I read Roman Koz's article Level Pegging in PPM (February 1999 issue) with great attention, as staking plans have been an area of interest for me over the past 20-odd years.

I've known Roman for some three years now but I have never met the man. Why? Because we know each other through the AUSRACE email group and have had many interesting and varied discussions on punting topics in that time.

I respect Roman's knowledge and punting ideas and one of his articles, Ratings By Pairs (PPM, March 1998), contained one of the most innovative ideas that I have seen for a long time. But I cannot agree with his ideas on progression staking as I believe there are three certainties in life for us mere mortals. They are death, taxes and Going Bust on Progression Staking Plans.

It may not happen today but it will happen. I assert that going bust on a progression-staking plan m the long term is a near certainty. Roman recommends that to use the Six-Point Divisor Plan (SPDP) for a seemingly small target of $12 and an initial bet of $2, you need to set aside a bank of at least $2400.

Good advice but it doesn't go far enough. I believe that punters need to examine their own betting profile before deciding to use the SPDP. Why? Because the 'risk of ruin' from a progression-staking plan is inexorably based on the expected longest-losing sequence which, in turn, depends upon the average odds of the horses that you support.

If the average odds of the horses you support are 7/1, then using the SPDP will be an unmitigated disaster. However, if your average odds are 6/4, then the SPDP may be a workable plan.

Let me explain my philosophy for deciding whether a progression staking plan or, in particular, the SPDP is the plan for YOU.

19990404

Step No. 1: Determine your Betting Profile.

You need to have a very good idea of the odds spread of the horses that you support, both winners and losers. If you keep records of your past bets (hey, doesn't every serious punter?), then go back and work out the average odds of the runners that you support.

You do this by adding up the odds of each-and-every bet that you have made and then dividing the total by the number of bets * For example, if you had bets of evens (50%), 2/1 (33.3%), 4/1 (20%), evens (50%) and 9/1 (10%), then this adds up to 163.3% which when divided by 5 equals 32.7%, close enough to 2/1. Alternatively, you can tally up your distribution of bets as shown below:

Here the total of all odds for the 155 bets made is 48.03 and the average is 48.03 ÷ 155 = 0.31 (31% expressed as a percentage), which is about 9/4.

Step No. 2: Determine how many bets you are likely to have in a year.

I average about 10 bets per week. This means about 500 bets per year. So I use 500 bets as a good guide to my year's betting action. Note that you need only include bets that you intend to make using the SPDP.

Step No. 3: Understand your risk of ruin.

Determine how many years you expect to be betting on the SPDP and multiply this figure by the number of bets that you make in a year. I would like to feel confident that I would 'survive' for seven years.

For me, that's 3,500 bets where I am hoping that I will not strike a horror run of cuts. But this is the grand uncertainty of punting. No-one knows when the fickle finger of fate will arrive.

But it's worthwhile to realise that every extra bet you make using a progression staking plan increases your chance of having a horror stretch and blowing your entire bank. For example, if I placed 3,500 bets with average odds of 9/4, then I have: a 99% chance of a 15-plus longest-losing sequence; a 50% chance of a 20-plus longest-losing sequence; and a 2% chance of a 30plus longest-losing sequence.

A progression staking plan that can recover from a continuous losing sequence of 30 ... I'd like to see that!

But if I restricted myself to just four bets a week (at least on the SPDP), then that's some 1400 bets over seven years with a 54% chance of a 17-plus longest losing sequence, whilst a 13-plus longest losing sequence has a 97% chance of occurring.

Another approach is to only bet on the SPDP when your bets are 6/4 or less. If your average strike rate for these bets was, say, 50% and you have restricted yourself to 1400 bets, then the expected longest losing sequence is 9-plus with a 62% chance, whilst a 7-plus longest losing sequence has a 98% chance of occurring. Food for thought, don't you think?

LEVEL STAKING OR PROGRESSION STAKING ... OR NOTHING?
Question: Why does everyone want to know what the level-stake Profit on Turnover (POT) is for a system?

Answer: Because deep down they don't trust progression staking plans to reveal the true intrinsic worth of the system.

So, when we are considering what staking plan to use, we need to know not only our betting profile but also whether our selections will show a Long-Term Level-Stake Profitability.

If the answer is NO, then you should use progression staking because you may still win and, besides, level-stake betting will be a slow and painful death. An alternative to this scenario is to not bet at all but, as the saying goes, 'hope springs eternal to the human breast' (or, as my mum used to say, "yolk clings eternal to the human vest") and so punters keep on punting.

If the answer is YES, then there is still the dilemma of whether to use level staking or progression staking. So read on ...

How do you determine long-term level-stake profitability? It's pretty simple really. Remember how you have kept a record of all your bets over the past 12 months? Well, if you have the SP odds for each and every runner, then add up their odds (e.g. evens = 50%, 2/1 = 33.3%, 4/1 = 20%, etc.) and then divide the actual number of winners by this total percentage.

This may give you a figure of, say, 1.08 which means your levelstake profitability is 8%. If you don't have all this info, then the next best guide is to use your levelstake profit on turnover. Find this by dividing the $ amount that you won (or lost) on the punt over the past 12 months by the $ amount of your average weekly bets multiplied by 52. For example, if you won $260 last year and you averaged $100 of bets per week, then this equates to a one-twentieth, or a 5% profitability ratio.

USING A SIMULATION TO COMPARE LEVEL STAKING WITH THE SPDP
The maths of doing a theoretical comparison between level-stake betting and the SPDP would be a long and tedious exercise. However, modern-day computers give us the opportunity of doing quite complex comparisons using Statistical Simulations.

I have developed a very comprehensive Staking Plan Simulator using Excel's powerful facility provided by the Data Analysis Tool Pak add-on and I use it to compare the same simulated results across a variety of different staking plans. Currently, I have programmed in six such plans. These are:

  1. Level Stake Plan: Level Bet of Minimum Bet for every bet.
  2. Recoup Overall Bank Loss: Bet wagered is Minimum Bet but if Bank is losing, then next bet is the amount required to return the Bank back to a no-loss position.
  3. Recoup Current Sequence Loss: Bet wagered is Minimum Bet but if the current sequence is losing, then next bet is the amount required to return the current sequence to a no-loss position.
  4. The "Jean Le, Rond D'alembert" Staking Plan: First Bet is as per Initial Bet. If a winner, then decrease next by Bet Adjustment. If a loser, then increase bet by Bet Adjustment. If bet at any time becomes zero, then change it back to the Initial Bet.

Click here to read Part 2.

By 'Anomaly' Nick

PRACTICAL PUNTING - APRIL 1999