The point of this article is to try and highlight to you how a betting style with less volatility ( up and down swings of profit / loss ) can be beneficial when it comes to growing your betting bank.

To help demonstrate that I will be using some real world data from our own service. However bear strongly in mind is that the key principles of the thought train are what is important. You can take those with you and apply them to whatever tipster, racing system or method you care to be involved with over your future betting career.

Anyhow this all started when a client here asked me

"Would it be a silly idea to take any Account Bet here that is suggested each way and instead just bet it win only?"

[ By way of extra explanation Account Bets here are one component of the overall service.

Think selective bets that are deemed value AND they also have to be in more liquid races as it is easier to get bigger stakes down on stronger race markets.

See an old sample message ==> here. The Doncaster 3.45 race was the Account Bet within it.]

That was a fair and interesting question I thought. I did not know the answer to it however so I had to go off and rejig my spreadsheet to churn out the desired data. The time period I used was between Jan 22nd 2022 and 23rd Nov 2024. The 2022 date is not arbitrary. Rather it matches an historic message format change here.

From here on in I am going to refer to the original staking advice as Default Style.

ie if it was tipped as each way then each way is what we report figures on here

Win Only style then is converting any each way bet into a win only bet.

So now let's compare the two variations using simple level stakes of 1pt total per overall bet.

The key metrics here are:

So very good for the Default Style. 60 odd pts profit would be £6000 for someone betting £100 on each bet. Edge is very strong as well. Compare 39% to say the 2.7% edge the casino would have over the player in roulette ( European tables )

But what then about Win Only Style?

Well that win only idea looks sort of better does it not? Level stakes profits have jumped up. Profit on turnover has jumped up too. Perhaps our client friend has asked a shrewd and clever question. More profits and greater edge. Happy Days. Off to Barbados we go? OR is there more to think about?

Well yes.

Note the max drawdown figure has increased. I do not think there is any shock horror surprise there. Most punters understand that each way betting offers a lot of stake protection when instead of winning as they are supposed to, the badly trained beasts you select annoyingly come second or third.

Each way tends to smooth the ride a lot. Both ups and downs or volatility, will be a little less extreme. A run of seconditus can cause little damage to a betting bank when each way is employed. The same run of post hitting with win only style can knock a bank down to a much bigger degree.

A common approach to betting bank management is to split your bank into a number of points. eg let's hypothetically say you had £1000 as your betting bank. You decide that you want to use a bank of 50 points. Each point then is worth £1000 / 50 = £20. That stake you then always use moving forwards.

There are some cons to the approach but the most appreciated upside is day to day simplicity. One still however has to initially decide the question of "How many points should be in my bank?"

There is no single answer to that question. It really depends on the style of betting one is doing. A method focussed on long shots will require more points in ones bank than say a method place only betting on favourites. Long odds decrease strike rate and increase volatility.

Of course the real life of using most methods, systems or tipsters is that odds can vary quite a bit across different horses. So how does one go about coming up with a half baked semi sensible answer to the "How Many Points" question?

Well one angle into it is to ponder the historic Max Drawdown figures of your method / tipster etc. Picture the law of sod was strongly in play. The day you decided to start betting was the very start day of the worst ever period in the recorded history. How many points would have needed to be able to survive that? Perhaps one more than the biggest ever historic drawdown.

The future however can not be guaranteed to be a perfect mirror of the past. The greater the sample size you have of historic action the increased confidence you can have in any historic poor patch being indicative of the worst the future holds. It is always prudent to build in a safety factor. So you can take that worst period ever and then add a safety factor of your choice to it. Let's say you choose a factor of 2. In that case you would be using a bank that would have been double the worst ever drawdown.

If you feel more comfortable with a safety factor of 3, all good. You need to design things to suit your own unique tolerance to up and down swings and also for your appetiteto risk and reward.

Let's return now to the nitty gritty of our real data example.

