Most punters have, at one time or another, developed a system they hoped would be the mother of all systems, however, what looked so promising generally became so bad it was eventually thrown into the circular filing system commonly known as the rubbish bin.

The hard fact is that poorly designed systems generally suffer from the "what used to work doesn't work now" syndrome, because their premise was based purely on past results. But doing the form is traditionally based on past results, so where does it all go wrong?

I believe most system punters get it wrong by being too complicated. They have too many system rules to the point where they are virtually attempting to split the atom, whereas a better approach would be to introduce a small number of filters (rules that weed out non-productive selections) and THEN do the form in a traditional way.

In other words, use some basic rules which, as a bare minimum, have the punter sitting in the right ball park where the action at least starts with some promise.

The definition of "some promise" needs to be clarified in that I believe any plan/system that loses about 5 per cent on turnover has really proven to be 50 per cent better than betting on the best yardstick, "the favourite", which loses 10 per cent on turnover on average.

From this stage it is really up to the punter to apply serious form considerations to the remaining contenders in an attempt to crack the "break square" barrier or, at worst, be very, very close to it.

Such form considerations as runs from a spell, age, sex, distance, track conditions, barriers, jockey, trainer, track layout like or dislike and running style are classical issues that need to be attacked towards the "some promise" contenders. With the application of commonsense formwork and a sensible staking plan the small loss on turnover could be turned into a small profit and have us cooking with gas.

Fellow Ausracer (a generic name for members of the Internet discussion group Ausrace) Leigh Rolls and myself have delved into the excellent CD-ROM database, produced by Dennis Walker from The Rating Bureau, called Winline GTSi, with the aim of attempting to set some general commonsense rules that would eliminate poor betting possibilities.

Our modus operandi was to interrogate Leigh's database only on well-fancied horses and it was decreed we would only research those that had started SP 1 /1 to 5/1 on city tracks in Melbourne, Sydney, Adelaide and Brisbane covering all days of the week.

Additionally, we decided to break down all the 1/1 to 5/1 chances into five separate SP price brackets: 1/1 to 15/8,2/1 to 5/2,11/4 to 3/1, 13/4 to 15/4 and 4/1 to 5/1 and each of those brackets were subdivided into the track conditions of Fast, Good, Dead, Slow and Heavy.

In order to ascertain which filters should be seriously homed in on, it would be best to begin with the overall figures covering all horses starting between 1/1 and 5/1 and then gradually provide the various situations uncovered for you to determine which "rules" you would apply.

In order to better understand the statistics, we shall call each set of figures a "Chart" where POT means Profit On Turnover and LOT means Loss On Turnover.

Chart 1 - All runners

Price RangeRunnersRetns LossLOT%
1/1 to 15/8313628912457.8
2/1 to 5/2 3892339549712.8
11/4 to 3/12834 2342 492 17.4
13/4 to 15/4 4355 4119236 5.4
4/1 to 5/1 9458 8377 1080 11.4
Overall 23675 21123 2552 10.8

The overall variance of the price ranges does not allow us to eliminate any price ranges even though the 1/1 to 15/8 and 13/4 to 15/4 price ranges certainly whet the appetite.

Chart 2 - State of the Track

Track Conds.Runners Reins.LossLOT%
Fast - Dead 18963 17052191110.07
Slow - Heavy4712406564713.73

Although we are not talking of a massive percentage, the difference of 3.67 per cent between the two combinations of track conditions has drawn a clear line in the sand and the dropping of rain-affected tracks seems necessary.

Some punters will swear blindly that races beyond sprint distances are littered with unevenly paced events and are therefore suspect betting races, so let us put this to the test by splitting the distances into seven ranges.

Chart 3 - Distance

1200 only 3415r3008 -11.91%
1250-13502131r1903 -11.98%
1400 only1700r1567 -7.82%
1500-15501063r1468 -8.93%
1600 only1646r1468 -10.81%

Overall the figures  show 10300r9282 = -1018 units or -9.88 per cent and any variances at the different ranges have not been significant enough, thus the figures in Chart 3 do NOT support distances being a major impairment in the punting process.

The next issue to investigate is one of fitness and for our purposes we must use the numbers of days since the last run as our guide. We have split the dates into 10 sections: 0 (first start in a race), 1-6 days, 7 days exactly, 8-13 days, 14 days exactly, 15-20 days, 21 exactly, 22-27 days, 28 days exactly and 29-999 days. Rather than display chart after chart in full detail, we shall display a shortened version of each chart and some comments thereafter. Chart 4 - 0 days 564r601 equals +37 units = +6,56% POT.

