Every time you study the formguide you are adopting what may be called a mechanical process in coming up with your selections.

I think it's true to say that most punters follow the same "pathway" in analysing the form. We all have an individual way of looking at the form, and each, in its own way, becomes a systematic approach.

So when we talk about systems, maybe we should really be incorporating form analysis in to the term? I think so, though many will disagree because they'll claim that subjective judgement is required to come up with the final selection when you analyse form, whereas no such judgement is exercised when it comes to mechanical systems.

That's fair enough, but I stick to my point that the PROCESS is more or less mechanical and thus much in line with the rules that are followed by system fans.

So what if we want to create a system of our own? How easy is it? How do we go about it? Is there much involved? Can YOU do it?

I'll try to answer these questions as we go along. But, first of all, yes, anyone can create a system and you can be as casual or as serious about it as you like. Often the very simplest of systems will prove more profitable than grander ones. It all depends on the ingredients.

Dennis Walker, the guru of The Rating Bureau, came up some time back with a summary of pointers for system creation. They are an excellent guide and I'll take you through most of them and make some observations. (By the way, you can contact Dennis via www.trb.com.au).

FOR CONSISTENCY, RELIABILITY
Win per cent, Place per cent, Consistency, Wins last 10 starts, Places last 10 starts, Age of horse, Number of career starts.

Now here we have an ideal starting point for finding a system. If we want consistent horses to be considered, then we have to look at what makes a consistent horse ... yes, the win and place strikes, the recent form, the age and number of career starts.

Form can be a trap. For example, a horse aged, say, 7 years may have a 35 per cent win strike but full analysis of his form might show that in the last year or so his win strike has been NIL! In other words, all his winning form was back in his earlier days and now he's a relic of the past.

Recent form is always the best form. Thus, the idea that perhaps we should base any consistency ruling around a horse's last 10 starts and not its last 40, or whatever. It makes sense to me.

DETERMINING CLASS (QUALITY)
Prizemoney earnings. Yes, a horse's earnings can be a good guide to its innate class or quality. We can work out an average by dividing the earnings by number of starts, though I favour isolating a horse's winning runs only and then dividing them by the earnings of those wins.

Thus, a horse with total earnings of $100,000 from 15 starts could actually be examined by referring only to its wins. Let's say there were four wins worth a total of $65,000. The horse's average WIN earnings, then, are $16,250.

Maybe if you were putting together a system you could use such an average instead of the usual "all races" divided into all prizemoney? I think the wins-only average gives a much clearer picture of the value of each win.

DETERMINING FITNESS
Number of runs from a spell, last four starts, Distance of the race.

This is one of the toughest factors in horse-racing to work out satisfactorily. We can only really GUESS about a horse's fitness, and use our knowledge of what's happened before to make a reasoned assumption.

The number of runs from a spell to determine fitness may vary from horse to horse. Some horses can reach peak in two runs, others take four, five or more races. So it's a difficult area. When you're building a system you might have to take a broad approach and hope that you can capture the right mark with more horses than not.

You might care to take into account the length of a spell. For example, you could say that a horse that's been out for 60 to 90 days will only need two runs to reach peak, a horse out for 91 to 120 days might take three runs, and a horse out for 121 days or more will take at least four runs.

You'll be wrong some times, but overall you might be right enough times to keep ahead of things.

The distance of the race is important. Sprinters come to hand more quickly than stayers. So you might decide that if it's a sprint race, a horse can be declared fit even if it is returning from a break, or has had only one start since a break.

THE LUCK FACTOR
Field size, Barrier draw, Race distance, Jockey ability.

Lady Luck plays a vital role in most races. We've all cursed our luck at some stage in our betting lives. And the factors mentioned under this heading are important.

Field size is important. In a big field, much can go wrong. Especially if the barrier position is not a good one, and a horse is going to be caught back and wide, or trapped in the middle behind a wall of rivals.

Your system must try to take into account these likely unfavourable happenings. You can use eliminations to avoid being on horses drawn wide or awkwardly in big fields, especially in races where the distance itself presents a problem.

In a short sprint, around a bend, a wide draw combined with an ordinary jockey may often spell disaster. So think carefully about this aspect of things.

When you draw up your rules you will want to reduce or increase the luck factor. Maybe you will restrict your betting to smaller fields? Say, up to 12 runners only. Maybe you'll restrict your betting to middle-distance races, or sprints?

