Following on from last time we look further at the noble pass time that is punting on the horses.

Last time - which you can read here - we discussed the importance of looking at races more probabilistically with the goal being not to find winners, but to find good bets.

This time we will continue down that path with an article best summed up by the phrase -

It doesn't matter that you like the horse, all that matters is how much you like the horse.

Many have phrased that line of thought a lot more eloquently than I, including the legendary Pittsburgh Phil.

“Every bettor should have such a definite conclusion as to the probable result of the race, that he can form for himself the market value of all horses in a race, and make a schedule of the prices as they should be in his opinion.”

We need to look beyond just which horse we think is the most likely winner, with the goal being to figure out how likely each horses chances are.

Put simply, if we are going to find good bets we need to have our own set of prices.

Coming up with our own priceline requires us to first make an assessment of the runners and put them into some order, what many would call ‘doing the form’.

Returning to our golf analogy from last time we can look at this point as our swing.

Everyone does things a bit differently, some methods might be downright ugly, but in the end we just need it to be effective. We just need to be hitting enough greens.

There are plenty of different methods for quantifying horse performance. Methods such as handicapping (by weight, speed or pace), profiling horses using the market, or other statistical approaches, are all valid - a pinch from all of the above is perhaps the ideal – but they all require more time and investigation than can be afforded to them here.

Whichever way we go about it – as mentioned last week the validity of your methods will become pretty clear when you look at your bottom line at the end of the year - we are ultimately looking to put the field in order. To give each horse a ranking.

From this ranking order we want to try and divide the field into what we consider key chances and the rest.

The table at the bottom of this article is a version of similar distribution tables discussed in Don Scott’s Winning series of books which provides a good, simple way of converting what we produce via doing the form into chances.

As an example lets price up a mythical super sprint where, after doing the form, we have come to the far from controversial decision that Black Caviar is our most likely winner.

Using a mythical set of ratings we have come to the conclusion that Black Caviar is 2.5 lengths better than her old sparring partner Hay List who in turn we have rated 1.5 lengths better than speedster Sepoy. Current top dog Terravista sits 0.5 lengths away as our fourth pick.

Using this table we say that for every 100 Black Caviar wins we get 50 Hay List wins, 25 Sepoy wins and 20 Terravista wins.

That adds up to be 195 races and so;

Black Caviar = 100/195 = 51% or $1.96

Hay List = 50/195 = 25.5% or $3.9

Sepoy = 25/195 = 13% or $7.50

Terravista = 20/195 = 10% or $10

Rounding aside, this adds up to 100%, but given that we are only pricing up those we have considered to be our ‘key chances’ we want to build in some margin for error.

Testing our selection methods to find out how often the winner comes from our ‘key chances’ will give us a good idea as to how much margin we should be building in.

If we are finding the winner in our chances in 80% of races then 80% is a good starting point as the percentage to set your market to.

This can be simply done by multiplying our likelihoods by 0.80.

Eg. Black Caviar = 51% x 0.80 = 40.8% = $2.45

Once we have our own opinion displayed as probabilities in the form of odds we are ready to go on our search for good bets.

We’ve worked on our swing, we are hitting the ball well (hitting 80% of greens in reg!), now we need to figure out which greens to attack or when we should lay up.

Tactics are key, avoiding bunkers can be just as important as hitting greens, and that is something to be discussed in the next article.