More than six months have passed since PPM printed an interview with the legendary US handicapper Mark Cramer.

His earlier books are a little hard to find, but his most recent work, Value Handicapping, subtitled "The art of making your own line and identifying overlays", is still readily available.

This is Part 2 of an extract from the book, which is a chapter introducing the thinking behind Cramer's 'value betting" philosophy. The book has received acclaim from many in the industry, including PPM's own correspondent, Barry Meadow.

I prefer the opposite approach. I tend to think in terms of odds, so I begin with the expected odds of each horse, then translate those odds to percentages. If the percentages for each horse in a field add up to 100, my line is complete. If they add up to a sum that is either greater or less than 100, then I must make adjustments.

For example, assume you like Bold Wilma in the third. She should be a 2/1 most-likely winner. You write down 2/1, then look at the table and see 33 per cent. But there are three other legitimate contenders in the race, and respectively you estimate them at 3/1 (25 per cent), 7/2 (22 per cent) and 4/1 (20 per cent). Three other non-contenders still have some chance to win. They are estimated by you, the linemaker, to be 8/1 (11 per cent), 10/1 (9 per cent) and 20/1 (4 per cent).

You then total the percentages of all the horses: 33 + 25 + 22 + 20 + 11 + 9 + 4 = 124 per cent. Whoops! You must have overestimated the chances of several horses, since a line based on reality should add up to no more than 100 per cent. It's adjustment time. Which horse(s) have been given too much of a percentage chance?

Your adjustment is crucial, and is tantamount to the "second draft" of your handicapping, when refined thinking can lead to better analysis. The odds line is forcing you to think it through more precisely.

Your line is adjusted from top to bottom, and you begin by making Bold Wilma 3/1 (25 per cent). Yes, Bold Wilma was your answer when the guy in the plaid sport jacket asked, "who do ya like?" but the other legitimate contenders must be reckoned with. They must get percentage points that reflect their reasonable estimated chances.

Besides downgrading Bold Wilma's chances, you are obligated to demote the odds of your other contenders, or it would be impossible to come up with a 100 per cent total. (It is mathematically impossible for any race to have either more or less than a 100 per cent chance of producing a winner.)

Your adjusted "second-draft" order of odds, beginning with Bold Wilma, is as follows: 3/1 (25 per cent) + 4/1 (20 per cent) + 9/2 (18 per cent) + 511 (16 per cent) + 10/1 (9 per cent) + 12/1 (8 per cent) + 20/1 (4 per cent) ~ 100 per cent.

According to your first draft, you could have bet on your estimated 2/1 Bold Wilma at a 50 per cent overlay (to odds), which would be 3/1. But after your second draft which envisions a more contentious race, you've got Bold Wilma at 3/1 on your line. Now, you can only take her at 9/2 or above.

At the track that afternoon, Bold Wilma goes to post at 3/1. She would have been an overlay on your first draft! But with your final draft of 3/1 for Bold Wilma, there is no longer a betting advantage and she cannot be backed. On the other hand, your second choice (4/1 on your line) is going off at 8/1 (a 100 per cent overlay to odds). Even though you "liked" Bold Wilma, your line instructs you to place a wager on your second choice because she is a much better investment.

This has been an abstract illustration of the line-making process, a textbook introduction. Do not be discouraged if it all seems too complicated. Some people like to begin with abstract concepts. But if you are like me, you learn better by going from the concrete to the abstract and not vice-versa.

The more lines you make, the fewer adjustments you'll need, but in the learning stages, you'll be making adjustments with just about every race. Beginning line-makers tend to overestimate the chances of their contenders, thus ending up with greater than 100 percentage points when all the horses' probabilities are tallied.

Why the total of 100 percentage points for a final odds line?

In any given race, there is exactly a 100 per cent chance that there will be a winner: no more, no less.

But why not use the pre-post market as an odds line and simply bet the overlays?

First of all, the pre-post market is estimating how the public will bet. It will simply incorporate the factors most often used by the public. We have already established the certainty that we must handicap in a way that differs from the public in order to make a profit.

A more important reason for not using the pre-post is the fact that it is not based on a 100 per cent certainty that the race will be won. If you convert the pre-post market to percentage probabilities, you will discover that there is a 120 (and even more) chance that there will be a winner.

The extra 20 per cent is roughly equivalent to the takeout plus breakage. By adding percentage probability to each horse, the prepost maker artificially reduces the odds. For example, what should be a 3/1 (25 per cent) on a personal line may become a 2/1 (33 per cent) in the pre-post market, depending on the number of horses in the race. The pre-post maker may give Bold Wilma 2/1 when he knows she should be 3/1, thus deluding the crowd into believing that her 3/1 toteboard odds represent an overlay.

By constructing our personal odds line on the basis of 100 per cent, we are assured that the overlay we see is the overlay we get.

But let's not get lost in money management abstractions, just how can the handicapper construct a personal odds line? In the body of this book, we will explore various line-making procedures using real races on which yours truly made wagers, including winning, losing, and pass scenarios.

Like all handicappers, except place bettors on odds-on favourites at the harness races, I lose more races than I win. Space does not allow me to reproduce my exact win-lose proportion within the sample races we are about to describe; let Norman Mailer indulge in writing 1000-page books. I have chosen a sufficient sample of losing races for the reader to witness how and why a personal odds line is not invincible in the short run, and how it is geared for what really counts: long-run profit.

One more word about the real race illustrations in this book. My handicapping methodology is eclectic enough so that you will be walked through a variety of linemaking procedures.

Some of these approaches may resemble your own handicapping affinities. But ultimately, our job will be to adapt the form of personal odds line to the content of your own unique style of race analysis.

Each time you study a race that is illustrated with my personal odds line, you should read the past performances and experiment with your own line.

This extract from Value Handicapping was compiled by PPM's Tasmanian-based contributor Steve Wood.

The book is available from the publisher, City Miner Books at the web site

By Mark Cramer