For the sake of argument let's say we settle on a safety factor of 3 as suiting our temperament.

Default style bank of points = 12.88 (drawdown) * 3 (safety factor) = 38.64 pts

Let's round that to 40

£1000 divided by 40 = £25 per point

60.16 pts profit thus = £1504

Now For Win Only Style

17.23 * 3 = 51.7 pts

Round to 50

£1000 divided by 50 = £20 per point

78.70 pts profit thus = £1574

With this outlook the true value of all those extra level stakes points starts to diminish. Every point won is worth less real money because one needs a greater bank of points to deal with the extra volatility.

Level stakes analysis is arguably best reserved for judging the degree of edge or profitability a series of bets has. It is a very useful tool but you need to think a step further if you want to ponder the ability to best grow a bank.

Here is a quick test for you.

Two selection methods each have a profit on turnover of 20%.

Over the past year Method 1 made 50 pts level stakes profit

whilst Method 2 made 100 pts level stakes profit.

Which is the better method?

#1 - Method 1 @ 50 pts

#2 - Method 2 @ 100 pts

or

#3 - I have not got a clue?

I would argue the answer is #3 on the grounds that there is insufficient data to decide. Methods 2 for example could be rooted in long shot golf tournament winner tips at 100/1. If so a huge huge bank would be required and thus each pt won is of little true value. Method 1 on the other hand may centred on some higher strike and less volatile style of thing. From a bank building perspective those 50 points level stakes could be worth a lot more real money that the 100 points.

Well he once proclaimed compounding as the "8th Wonder Of The World" so I guess he would bring an element of that to his method instead of betting eternally the same level stake. Let's say every year he could double his bank.

£1k ==>£2k

£2k==>£4k

£4k==>£8k

£8k==>£16k

And so on. Things would seem slow at the start but because he keep reinvesting his winnings, his bank compounds bigger and bigger over time. Arguably it is as hard a mathematical sweat to jump from £8k to £16k as it was to jump from £1k to £2k.

Now you can take that simple bank of points concept and bring an element of compounding into it. Perhaps one double the £ value ofa point when ones total bank has doubled. Or perhaps you go sooner and recalibrate at the point of say a 50% increase. One can do that yes but it is not as pure in theory as an approach dedicated to the compounding concept from the ground up.

Our friend Albert would no doubt give a thumbs up to the principles of Kelly Staking. If you have not already done so, worth a read is that Kelly Staking article Sections of it such as "Can Kelly Help Grow My Horse Racing Bank To Infinity?" are equally applicable to the thought train of this page here.

Kelly however only really works properly for simple win only bets and won't handle well the idea of each way betting or covering more than one horse in a race. So whilst therefore we can not use Kelly here for this real world horse racing results data, we can indeed aim for kellyesque analysis using % of bank staking.

Ok so what is % of bank staking? Well it is pretty simple. One always stakes a set percentage of your current bank on the next bet. As your bank rises or falls the real world £ value of your stake also increases or decreases.

Say you start with £1000.

You decide 5% is apt to the method you are following.

5% of £1000 = £50

The horse wins at 2/1 for £100 profit.

Add that to your bank and it now stands at £1100

Next day 5% of your bank will be £55

Sadly that day you had a loser

The bank now stands at £1045

The next day you will be staking 5% of that which is £52.25.

Spreadsheet analysis will work to many pedantic decimal places but in real life of course you are free to round stakes as you see fit.

A core question of course very similar to the "How many points in my bank" question is "What percentage of my bank do I stake?"

This is semi similar to Kelly staking thinking. However whilst Kelly will provide an answer for a single betting punt, here we are more so considering a percentage based on a larger historic data set of punts.

There are two components to that "What Percentage" answer.

#1 - You need to fit to the nature of your method. Similar to Kelly high strike higher edge selections will merit increased staking aggression than lower strike / lower edge.

