October 16, 2020

Sports Betting As An Investment

SPORTS BETTING SYSTEMS – Money Management

 

(Photo by Adem KAYA from FreeImages)

In this article, we will discuss the importance of money management when you are sports betting or sports trading.

 

 

THE NEED FOR MONEY MANAGEMENT

 

If you are only sports betting for fun and do not regard it as an investment this article is probably not very relevant for you – but then, in fact, this whole site is probably not very relevant for you, as we definitely take sports betting and especially sports trading very seriously and do regard it as a serious investment which is supposed to deliver a steady profit.

Money management is obviously important if you want to treat this as an investment. You would never treat your pension plans as a bit of fun, would you? This should be treated just as seriously!

 

BETTING BANK

 

The “betting bank” is the budget you have allocated for a specific system. Typically, a betting bank will be measured in “points” rather than money as such. So a betting bank can be a 100 points bank or a 200 points bank etc. The reason we are using points for a betting bank relates to the staking since you will want to know how much of the bank you are expected to stake on each bet or trade (and especially, if the system is not your own, then using flat £-stakes would mean something different to each person vs the actual bank).

If a system dictates that you spend 5 points per bet, having a 100 points betting bank means you will risk 5% of your bank on each bet. If you have set £1,000 aside for the system, it means that you will risk £50 per bet, as 1 point is 1/100 of the bank, i.e. 1 point = £10.

The size of the betting bank (that is, the number of points) depends on the strike rate and the average stake the system is based on.

Obviously, the lower the strike rate, the more points the betting bank should have and likewise, the higher the average stake (in terms of points), the higher the number of points the bank should have.

If you have a system with a low strike rate and a high average stake, then obviously you will expect that system to have a very high number of points.

On the other hand, if the system has a very high strike rate and only stakes at low points-levels, then the system should cope with a low number of points for the betting bank.

 

THE STRIKE RATE

 

The strike rate is essential to know for a couple of reasons.

The strike rate obviously shows high often a system wins. So ideally you’d like a system with a high strike rate, but with that comes low odds winners, as the odds are very tightly connected to the likelihood of an event to happen.

Strictly speaking, this is the formula:

1/Odds = Probability

So, if you back something at odds 2.0, the probability of it to actually happen is 1/2 = 0.5 or 50%

I.e., odds 2.0 means there is a 50% chance of something happening. This is the case if you flip a coin, where there is an equal chance for heads or tails. You should be offered odds 2.0 by someone if you were to back either heads or tails. If you get higher odds, you have an advantage (aka an “edge”), while if you are offered below, it’s the opposite (and you shouldn’t take the bet).

Another example: On a dice, there are 6 potential outcomes, i.e. each outcome has 1/6 chance to come through. If you are offered the chance to bet on one of the numbers to occur, the odds should be 6.0.

The strike rate is 1/6 (or 16.67%), so after 6 throws you’d expect each number to arrive exactly once. Now, this obviously doesn’t necessarily happen like that. You could actually see, say, the “5” side come up 6 times in a row. What the strike rate tells us is that on average each side will be “the winner” every 1/6. So, if you threw the dice 1,000 times, all 6 sides will be expected to show approximately 167 times each. In reality, it will be close to that but with a variation of plus/minus something. The higher the number of throws you make, the closer to the average you’ll get.

 

ESTIMATING THE LONGEST LOSING/WINNING RUN

 

We actually touched upon something very important there. The strike rate shows us what the average expected number of winning outcomes is (shown as a percentage of the total number of outcomes).

As a bettor or trader, you want to know how many losing bets in a row you can expect and the strike rate can be used to calculate that but in itself, it doesn’t give us the answer. If you flip a coin then, yes, the probability of each side to come on top is 50%, but don’t expect it to always change “face” from flip to flip – i.e. if the first flip showed the “head” then you can’t be sure that the next one will be tails.

If it was so simple, it would be extremely easy to make money on simple gambling…

So, we need to know what the actual losing run is, i.e. how many times in a row can we expect the unwanted outcome to occur, given a specific strike rate?

This can be calculated in a number of ways – both of which will be made much easier if you have access to a spreadsheet.

The first method is the simplest one, as it is just a formula.

You need only the likelihood of something to happen and the number of total outcomes.

The formula is the following:

LN(#Bets) / – LN(probability)

LN is the Natural Logarithm

So, let’s say you have a system that has a strike rate of 25%. You now want to know what the longest losing run is.

If the strike rate is 25%, the likelihood of the system NOT delivering a winning bet/trade is 75%. Let’s see what the estimated longest losing run over 1,000 bets is.

LN(1000) / – LN(0,75) = 32

So the estimated longest losing run (LLR) for a system that has a 25% strike rate is 32.

Likewise, if we want to calculate the estimated longest winning run (LWR), the number in this example is 6 (rounded down).

The estimated longest losing run is a great indicator of what you in real life can expect from the system. However, it’s actually not quite good enough. Yes, it may be that the system above at worst delivers 32 bets in a row that loses – but there is no guarantee that the sequence isn’t like this:

1 winner

20 losers

1 winner

32 losers

1 winner

17 losers

Etc.