Maybe you will very seriously consider the jockey factor, and only bet when you know for sure that the jockey is triple A rated! I've lost count of the number of times a moderate jockey has left my cash in the bookie's bag,

Only recently, at a Mornington meeting, a well-known young jockey rode like a sack of potatoes on a horse I had backed ... missed the start, trailed last, went wide too early and then battled home wide in the straight. Ugh!

RACE SUITABILITY
Won or placed at the track, Won or placed at the distance, Won or placed first-Lip previously, Won or placed on slow or heavy going.

When you put your money on a horse, you need to satisfy yourself that it is suitably placed. You know, in the right race, the right distance, the right track, the right going ... all these points are so important.

Any system you compile will need to take such factors into account. If your rules lead you to Horses A, B and C as the contenders, the system then has to demand to know if the race and the conditions are suitable.

Can Horse A win, say, a 1600m Open Handicap at Caulfield on a track rated Heavy? Has it done it before? Can it do it again? Does its record show that it has a good chance of doing it? The same thinking must be applied to B and C.

TRAINER STRATEGIES
Jockey changes, Jockey previously successful on horse, Weight change, Class value changes, Distance change, Prizemoney change.

Maybe you can frame an entire system around trainer strategies? A system that zeroes in on selections by looking at jockey switches and weight changes may well be successful.

Always remember when you build a system that you are looking for an angle that no-one else has thought about. Something different, something hopefully unique.

What about jockey switches from an apprentice to a top jockey? It often happens that such switches lead to winners. Horses that, say, ran well last start, without winning, when ridden by a kid, and which now are to be ridden by a senior jockey.

Used in conjunction with other factors (race suitability for example) this approach might throw up something very special.

RUNS WHICH ADD VALUE TO BETS
Poor barriers, Lesser jockeys, Last start beaten more than 6 lengths, Unplaced last two runs, Big weight, Distance change, field size.

Ah, now we come to the aspects of system creation that can make or break us. Do we ignore a bad barrier and allow a selection in the hope that it will win anyway and provide us with a bigger-priced collect?

The factors mentioned usually mean a horse is sent out at a good price. Punters love to see 1, 2, 3 next to a horse's name and when they see 4, 5, 6, 7, 8, 9 and duck eggs then they tend to be wary.

In some instances, however, you shouldn't be put off by unplaced runs. They may well have been excellent performances.

So, think about any rules you make that may cut out the prospects of a big-divvie collect, because any system ever devised relies on such windfall returns every now and then to ensure an annual profit.

It's a matter of weighing up the pros and cons. It's tough and as the system creator you have to do it.

RUNS WHICH DETRACT FROM VALUE Barrier, Jockey, Recent wins/ form, Win per cent, TAB number (1 to 6).

Here's another problem. Often, GOOD factors can kill the dividend and rob your selection of any value! (Yes, this is a tough business!) A well-fancied horse in a good barrier with a good jockey and with good recent form ... well, can you see yourself securing any value?

It's a fact, too, that any runner with a number from 1 to 6 will always be more heavily backed than runners further down the list, be they in form or not.

But, of course, you can't turn your nose up too much at such horses because they do win their share of races. And you don't want to be throwing out too many winners. That's not the aim of the game at all.

GENERAL RELIABILITY RULES WORTH CONSIDERING
Limit age to between 2 and 5, Career starts 1 to 20, Avoid maiden races, Avoid Class 1 and 2 races, Stick to distance range 1200m to 2000m, Avoid horses rising 5kg or more in weight. Look for lightly-weighted horses on wet tracks.

These are some excellent examples of aspects of a race that you can factor into a system. They are very sensible. Yes, limit the age range to between 2 and 5, who needs the greybeards?

Lightly-raced horses, too. But no horses who haven't won a race. And by sticking to a certain distance range you will be helping your system to concentrate its attention on a select group of horses.

So, there you are, some pointers to guide you along the way to creating for yourself a system that may well change your punting fortunes for the better. You can't use all the factors; you must pick and choose and then frame your system in the best way you know how.

What you need is a "theme", as Dennis Walker so aptly puts it. It could be, for example, a horse dropping in distance or class, it could be looking for horses that are trained by certain trainers or which are to be ridden by certain jockeys.

Sort out your "theme" and go for it. You could well end up with a ripper of a system.

By Denton Jardine

PRACTICAL PUNTING - MARCH 2003