Winning & Losing Cycles

What is Variance?

Variance is the difference between what you can expect on average in a given cycle of bets and what actually happens. It is your experience of winning and losing runs as a punter.

For example, you might average a 35% strike rate and +10% profit over a large number of bets, but your results within a given group of 10, 20, 50, 100 bets or more will rarely if ever be equal to that average.

Some results will be much better and others much worse. In one group of 100 bets you might make +20% profit on turnover and in other group of bets you might lose -10% on your turnover.

It will have nothing to do with your strategy, mental focus or being in / out of form as a punter… it’s simply random statistical variance that comes from trying to predict events where your probability of success (i.e. strike rate) is typically between 20% and 50%.

A good way to demonstrate this is to think of coin tosses. You know that you have a 50% chance of guessing the correct result and winning each time, which is just like backing a true even money ($2.00) chance in a race.

If in a group of 20 coin tosses you only get six correct, does that mean you are in poor form or have a bad strategy for guessing the toss of a coin?

Of course not, you know there is an element of randomness to the results (i.e. variance.) that is outside of your control. You recognise in this case that you have simply been unlucky.

The same applies to betting on the horses, or any event for that matter, yet we tend to associate a small sample negative run of results as meaning that we’ve done something wrong.

If you back 20 horses that are a genuine $2.00 chance in their respective races at a price of $2.30, then on average you have +15% profit edge… it’s a great strategy!

However, in one group of 20 bets you might only land 6 winners and you will lose 31% on your turnover. It doesn’t mean that you were on the wrong horses, they were all great betting value… it was simply bad luck (i.e. negative variance.)

Understanding winning & losing cycles

Understanding the nature of winning and losing cycles that can occur as a punter is essential knowledge.

Why? Because it doesn’t matter how good you are, you will experience big winning runs and big losing runs… it’s the nature of the game and there’s nothing you can do about it.

Dealing with winning runs is relatively easy, but when things aren’t going so well you need to call on that knowledge of winning & losing cycles to help maintain your confidence.

You need to rationalise that you aren’t necessarily doing anything wrong and that a natural slump in results will soon turn around.

Without that knowledge, a losing run will see you lose confidence in your selection method and either give up on it or adopt other destructive behaviours such as loss chasing, which will eventually send you broke.

It’s possible that you have given up on profitable selection methods in the past because you interpreted a slump in results as a sign that the method doesn’t work, when in fact that slump was within the normal expected range for a profitable betting method.

What to Expect?

While we can’t predict whether our next series of bets will be a big winning run, a big losing run or somewhere in between, we can use the statistical function of standard deviation to measure the range of possible winners in a given number of bets and therefore the results that can be expected.

Standard deviation is calculated by taking the square root of (the number of races x the win probability x the losing probability.)

For example, if you average a 35% strike rate then in 200 bets you would expect 70 winners. The standard deviation for this set is calculated as the square root of: 200 (number of races) x 0.35 (win probability) x 0.65 (losing probability.) = 6.75 wins.

In a statistical sense your actual number of winners will fall within one standard deviation 68% of the time (63 to 77 winners) and two standard deviations 95% of the time (56 to 84 winners.)
For the sake of simplicity, we will ignore the likelihood of results falling outside the 95% range.

If your average dividend is $3.14, providing a long-term profit of +10%, then natural variance in your number of winners over 200 bets could see your results fall anywhere within the following ranges:

200 Bets	Winner			Profit %
Period	From	To	Avg Div	From	To
Average	70	70	$3.14	9.90%	9.90%
68% of the time	63.25	76.75	$3.14	-0.60%	20.60%
95% of the time	56.5	83.5	$3.14	-11.20%	31.20%

After 200 bets your profit will be somewhere between -0.6% and +20.6% POT most of the time, but you could be showing as much as -11.2% or +31.2% POT.

That variation in results may shock you, but it’s the reality of what you can experience in a 200-bet cycle.

Time vs Number of Bets

If you only bet on Saturdays, then a 200-bet cycle could take you 4-6 months to accumulate.

If you were making +20% POT after 6 months you could be forgiven for thinking that you were one of the smartest punters in the game.

However, the 6-month period is irrelevant. The important factor is the number of bets and what that means for your likely range of results.

In 200 bets a +20% profit may or may not be reflective of the long-term edge, it could have a large part of positive variance built in.

You could follow the same selection process and make the same types of betting decisions over the next 200 bets (6 months) and lose 10% on turnover. It doesn’t mean you’ve done anything wrong, that cycle may simply negative variance from your long-term average.

One of the great uncertainties and challenges of betting is that we never really know whether our most recent sample of results is representative of the long-term or reflects positive / negative variance. This is only revealed through time and much large sample sizes.

The Importance of Sample Size

The larger the sample size the closer to the long-term average your results will be.

For example, assuming the same 35% strike rate and $3.14 average dividend, after 5,000 bets you can expect somewhere between 6% and 14% profit, with your result falling between 8% and 12% profit most (68%) of the time. The highs and lows are much closer together.

The reverse applies when dealing with smaller samples. In a 50 bet cycle you can expect anywhere between a -32% profit and +52% profit.  The smaller the sample size the much greater the gap can be between highs and lows.

Using my example of betting only on Saturday’s, 50 bets might take you 6 or 7 weeks to accumulate.

If you were showing -20% POT or even -30% POT over 6 or 7 weeks betting then it’s very likely you would assume that you are doing something wrong and look to change your approach.

However, you could be following an approach that delivers a good long-term profit and simply be experiencing extreme negative variance, but still within normal boundaries.

Practical Implications

The concept of standard deviation shows just how volatile and unpredictable punting outcomes can be over what we might think is a large number of bets that should provide some stability.

When you are dealing with winning probabilities (i.e. strike rates) that are well below 50% then natural variance can be huge.

It’s this concept more than anything else that makes successful betting such a challenging pursuit.

When you are losing 20% on turnover after 50 bets is that just natural variance? Or is the method you are following a long-term loser?

When you make 10% POT in one 200 bet cycle and lose 10% in the next 200 bet cycle, is that just natural variance or is your approach no longer profitable?

How long do you persist to find out? Do you wait another 200 bets and hope things get better, only to lose another 10% on turnover? Or do you change things now and look to improve?

What happens if you do make some changes and start off with a losing 50 bet cycle?

The truth is that you never really know. It’s that constant state of uncertainty that is eventually the downfall of many punters.

Having a profitable selection method is just one piece of the puzzle and in itself does not guarantee success.

You must also have the knowledge of how far your results can deviate from that average in a given number of bets and most importantly the capability to manage yourself through those periods.

To be a successful punter you need to be able to understand, accept and manage the inevitability of going 50, 100, 200 bets where you might be losing or at best making virtually nothing.

You also need to be able to keep yourself in check which you go through the same number of bets making a massive profit. Neither represent your likely long-term result as a punter.

There are never any guarantees and it’s only through time and experience that you can develop the confidence that you do in fact have an edge over the game.

What all of this means is that from time to time we all need to ask ourselves the question:

“do I really want to be a punter?”

In my opinion there’s no more challenging or satisfying intellectual and psychological pursuit. But answering “yes” to that question means accepting the nature of the game.

You need to have persistence and determination to work through inevitable losing runs and the wisdom to put big winning runs in perspective. If you can do that, then any goal you set for yourself as a punter is achievable.