I was amazed at this figure as it seems first-starters at SP 5/1 or less are excellent betting propositions. Could THIS be a system on its own?

Chart 5 - 1-6 days
896r993 = +97 units = +10.82% POT

Without doing any formwork and just betting all runners at 1/1 to 5/1, we found a 20 per cent turnaround from the favourites' normal 10 per cent LOT!! Wow! Without doubt, add this to Chart 4 for deep consideration.

Chart 6 - 7 days exactly
1219r1092 = -127 units = -10.41% LOT

What happened?

Chart 7 - 8-13 days
3913r3723 = -190 units = -4.85% LOT

This is not too bad and fits the "some promise" premise I set earlier. With some serious formwork this might be turned around.

Chart 8 - 14 days exactly
3735r3063 = -672 units = -17.99% LOT

This is terrible!

Chart 9 - 15-20 days
3573r3188 = -385 units = -10.77% LOT

Chart 10 - 21 days exactly
1430r1209 = -221 units = -15.45% LOT

Chart 11 - 22-27 days
1161r1026 = -135 units = -11.62% LOT

Chart 12 - 28 days exactly
355r327 = -28 units = -7.88% LOT

Chart 13 - 29-999 days
2681r2434 = -247 units = -9.21% LOT

Phew! There is an enormous amount of data to digest in the charts relevant to Fitness but it is quite clear that 0 days (first-starters - I am still amazed) and 1 to 6 days are very lucrative betting areas. The 8 to 13 days spell dates need working on while from 14 days and onwards you are back to square one.

I personally still cannot back first starters but I can assure you I will be homing in on 1 to 6 days with a vengeance in coming times. There is just one more area to delve into the issue of current form, that is, last-start positions.

Chart 14 - Last-Start Winners
5284r4744 = -540 units = -10,21% LOT

Chart 15 - Last-Start Seconds
3843r3401 = -442 units = -11.50% LOT

Chart 16 - Last-Start Thirds
2751r2556 = -195 units = -7.08% LOT

Chart 17 - Last-Start Fourths
2002r1726 = -276 units = -13.78% LOT

Obviously, Last-Start Thirds stand out and perhaps this is the start of a system but overall I believe Current Form figures for 1st to 4th (13880r12427 = -1453 units = -10.46% LOT) have not shown any major trends for us to deem last-start form as a major worry.

For the intentions of this article, which was to see if we could isolate some "commonsense rules to eliminate poor betting possibilities", I believe we have clearly shown that the recency of the last run to be the MOST important factor we have delved into.

It's interesting to ponder why 7 days exactly does so poorly, yet a day or two less and the figures turn around dramatically. Are we dealing with runners who are racing from the same grade to the same grade and is it harder to win under those circumstances? Is this hypothesis supported? A quick look at Wednesdays and Saturdays only shows these figures.

Chart 18 - Saturday to Saturday: 7 days only
934r862 = -72 units = -7.70% LOT

Chart 19 - Wednesday to Wednesday: 7 days only
219r189 = -30 units = -15.85% LOT

So, the hypothesis is supported for Wednesday to Wednesday, possibly because we are dealing with lower class racehorses of varying ability however, on Saturday to Saturday there is a quite different scenario, but why? I guess we can consider several arguable reasons but the figures are the figures and that is that!

In the course of this article we have unearthed two major factors in the 1/1 to 5/1 price bracket: First-Starters and 1-6 days that show an actual profit on turnover WITHOUT any formwork whatsoever.

Obviously, with the first-starters, the majority will be two-year-olds who have trialled well and/or have come from high-profile stables renowned for producing two-year old winners. The 1-6 days runners will seldom be two-year-olds but more often be horses who have been targeted for a quick backup. Because of nomination and acceptance deadlines, when a horse is nominated in the first place, serious consideration must have been given to BOTH races.

There are some trainers who rarely give their horses a quick backup but, when they do, you can rest assured that their horses will perform very well. However, there are some trainers who are poor practitioners of this art. As a punter, you will need to keep records to know which trainer is in which group - you will find the effort will pay rewards.

I hope this article has stirred enough interest for you to get on with creating your own systems. The "rules" may be as simple as "back 1-6 backup horses when 1/1 to 5/1". However, the more adventurous punters may wish to delve into the approximately 5 per cent or less LOT runners and add personal "rules" they have developed through their punting career.

By Roman Koz