This month we wrap up the series with a continuation of our beaten margin statistics and a look at the effects of weight. Before we do however a quick look back at the market percentage analysis we started last month.

CITY OR BUSH?
What we didn't do in part 5 was clearly answer the question posed by the article's title, namely City Bets - or Bush? We will do so now.

You will remember that the total market percentage is calculated by converting all the bookmaker's prices for a race into percentage probabilities and adding them all together. As discussed last month, the market percentage is usually over 100 per cent and the excess is referred to as the bookies theoretical advantage.

I use the word 'theoretical' deliberately because in practice the bookmakers don't automatically make more money in races where the market percentage is higher. The reason for this was touched upon last month and basically revolves around the fact that when there is less money around, the bookies cannot balance their books and can actually lose on certain horses in a race. In these circumstances the theoretical margin must be higher so as to compensate for these loses.

Last month we found that Sydney markets were framed to an average of 116% at jump time, Melbourne markets to 177% and Brisbane markets to 128%. Because of the reason just given, we cannot however conclude that bookies in Sydney make more money than bookies in Brisbane. They might do in absolute dollar terms because of higher turnover but in percentage terms their profits are unlikely to be very different.

What we did conclude last month was that horses starting at even money and 6/4 win more often in Sydney than in Melbourne or Brisbane. The reason for this is that horses at given starting prices win less and less frequently as the theoretical market percentage increases.

Now we can answer the question posed by the title to last month's article, namely City Bets or Bush?

The fact is that you are more likely to win by backing favoured runners in races where the market percentage is below 125%. Most races in Sydney, Melbourne and Adelaide fall into this category.

The percentages will be further against you in races where the market percentage is between 125% and 135%. Most races in Brisbane, Perth and at the major Victorian and N.S.W provincial tracks fall into this category.

Your chances are worst in races where the market percentage is greater than 135%. Races at Queensland, South Australian and West Australian provincial tracks generally fall into this category as do nearly all country and picnic races.

It is important to realise that these conclusions do not mean you shouldn't bet on races in Brisbane or in provincial or country areas. They just mean that the task for the punter simply following the markets will be more difficult. If you use ratings or a betting system to make your selections then you should not be afraid of betting at these venues provided there is sufficiently disclosed form.

BEATEN MARGINS
Last month we gave a few average beaten margin statistics to whet your appetite. We claimed that horses finishing second were on average 1.5 lengths behind the winner, horses finishing third were 2.9 lengths behind and horses finishing fourth were 4.7 lengths behind.

Details for all finishing positions up to 16 are as follows.

FINISH
POSITION
MARGIN
10
21.5
32.9
44.7
56.2
67.4
78.8
810.3
911.6
1013.4
1114.8
1216.9
1317.8
1420.1
1521.3
1624.5

The relationship of beaten margin to finishing position is almost linear as you can see by noting that the average beaten margin is almost the same as the finishing position.

For those of you looking for a formula to calculate average beaten margins, multiply the finishing position by 1.5 and then subtract 1.5 from the result. For example a finishing position of 6 becomes 9-1.5 or 7.5.

So what point is all this? Well, beaten margins are useful whether you are calculating ratings or just following a system that looks at last start margins. Fortunately most of us can find exact beaten margins for any horse's past form from detailed formguides like the Sportsman. There are two situations however when beaten margins are not available and the preceding information will be useful.

First is when you are examining a country or provincial race meeting for which little other than finishing positions are available. The second, as we mentioned a couple of months back, is if you live in New Zealand and your formguides don't list beaten margins for horses finishing further back than fourth.

While finishing positions can be used to estimate beaten margins, it must be stressed that they are approximations only. Sure they are better than nothing, but they are no where near as good as the real thing.

You should all be aware that one horse can finish second and be five lengths behind the winner while another can finish sixth yet only be three quarters of a length behind. Clearly actual margins are more important.

WEIGHTS
There is one last form factor I would like to analyse before concluding this series and that is the effect of weight carried by a horse.

We all know that weight can slow down a horse but wouldn't it be nice if we had some comprehensive statistics from which we could determine how great the effect really is.

Most rating methods assume that 1.5 kilograms slows a horse down by a length but I am sure a horse would be slowed down more than this over longer distance races and over wet surfaces. In fact the relationship is probably different for each individual horse with some horses being unable to perform over a certain weight no matter what the conditions.

Unfortunately time prevents me from carrying out a detailed analysis of the effects of weight so all I have done is ask my computer to look at the strike rate of horses at different weights, for example 50kgs, 51kgs, 52kgs etc. Even taking this simple approach, the results are quite revealing.

Just pause for a moment and consider the following:

WEIGHTWIN %
50-50.55.3
51-51.56.0
52-52.57.7
53-53.59.5
54-54.510.8
55-55.513.5
56-56.516.5
57-57.514.0
58-58.517.2
59-59.515.4

See how quickly the strike rate improves from just 5% to more than 15%.Clearly the lower weighted horses are to be avoided. The horses to concentrate on are those carrying 55 kilograms or more.

Certainly the more heavily weighted horses are considered superior by the handicapper or the weight for age/set weight scale by which their weights have been allocated. And accordingly you might expect them to perform better.

I for one however didn't expect their strike rate to increase with weight as quickly as suggested by the figures just tabulated. After all, weights are meant to even out the chances of horses in a race.

Although the trend is more extreme than I expected, its direction nevertheless makes sense when you consider that the official handicapper is forced to allocate weights between a maximum weight (the top weight) and a minimum weight (the limit weight).

In handicap races the difference between the top weight and limit weight is usually around 5 or 6 kilograms. Since the difference in ability between the best horse in the race and the worst is usually much more than this, the handicapper is unable to accurately handicap all horses in a race.

You will generally find that horses above the limit weight are correctly handicapped (at least as far as the handicapper is concerned). Those on the limit however may or may not be correctly handicapped-some will be but many should really be on a much lower weight which the handicapper is unable to allocate.

This being the case, you can see why the strike rate of horses in the range 5051.5kgs is extremely poor.

Since it is not uncommon to find handicap races where nearly half of the field are on the limit weight, excluding limit weight horses from your betting appears to be a very good way to reduce punting risk. Incidentally, you will see that John O'Sullivan has successfully used this very approach in this issue's Plan Of The Month on page 34.