## Variance, a statistic concept

### The coin example

If we toss the coin 3 times it could be tipped 3 times and tails 0 times and vice versa. It could also be tapped twice and tails twice, or the reverse twice tails and tails once.

The purpose of variance is to find out whether the probabilities associated with the result obtained (twice heads and once heads for example) are greater than the theoretical probabilities. Let’s do this calculation together. On 3 launches, we can obtain the following results (note P for Pile and F for Face): PPP / PPF / PFP / PFF / FFF / FFP / FPF / FPP

That is to say 8 combinations in total and 3 possibilities of falling on 2 tails and 1 time tails: PPF, PFP, FPP.

So we have 37.5% to fall on a favorable draw. If from the 1st try, we get 2 tails and 1 tails, then we had a 100% success, ie a difference with the theoretical probabilities of 63.5%! We are in a very lucky time. However if we get this draw at least 10 attempts, we have had that 10% success is 27.5% less than the theoretical probabilities would like: we are in a period of “bad luck”.

Thanks to this calculation, we can therefore know if the results we obtained will be less favorable or more favorable if we continue to flip the coin 3 times in a row. Indeed, probabilities and the law of large numbers indicates that the more we flip the coin, the closer the probabilities of the result we have obtained will be to the theoretical result of the probabilities.

## Variance in betting

Conversely, if we know that we are with lower results than predicted by

probabilities, this means that we will soon have a better period and therefore more gains than we should (always relative to the probabilities). So this is where you have to keep your methods, have confidence in your refined analytical frameworks.

This is one of the reasons that motivated us to develop the functionality of the analysis frame on the Bet Assistant platform. Indeed, in bad times we often question everything. But if we have the tools that help us stay the course and rely only on statistics that don’t change in the event of a bad result, we will be able to continue to carry out our analyzes by taking less account of our emotions.

With this automated analysis framework, information is collected automatically, your analysis logic is intact because it has already been thought through and you can continue to bet while being at ease to get through the period of poor results.

## What do to against variance ?

### Bankroll management

### Valuebets

The other technique takes us back to our previous article on valuebets. By playing this type of bet (identifying odds at poorly adjusted bookmakers) you change the theoretical odds in your favor. And yes ! 🙂

If you read the article you saw the difference in results on just 100 bets. Imagine the results on 1000 bets, 10,000 bets! By playing valuebets you gain an advantage over your theoretical probability (you bet on a 50% probability of winning while in case of winning you will be paid for a 40% probability).

Let’s say you flip a coin 10 times and still bet heads because it is a value bet (heads are paid 60% instead of 50% or a 1 in 2 chance of heads being paid). You are “unlucky”, the coin is tails 6 times and heads only 4 times. By having played valuebets on the face, you may be unlucky but you are less losing than those who do not play valuebets.

Indeed, you will win 6.66 units against 5 without playing the valuebets. Your loss is therefore less important and will allow you to better manage the period of poor results.