One of racing's perpetual debates is on the worth of staking plans - do or don't they enhance a punter's selections by improving profits over those at level stakes.

Practical Punting Monthly over the years has produced many, many articles about staking plans, as well as having produced at least two books on the subject. I notice that on the inside back page of each issue is an advertisement for "Back Issue Specials", which are a series of PPM magazines devoted to certain topics, one being the Great Staking Debate, a five-issue pack about level stakes versus progression betting.

I would encourage anyone interested in staking to grab this package while it is still available.


Back in 1998, fellow PPM contributor Roman Koz and myself had a long discussion about staking plans, in particular "target" staking plans. Roman has always been a believer in target betting, having meddled around successfully, and, on occasion, unsuccessfully, with various plans.

For myself, been there, done that and vowed never to do so again after a heart-stopping flirtation with target staking about a decade or so ago. At the time, the real issue wasn't so much whether the staking plan would work or not; it was more a question of not understanding exactly what the "comfort zone" factor meant in real terms. A run of cuts soon brought home the realisation of what the comfort zone was all about.

The two of us decided to carry out a test of a target staking plan to see how it would go and if it could improve profits or even turn a loss into a profit. Using a couple of very simple rules in regard to fitness, we used the tipsters' poll from the weekend edition of The Australian newspaper as the selections.

The two rules were:

  1. To be a selection the horse must have started within the last twenty days.
  2. It must also have had at least two starts this preparation.

In each race, commencing with the first-named horse in the Most Favoured column, the selection was the first horse to qualify under the two rules above.

Starting at the beginning of the racing season in 1998, we ran the test for a full year, ending up with some 1410 selections, an average of 27 selections each Saturday, and 347 winners, a win-strike rate of 24.6 per cent.

With a return of just $1318 ($1 bets), a level-stakes loss of $92 was incurred, giving a loss on turnover of 6.5 per cent.

Turning a level-stakes loss into a profit is quite a challenge for any type of staking plan to overcome.

The staking plan, which was targeted at only a very small profit of a mere 10 cents per race, did, however, turn the $92 loss into a reasonable profit of $111, which on a turnover of $910 amounted to a profit on turnover (POT) of 12.2 per cent.

However, the profit was gained with a fair degree of risk involved, for the largest bet required was $29, quite substantial given that the base bet started at a mere 50 cents.

The drawdown on the punter's betting bank reached $63.50, which was again quite substantial in the circumstances. If nothing else, the drawdown emphasised that being under-capitalised is one of the biggest issues a punter needs to consider when embarking on a quest such as the one we did. Notwithstanding, the staking plan had achieved its aim (target) and more so, over 200 units turnaround between a level-stakes loss and the targeted profit.

At the time, the mathematicians stated that it wasn't possible - that we got lucky and had the test gone on long enough a loss would have resulted, staking plan or not.

We didn't believe them, for when pressed as to what they considered a reasonable test of any staking plan, invariably they resorted to such terms as "infinity" and chose to ignore the significant test such as the one Roman and myself conducted.

There are many others who have also proved the worth of the right type of staking plan and while such plans will invariably work best with selections that show a level-stakes profit, they can also be used to turn small losses into profits. Small losses not big ones!

It would appear that many of these mathematicians fail to come to terms with the difference between casino-type betting and wagering on horse, dog or harness racing.

Unless the gambler has the ability to count cards, the difference in the odds offered by casinos is fixed; that is, a gamble on red or black, odds or evens and high or low on the roulette wheel is always a negative bet for the gambler - the odds are always the same and always in the casino operator's favour.

This is not the case with wagering. Bookmakers can and often do lose because the odds they offer are not "true" (as with casino gambling), being nothing more than "opinion" odds, the main reason why the mathematicians get it so wrong.

In August last year, I revisited The Australian's tipsters' poll and commenced testing them once again, some four years after the original test. This time I had revised the original staking plan, adding in far greater risk controls and restricting the maximum bet-size limit.

In next month's column I will continue this theme on target staking plans, detailing what has occurred so far this time around.

By E.J. Minnis

PRACTICAL PUNTING - FEBRUARY 2003