#2 - You need to fit to your own temperament. There is no point you setting off on a high aggression path even if mathematically sound, if the violent swings of up and down quickly scare you out of the saddle. Go less aggressive and the ride will be smoother and you will be more prone to stay on the horse longer term.

Ok with the basics of percentage bank staking out of the way let's revert back to the Account Bets dataset here and the question of default style with each way bets versus the on the nose win only approach.

The spreadsheet I am using is here ==> win-or-each-way.xls

So my basic process was to ask the spreadsheet

What % of bank as a stake produced the maximum growth of a £1000 starter bank?

Then

Show me the money. What was the final bank figure?

This same question was asked of both the default style with each way bets when advised and of the win only style. In effect we are asking for the absolute optimum back fitted staking percentage that would have produced the most wonderous bank growth possible.

The output was

As anticipated the staking percentage ( to achieve maximum back fitted profit ) for the increased volatility win only approach is lesser than that of the more steady default style that incorporates each way.

Similar as to with the bank of points approach, each win thus produces less real world profit.

I found it quite surprising the difference in both stake % and end bank £ figures for that unique dataset in question.

I would suggest that would be a bit risky. I would see those figures a similar in nature to full Kelly betting. Similar arguments as made in the Kelly article would apply here too. With Kelly and sports betting one should assume that one's true odds figure is imperfect. Bear in mind over aggression tends to carry enhanced risk compared to under aggression and application of a Safety factor to Kelly is usually sensible.

Similar applies here when asking what percentage for a dataset.

History is reasonably set ( but in the real world each individual may get slightly different odds due to getting barred at different bookies etc) however what we do not know is the future. It is only an educated guess that the future will mirror somewhat the past. Add such uncertainty into the equation and it is prudent to sandbag expectations. Again the thought train would be one of building in a safety factor. We can take that back fitted optimum percentage and reduce it by a safety factor of choice.

If again we somewhat arbitrarily decide a safety factor of 3 is our choice then we would get

Whilst end bank figures are obviously lower than the full on bigger back fitted percentage, the approach has additional safety for an uncertain future. Plan for the worst. Hope for the best.

Adding in my own personal tuppence worth view of the world I would add in the question

One answer to that is over aggression. They want too much too fast. They stake too aggressively. A bad patch surprises them and their bank gets wiped out.

Better to have the mind of a tortoise and think longer term. Get yourself EDGE then permit sufficient TIME for compounding magic to do its thing. The hares tend to set off too fast and they end up wrecking themselves at the first corner.

Note how those numbers despite being smaller than the first set still put you well on the curve compared to the annual doubling of Einstein above. The hare will favor get rich quick.The tortoise will favor longer term progression along a path.

Sticking that £1000 in your bank account for two years would leave you with what? Not much more than you started with I guess. £1060 if getting 3% interest.

Bear in mind real inflation and that £1060 would probably buy fewer loaves of bread than the original £1000 would have done. In the real world excess safety can be very dangerous.

Note that there is no one true answer to safety factor.A bit of maths on the numbers is one component. After that individual temperament to up and down swings comes into play. Circumstance too can have impact. Picture for example BOB who has a significant proportion of his wealth invested in something. Preservation should be a key part of his thought train. Jim on the other hand may be working with a small sacrificial pot he could fully lose without sweat. Jim's attitude to risk and reward would probably be higher up the scale towards the boom or bust direction.

Another potential circumstance further down the line is that one has already doubled one's starting bank. At that point original investment is removed and you move forwards playing only with the house's money. Such a scenario can alter one's happiness to increase aggression or risk factor. So even a single individual may opt to alter safety factor due to circumstance at the time.

Try not to over focus on the precise numbers from the unique data set from here that I was working from. More important for your longer term is grasping some of the core principles and using them to your advantage over your future punting career.

General dogsbody and assistant to Guy Ward ( our horse racing analyst )

here at the www.Mathematician-Betting.co.uk website.