I am recommending a progression which goes:

1 - 1 - 2 - 3 - 4 - 5.

That allows for 6 bets and if there are that number of losers (phew!) you switch back to a new series, starting with a 1 unit bet again. As soon as any progression is in profit, no matter how small the sum, a new series is begun.

The worst scenario is a complete wipeout of all 6 bets. But if you had backed 5 losers and were going into the 6th bet, you would be losing 11 units. With a 5 unit bet on the final leg you would need a winner at around 2/1 to just about break even (almost!).

For the middle-priced selections, say horses between 3/1 and 5/1, with an expected 25 to 30 per cent winners, the maximum losing run, I estimate, would be 9 or 10 (some may argue with this, but that's been my experience).

The suggested progression for these selections is:

1 - 1 - 1 - 1 - 1 - 2 - 2 - 2 - 3 - 3.

If you strike 10 straight losers, then crawl away and hide somewhere! Seriously, you start all over again and hope to goodness that you will be able to retrieve losses.

Now we come to the final bracket, in which we look at how to bet the longer-priced selections. These are the horses priced at anything from 6/1 through to 10/1. My estimation is that you can strike 10 per cent winners and the longest losing run likely to be encountered would be between 20 and 25.

The staking progression I believe should be followed is:

1 - 1 - 1 - 1 - 1 - 1 - 1 - 1 - 1 - 1 - 2 - 2 - 2 - 2 - 2 - 2 - 3 - 3 - 3 - 3 - 4 - 4 - 4 - 5 - 5 - 6 - 6 - 7.

This allows for 28 (gulp!) successive losers. Were you to strike 27 losers your next bet would be 7 units, so you would need to retrieve losses of 68 units. Thus the winner required would have to be around 9/1 or 10/1.

These progressions, I believe, will work nicely most of the time but a lot will depend on the price of the winner which breaks any run of losses. If the winner is in the expected price bracket then the progression will be effective.

There is another tack you can adopt. This depends on lifting your stakes when you reach a certain level of profit. Let's say you are betting level stakes at $10 each. If you get, say, $60 in front, you can lift the stake on the next set of 10 or 20 bets to $15.

When your profit level has reached $120, you can raise the bet again, this time to $20. Once you have achieved $180 profit, the stake can be lifted again to $25. Once you hit $240 profit, lift again, this time to $30.

So, each $60 profit you make, you increase the stake by $5. You are always betting level stakes, except that as profits accrue you are raising the level stake to be bet.

I am assuming here that with a method like this you are going to be backing well-fancied selections and not silly flights of fancy! Level stakes betting is always a safe, predictable way to travel if you are a punter desirous of avoiding risk.

The good thing about it is that you always know how much your next bet is going to cost you. You do not have the worry over rising stakes.

Of course, with the methods I have outlined any rise in outlay is a gradual one and will not do you a great deal of damage. The only sustained damage can come from a series of losers, and you will sustain loss by backing losers no matter what staking plan you use. That's a fact of life, isn't it?

Staking, as I have indicated, draws many different responses from different punters. Each person has fixed ideas about it. I know punters who have absolutely no idea about money management; they just continue on their merry way, plonking down bets here, there and everywhere with no thought at all as to the 'method' being employed.

Yet these same chaps will often spend hours during the course of a week studying form! The last thing they seem interested in is HOW to bet, or HOW to judge the comparative probabilities of winning among the top contenders in a race.

Selection processes should mean careful study In his book Horse Racing Logic, the respected author Glendon Jones writes on this subject: "The real secret of beating the races is to be selective, by passing races (and missing winners) when the logical choice is overbet, and by restricting bets to selections that are underbet.

"A player who is able to achieve an average of 25 to 30 per cent winners cannot possibly reap consistent and substantial profits over long periods if betting each race, but will show handsome returns restricting bets to horses that are solidly underbet in relation to their true chance of winning."

So it is worth remembering when you are making your selections in readiness to bet them according to my outlined staking plans, that you steer clear of horses that are underpriced.

Whether you do this by instinct or by reasoned assessment is up to you.

But you have to be satisfied that the horse you are backing offers you some value. If it has a fair 2/1 chance in a race, you should refrain from backing it unless you are assured of getting this price or better.

It takes discipline, and you will miss a few underpriced winners, but in the long run it will pay off.

It's been pointed out before in this magazine that overbetting is a killer. Bet too many races and your long-term prospects drip away (in some cases, flood away!). Be patient.

By Mark Merrick