A professional punter I know reckons the trifecta remains the best way to get big profits on Australian racing, but he says the exacta runs a close second.

He warns, though, that "exacta science" is definitely not an exacta science!

With this form of betting, you need to select the 1st and 2nd horses in the correct order. With a quinella, you have them in any order.

Exactas obviously present punters with a tougher task than an ordinary quinella, but an easier task than trying to land a trifecta.

The mathematics of an exacta is similar to that for a trifecta. Example: You select Horse A at 3/1 and Horse B at 4/1 to run the exacta. The chance of Horse A winning is 3/1, and the chance of Horse B finishing 2nd is the same as the chance of that horse winning if Horse A is ignored.

The way to arrive at the correct mathematical odds is to express the odds as a percentage. Horse A at 3/1 has a 25 per cent chance of winning. If A is ignored, the percentage of the remaining chances is reduced to 75 per cent.

So, if A is ignored, B has a 20/75 x 100 chance of winning, or 26.67 per cent. The odds about B finishing 2nd are 100/26.67 equalling 3.75 minus 1 equals 2.75, or 11/4.

As chance of winning is 3/1. If A wins, B's chance of finishing 2nd is 11/4. So the combined chances of A running Ist and B running 2nd are 4 x 3.75 equalling 15, which is 14/1.

You can use this formula to work out any exacta odds. My approach, and one that gives a pretty good idea of things, is simply to multiply the two prices and add one.

Thus, a 3/1 chance into a 4/1 chance would be 3 x 4 plus 1 equalling 13, which, as you see, is pretty close to what the strict formula works out.

Let's say you have a 2/1 horse with a 5/1 to run 2nd. That's 2 x 5 plus one equals 11. Expect around $11 for your $1 bet.

This is more or less a "rule of thumb" way to get an idea of the odds of an exacta pairing. It gives you an idea of the true value, odds wise, of the exacta.

Knowing when to bet the exacta is something of an art in itself. At the track recently, a horse which had won five of his seven starts returned after a spell and was at odds-on. My own fancy was another horse at 10/1, but I was fearful of the favourite.

The quinella would have paid peanuts because of the favourite, but, but taking the favourite to win and my selection to run 2nd worked out well for me.

The exacta paid $9.00. In reality, I had backed the odds-on favourite at odds of 8/1. There was enough in the bet to have reversed the exacta as well, to have my selection to win and the favourite to get 2nd (an even bigger dividend was in the offing).

The idea of reversing like this has to be carefully thought out. Many quinellas will pay about two-thirds of the exacta and so a bet of $10 on the quinella might prove more profitable than two $5 bets on the exacta.

Say the exacta paid $19 and the quinella $11. The profit on the exacta is $85 (5 x 19 equals 95 minus the 10 bet) and the quinella profit is $100 (10 x 11 equals 110 minus the 10 bet).

It's simply a case of thinking things through, though it can be a most inexact science. The following is a plan for the exacta when you have a standout chance.

* The field must be 11 or fewer.

* Decide on your standout banker selection.

* Use the banker with your two choices to run 2nd (for, say, $10 each).

* Back the standout $10 win.

* Back the Field to win and your banker to run 2nd for $1 each exacta.

If the standout banker wins, that's good. You stand a good chance of landing the exacta, and you have the win bet up. If the standout runs 2nd, you have a chance of "saving" on the transaction via the Field bet, especially if a roughie gets up.

When betting favourites in exactas, or quinellas, there are things to remember. Short priced favourites are usually overbet for 2nd in quinellas and exactas. These favourites automatically make good standouts to win but are mostly best ignored for 2nd place.

American form analyst Kenneth Strong has some interesting views on exactas. He says: "Similar to win betting, the key to betting exactas and quinellas is finding value.

"Prime betting opportunities for exactas and quinellas occur when horses are overbet in the win pool. Also, similar to win betting, if an exacta or quinella payoff listed on the tote before the race is lower than you think it should be, it represents poor value and should probably be left alone.

"The worst value in exacta betting is usually the favourite on top (in the win spot), with the second and third choices on the bottom (in second place). These types of combinations often pay lower than they should. Another type of exacta bet that often pays less than it should is the combination of two longshots.

"While multiple-horse exacta boxes are thought by many handicappers to be poor wagers, due to money being wasted on combinations that have little chance of winning, there is a situation when these types of bets can be very lucrative.

"If you determine that a favourite has a good chance of running worse than second in a race in which the secondary contenders are also weak, you can make some good scores using multiple-horse exacta boxes.

"Other types of exacta combinations that represent good value include favourites on the bottom, and combinations of medium priced horses. The latter almost always produces payoffs higher than it should.

"If you have determined the favourite has a good chance of finishing worse than second, try playing some boxes of medium priced horses with smaller saver combinations including the favourite on the bottom of the exacta."

My professional punter friend says that while he still backs exactas "on a limited basis" ' he learned a long time ago that if you think you can pick the exacta with good value horses then you should go one step further and back the trifecta.

He does this using the AB-BA Field approach, something we have discussed before in PPM articles.

It makes sense. You pick Horse A and Horse B to run lst/2nd, and you are prepared to back your judgement with an exacta bet, so why not extend the bet and attempt to garner a big paying trifecta?

Even in big fields, this is not an overly costly bet. For example, the field has 16 runners. You choose A and B to run lst/2nd, so that leaves 14 horses to fill the Field section for 3rd. The cost of such a bet is determined as follows:

A-B//A-B//FIELD (14) making total bet of $28, or just $14 if you be in half units. And of course you can use Flexibet to reduce the bet further if you wish (you get a percentage of any payout depending on how much you bet).

Should a roughie arrive in 3rd slot, you could be in for a big payout.

Finally, a look at the approach of the US form analyst Cordon Pine. He says: "The notion behind my exacta wagering method is, if a horse is overlayed in the win pool, it's likely to be overlayed in the exacta pool.

"Combine that with the fact that one of the top two favourites is going to be in the exacta more than 80 per cent of the time, and you've got my method. If any of my overlays runs at least second, I'm probably going to make money in the race.

"If I have multiple overlays that run one-two, I'm going to have a bonanza. And I don't have to look at exacta payoffs for 12 different combinations with three minutes to race time - I just watch the win odds, mark my bets, and place them."

PRACTICAL PUNTING MONTHLY - MAY 2006

• Learn how to increase your profits from punting
• Discover all the information you need to win
• Find out how to beat the bookies and make more money!