By Matt Elliott

Countless times now, punters have asked me why just about always they lose more in a betting session than they theoretically should. For example they lose more than the 16% takeout in the win tote pool, or they lose more than the 0.6% (roughly) on turnover one would expect to lose betting blackjack basic strategy.

So if you make 10 x $10 bets, then your turnover is $100 and you can expect to lose 0.6% of that which amounts to 60 cents. Why is it then that a punter will more times than not lose all their bankroll when playing such a game? Obviously we are referring to social gamblers here and not professionals such as Andrew Scott.

We will ignore the most obvious (but practically incorrect), answer to this dilemma, being the punter backed many more losers than winners.

Why does this happen?

The reason is simple. These punters are playing a game that is too big for their bankroll. Bankroll fluctuations, or if you like the standard deviation, steps in to ensure you lose your initial bank if you are betting too big. Now if you are a punter who sets aside a bank and bets conservatively in proportion to the bank, then this article may serve as nothing more than a refresher.

You may even wonder why we are always talking about managing bankroll fluctuations and not concentrating more on the form. The answer is simple. If you don't truly understand bankroll fluctuations, you don't have a chance of staying in the game long term. Managing bankroll fluctuations is just that important.

I do hope you have downloaded a copy of our simulation program which is available for free through the 'Betting Downloads' section of the website. Not only that, but I hope you have had a good play around with it and gained something worthwhile.

Let us run through an example. You back racing favourites and your strike rate is 27% and your average dividend is 3.7. These are fair assumptions and would put you very close to a long-run, break even situation. What would happen if you started with a $200.00 bank and you bet $20.00 per selection which is a typical scenario for many punters? Well if we simulate 50 bets using the calculator, you will find that you have a stagging 40% chance of losing all of your bankroll. This is of course betting what should be a break even scenario. Even if you bet a smaller percentage of your 'roll' a bankroll fluctuation can wipe you out. The bank is constantly fluctuating. There will be a point where the bank will dip to a level where you would 'tap out'. You might have hit the next five winners however you can't do that if you don't have a bankroll!

For many punters, this article may be nothing more than a refresher, however judging by the questions that Jess and I receive, many punters still can't understand why they keep losing their bankroll. They immediately look towards their selection strategy or blame their staking plan rather than conceding that recent results are only a natural fluctuation. You can be the best tipster in the world, but if you bet too big for you bankroll you will go broke. The question isn't 'if' but 'when' you go broke.

There is an excellent publication titled 'Flutters' by a John Robson which describes this 'random walk' concept clearly.

For those punters who grasp the ideas above, you may not have considered just what these bankroll fluctuations mean to the skilled punter who bets within his means.

Always remember that highly fluctuating banks are a good thing.

It means that social punters can walk away as a winner over a particular night or even week. In the short-run, it is the bankroll fluctuations (or if you like, the standard deviation), that dictate the short term results. It is these occasional wins which bring the social punter back time and time again. This is good for us, if there were fewer social punters, then the ratio of 'sharp' money against 'social' money would rise, thus making it tougher to get an edge over the more informed and skilled punters.

Beware of the random walk!

This article is protected by international Copyright © Elk Publications Pty Ltd May 2005 Please contact if you wish to reproduce this article elsewhere.