After this point, it may pick up and delivers better, but within the 72 bets shown you actually only had 3 winners, equal to a strike rate of 4% and, depending on the winner odds, a massive drawdown. If the winning odds were 2.0 for instance, the drawdown would be -66 points. Would you be able to get through that and still believe in the system? Most wouldn’t!

In other words, we need to know what the expected drawdowns are, especially the maximum drawdown, in order to decide on the correct betting bank for the system.

Now, it is not as simple to calculate the Maximum DrawDown as it is to calculate the LLR/LWR. Some just use the LLR as the foundation for deciding the betting bank as they take the LLR and multiplies it by 3 since this is a pretty safe factor to use. So if your LLR is 32, then the betting bank should be 32 x 3 = 96, which will then be rounded to 100, provided the average stake is 1. If the average stake is not 1, you should multiply the 100 by the average stake, i.e. if the average stake is 5, then your betting bank should be 500 and so forth.

 

MAXIMUM DRAWDOWN

 

The Maximum DrawDown (MDD) shows the biggest drawdown you can expect for the system to deliver. Let’s say we have a sequence of 1,000 bets. At bet no. 117 the profits reached 60 points which were the highest so far. From here it started to go down, then a bit up, then more down, then a bit up again, and then even further down. At bet no. 271 it reached the lowest profit (since it topped at bet no. 117) at only 3 points profit. In other words, there had been a drawdown since bet no. 117 of 57 points. After bet no. 271, the profits started to increase gradually again, and we never saw this kind of drawdown again within the 1,000 bets. So, in this case, the actual MDD is 57 points.

If you want to calculate the estimated MDD, the best way to do it is via Excel (or any other spreadsheet) and the use of the so-called Monte Carlo method.

The Monte Carlo method is basically a method that is based on testing the outcome of a certain event on a very large number of events. It can easily be used as a method for calculating the longest losing or winning runs as well (in fact, this figure is included when you use the Monte Carlo method for calculating the MDD so if you intend to do this, there is no need to do the above-mentioned formula for calculating LLR or LWR).

I will not be going into details here on how to actually do it, as it’s a bit tricky to explain without showing in Excel how it’s done, but I have created an Excel sheet that you can get for free. Just contact me at slayde@serioussportstrader.com and I will send it to you.

In the meantime, if you want to know how to easily create an Excel sheet for the Monte Carlo system, I suggest you look at the following YouTube video: https://www.youtube.com/watch?v=UeGncSFijUM&t=304s

Once you have calculated the MDD, this figure should be used as the foundation for setting your betting bank. I would multiply it by 2 or 3 in order to set my actual betting bank, but this is personal. Obviously, you shouldn’t just use the MDD as the number of points for the betting bank since it basically means that you will go bankrupt at some point… Remember also to include the average stake in the calculation the same way as I showed above.

 

STAKING PLANS

 

Now, the subject of “staking plans” is worthy of an article in itself – and it will be, so we’ll only be discussing this in brief here.

It is, however, closely related to the betting bank, since the way we calculated the betting bank in this article is pretty much based on a simple flat staking system, i.e. an approach where each bet has a fixed 1,2 or 5 points stake (or the like) and you know the average stake per bet.

There are numerous different staking plans available for free on the internet – just google it and you’ll see. It is our intention to collect the best known and perhaps the ones we would recommend the most (as well as the ones we definitely recommend against, such as the infamous Martingale staking plan).

Just note this: If your betting or trading system doesn’t deliver a profit at level/flat stakes, no clever staking plan will change that. Most likely, it will make sure you lose your money faster.

Having said that, using the flat stakes is never a bad idea and, simple as it is, is one of the best plans to follow. Now, if you do have a system that gradually grows your betting bank, it is sensible to consider compounding your profits, i.e. to let the points stake reflect a percentage of your betting bank at any given time, alternatively to change it on daily, weekly or monthly basis. This will grow your betting bank much faster than just by staking the same amount on every bet.

In the diagram we have shown an example of a system, that has the following data/statistics:

  • The bank (capital) is £1,000
  • The average ROI per bet/trade is 5%
  • The average points per bet/trade (i.e., the average stake) is 2
  • The number of bets/trades per month is 100, i.e. just above 3 daily

The flat stake approach will deliver a monthly profit of £100, i.e. 10% of the capital (so ROC, Return Of Capital, is 10%)

The Compounding method obviously delivers the same in month 1 but after that, it begins to increase, and it’s easy to see the difference after the first year.

After 12 months, the bank at flat stakes is at £2,200, i.e. an increase of 120%, while the compounded staking bank is as £3,138 – an increase of 214% and 78% more than the flat staking. This is in itself a very respectable improvement, but it’s after 24 months that the difference is enormous.

While the flat staking plan is obviously just rolling along at the same steady pace and now is at £3,400, another £1,200 increase of the bank, the compounded staking bank is at a totally different level. It now stands at £9,850 – almost 3 times the capital vs the flat staking bank!

Einstein is supposed to have said something along the lines of:

“Compounding is the 8th wonder of the world.

He who understands it earns it.

He who doesn’t pay it.”

He was no fool.