I was told many years ago to forget about single-horse betting. Why, a friend asked me, bet one horse in a race when you can bet three, even four, and make a profit if one of them wins?

A good question. I suppose it's a case of greed. If we can land the winner in one bet we get the lot. If we back three, we may win but we win less. At least, that's the theory.

This friend, I should point out, made a living (not a grand one, I admit) from Dutch Book betting. He followed the basic textbook Dutch Book technique.

(By the way, the term Dutch Book was supposedly invented when Dutch insurance companies that insured ships in the 19th century would organise and combine insurances in such a way as to make money whatever contingency occurred.)

And this is what Dutch Book punters attempt to do, except that it is impossible to cover all contingencies because you can only back so many horses and there is no guarantee any of them will win. So the punter has to exercise great skill and patience in framing his bets so that the number of winners he hits over 12 months are enough in number and PRICE to ensure an overall profit.

Nick Aubrey, of Twonix.com, recently wrote an article about Dutch Book betting so I will let him explain the niceties of  "basic Dutch Book betting". Nick (known to all and sundry as Anomaly Nick) says:

The goal is to win the same amount no matter which of a number of selections wins. The original concept is well suited to betting with bookies where your exact return is known in advance.

So let's look at an actual race, race 8 at Flemington on July 9. Let's say you wanted to back the following runners whose IAS (International Allsports) odds were as follows:

(1) UMBULA  8
(2) ASCANA  10
(7) CABELLA  51
(8) ACTION SHOT 8

The fixed IAS odds shown were those available on IAS (www.iasbet.com) just before the start of the race.

A basic Dutch Book bet on your selected runners consists of backing each one in proportion to their odds so that no matter which one wins you make the same amount of profit. The Dutch Book bet amounts are calculated using the following two step process:

STEP ONE:
For each selection divide 100 per cent by the IAS odds to get the Win Odds per cent. Then add up the percentages to get the Total Dutch Book percentage (known as the TDP).

(1) UMBULA ... Odds 8 ... Win Odds per cent 12.5
(2) ASCANA ... Odds 10 ... Win Odds per cent 10.0
(7) CABELLA ... Odds 51 ... Win Odds per cent 2.0
(8) ACTION SHOT ... Odds 8 ... Win Odds per cent 12.5

STEP TWO:
Decide what you want to outlay and let's say this is \$100. The Dutch Book bet is calculated for each selection by MULTIPLYING \$100 by its Win Odds per cent and then dividing by the TDP.

For example, for Umbula the bet amount equals \$100 x 12.5 per cent divided by 37 per cent which equals \$33.82 but rounded to the nearest \$ the bet is \$34.

The final tally of bets would be:

UMBULA  \$34
ASCANA  \$27
CABELLA  \$5
ACTION SHOT \$34

You can see from Nick's workout that the four horses to be bet had odds totalling 37 per cent. So what does the punter hope to achieve with this bet?

Obviously, he stands to make a decent profit NO MATTER WHICH ONE OF THEM WINS. In fact, the race was won by Ascana, so the winning bet was \$27 at actual odds of 9 / 1 (\$10. 00 as quoted). The return on this bet was \$270. This gave a profit on the race of \$170.

Now this all sounds grand in theory but, as we quickly realise, the actual odds on the total bet are only around 7/4. In reality, you have backed four runners at an overall price of 7 / 4 to win the race.

This 7 / 4 was achievable only because the four chosen horses were on offer at generous odds. What if you were to back four horses at shorter prices? Would it be possible to back them all and make a profit? A worthwhile profit?

Well, yes, sometimes it would, but often the percentages just wouldn't stack up.

Take for example a Dutch Book bet with four horses at 2 / 1, 7 / 2, 9 / 2 and 9 / 2. What sort of a position are you in with this quartet? Okay, let's look at their percentages:

2 / 1 equals 33 per cent, 7 / 2 equals 22 per cent, 9 / 2 equals, 18 per cent and 9 / 2 again equals 18 per cent. Add these up and the total is 91 per cent, leaving you very little "edge" to make a Dutch Book bet. In fact, your bets would be \$36 on the 2 / 1 chance, \$24 on the 7 / 2 chance and \$20 on each of the two 9 / 2 chances. This totals the \$100 you want to bet.

Your return? If the 2 / 1 chance wins, you'll get back \$108, a profit on the bet of only \$8. And the 7 / 2 chance? If it wins, your bet returns \$108. Same again! Need I say more. You have bet \$100 to make \$8.

Who needs that sort of betting? In cases like this, you could consider reducing the bet to three chances. Let's say you axed one of the 9 / 2 chances. Your bets would now have a total percentage of 73 per cent. Using the formula, the final bets would be: On the 2/1 chance you'd bet \$45, on the 7 / 2 chance you would bet \$30 and on the 9 / 2 chance you would bet \$25 for a total of \$100.

Now you have the situation that a winning return on the 2/1 chance winning would be \$135. Same for the others. You have bet \$100 to win \$35.

To a lot of punters that's quite acceptable. Many others, though, would demand a greater return on three or four backed runners.

If you stick to horses at 4 / 1 and longer you give yourself a good chance of securing good enough odds, as well as a fighting chance of coming out on top over 12 months.

Another example: You back four horses at 4/1, 6/1, 8/1 and 9/1.

The Win Odds per cent for each are: 20 per cent, 14.28 per cent, 11 per cent and 10 per cent. This totals 55.28 per cent. Okay, you want to bet \$100 again (this can be any amount you choose of course from, say, \$10 upwards).

The 4 / 1 horse's bet is worked out as \$100 x 20 per cent divided by 55.28 equalling \$36 to be bet. The 6 / 1 chance is worked in the same manner and the bet will be \$26, the 8 / 1 chance will carry a bet of \$20 while the 9 / 1 chance will have a bet of \$18. This is a total of \$100.

If any one of the foursome wins you will get a return of \$180, for a profit of \$80. The actual odds of the bet were 4 / 5.

You will understand, then, that the LOWER the total of your selections' odds percentages the more you stand o make if one of them wins. The HIGHER the percentages the less you'll make.

The cut-off point? Personally, I'd make it at around 80 per cent. If the totals of your percentages add up to 80 per cent or more, then forget the bet or drop one- of the selections.

I guess the main aim is to try to bet four chances. If you can do this and make a decent profit on the race, then all well and good. But please remember that this sort of betting needs CONSISTENCY. You don't want to be missing out on too many races.

If, say, you are averaging a Dutch Book. returning 4 / 5 on the foursome, then you must be able to have a strike rate with your bets to accommodate .this figure in order to make a profit.

If you have 100 bets in a year at \$100 each that’s an outlay of \$10,000.

You'll need to hit at least 55 per cent of the Dutch Book bets just to break even. To make 26 per cent profit you will need to notch 70 out of 100 bets. Can you do it? Think about this question very carefully.

** Nick Aubrey's website is at www.twonix.com and contains further information on Dutch Book betting. You can download a desktop calculator from www.practicalpunting.com.au. It's a beauty.

By Jon Hudson

PRACTICAL PUNTING - SEPTEMBER 2005