Many of us go through life thinking we understand something, or assuming something to be right, when all the time we are wrong. I spent 30 years of my life believing that the word 'nonchalant' was pronounced 'nonkalant'.

The notion of the 'misunderstood truth' is more than apparent in horse-racing, especially among punters. Too many of us follow ideas that are way out of line with the truth, or with statistics drawn up over decades.

I know punters who fervently believe they can back every favourite and end up with a profit at level stakes. Alas, it isn't so.

The American author and handicapping expert Dick Mitchell, who's written some of the best books published about betting and selection, has made a special study of the fallacy syndrome and identified five areas where punters invariably have it all the wrong way.

He lists them as follows:

  1. Pass contentious races if a race doesn't boil down to one or two horses.
  2. Never bet for a place. Place betting is for suckers.
  3. Always insure a win bet with a place bet.
  4. Don't play exactas; they're too risky.
  5. If you're behind, increase your bet size; a winner is due.

Says Mitchell: "I've heard or read all of the above admonitions more than I'd like to admit. Each time I cringe." Mitchell is adamant that anyone who adheres to such approaches is heading for trouble.

This is how he analyses the 'five deadly blunders':

This is very poor advice. Consider, for example, a race in which there are 10 horses. A handicapper scans the past performances, and decides that six of these horses have very little chance to win.

However, there are four horses that are virtually indistinguishable. The handicapper can find no good reason to eliminate any of these horses from contention. In fact, they're almost all equal in his judgement.

So he throws up his hands in frustration and wails, "The winner can come from any one of four horses!" So, he passes.

This is a big mistake. Instead of passing, the handicapper should have rejoiced. Here comes a wonderful opportunity to make a score. There will be a discrepancy in odds when he gets to the track.

Suppose the public odds of his four picks were as follows:

A 3/2
B 5/2
C 6/1
D 8/1

Since each of these horses was considered equal, the handicapper should consult his records for his contender-selection percentage for this particular type of race; that is, how often is the winner found among his top four picks?

For the sake of this example, suppose this number is 72 per cent. This assumes each of his selections has an 18 per cent chance to win. This means fair odds of 9/2 for each of his contenders.

Let's look at his mathematical expectation or edge for each possible bet. Edge is calculated by multiplying the probability of success times the odds-to-one, then subtracting the probability of failure. (Please note: 3/2 is the same as 1.5/1, and 5/2 is the same as 2.5/1.)

E(A) = .18 x 1.5 -.82 - -.55
E(B) = .18 x 2.5 -.82 = -.37
E(C) = .18 x 6.0 - .82 = +.26
E(D) = .18 x 8.0 -.82 = +.62

Horses A and B are to be avoided like the plague. They offer negative expectations, worse than any casino game. Therefore, the handicapper should consider win bets only on C and D. The win bet on D is the very best bet a handicapper can make in the win pool. If he's conservative, he can bet both C and D to win and still have a positive edge.

The exacta pool looks even better. There are 12 possible exacta combinations using four horses. All 12 are equally probable according to our example. The probability of a 9/2 horse coming in with another 9/2 horse is approximately 4 per cent.

This means 25/1 is a fair exacta price. We don't want fair prices. We want prices that are grossly in our favour. If any of these exactas are paying $75 or more, we have a wonderful opportunity.

Play them all.

The argument against place and show betting is the fact that you have to hit ghastly high percentages to remain profitable. For example, a place bet that averages $2.80 must be hit 71 per cent of the time just to break even.

The public gets about 52 per cent with its top pick in the place pool. Normally, an odds-on favourite that pays $1.50 to win will pay $1.20 to place. In order for it to be fairly priced, it should pay $1.35 to place. This seldom happens. Therefore, it's a sucker bet to consider a place wager.

This argument sounds valid. The logic is almost unassailable. However, Bill Ziemba and Don Hausch disagree. In their book, Beat The Racetrack, they make a very strong case for place, and show (in the US, 2nd place only), betting.

They claim that between two and four times a day at every racetrack horses pay disproportionate odds in the place and show pools. They further argue that a player can make a living at the track by just watching the tote board and betting when certain conditions appear.

Mark Cramer and I used to have a common betting pool to exploit these situations. We made a ton of money together. Before we disbanded our mutual fund we had a string of over 30 consecutive winning show bets.

Most horse-race betters don't like the idea of investing $2 to get back $2.60. They fail to realise that a 20 per cent edge is a 20 per cent edge, regardless of what the investment is.

You can't insure a bet. No insurance company offers this product. You're not insuring a bet; you're just making another bet!

Betting a horse to win and place or betting across the board is usually a dumb bet. That's because each bet has its own edge. Usually, these edges differ by a lot. It's best to calculate the edge on each bet, and wager all your money on the one that offers the biggest positive edge.

Betting a longshot to win and place is the height of stupidity, unless you're willing to accept the place price as an overlayed win price.

For example, you make a horse 3/1 on your line. The public lets it go off at 15/1 and it's paying over $4 (for $1) to place.

In this case, betting both win and place is a very smart bet. As a rule, place betting should be reserved for horses underlayed in the win pool, but overlayed in the backholes.


This advice is utter lunacy. You should never skip a wagering opportunity that offers so much value. Thank God the public doesn't listen to this bad advice.

The truth remains that most players abuse this wonderful form of wagering. They do dumb things like buy three- and four-horse exacta boxes. (A box is a bet on all the possible exacta combinations.)

A three-horse exacta box is actually six bets, with at least five guaranteed to lose. A four-horse exacta box is 12 bets, with a minimum of 11 destined to lose. Giving away this many losing bets is what practically guarantees the bettor will lose.

If you learn to bet exactas properly, you'll be taking advantage of a lot of dumb money The exacta pool offers much more value than the straight pools.

This advice is absolute madness. The reason you're behind is that you don't know what the hell is going on at your track that day. It's the Good Lord's way of telling you that you have to go back to your handicapping records.

The notion of a winner being due is responsible for a lot of Las Vegas players coming to town in a $20,000 vehicle  and leaving in a $200,000 vehicle. They came in a Ford, and left in a Greyhound bus.

If you never believe another word I say, believe this: In any sequence of independent events, there is no such thing as a specific outcome becoming 'due'. That is, if heads just appeared 20 times in a row on the flip of a fair coin, A TAIL IS NOT DUE. A head is just as probable as a tail on the next flip The probability of a tail doesn't increase!

This same concept is true for horse races. Each race is an independent event. There is no memory of prior events. Your win percentage for this specific type of race doesn't increase after a string of losses. This is counterintuitive, but true nonetheless.

When you increase your bet size after a string of losses, you're courting disaster. This is known as 'chasing'. Ask any pit boss in Las Vegas or Atlantic City "What's the quickest road to the poorhouse?" You'll get a one-word answer. Care to guess the word?

If you're behind at the eighth race don't try to get it all back by increasing your bet size in the remaining races. If anything decrease your bet size. The only time you should increase your bet size is when you're winning, and then the increase should be very gradual.

By P.B. King