Staking Plans

Like this lesson?

Share it with your friends!

Staking Plans

Betting Banks

Many professional punters like to have a set betting bank (size varies depending on wealth) from which they place all their bets. This allows them to easily keep track of profit and loss because all winnings and losses are coming from the same bank. It also allows them to stake set proportions of their bank on bets which reflect their confidence in the selection's chances. Profit from the bank is periodically withdrawn or withdrawn when it reaches a certain amount to be used for non-betting purposes.


You may have a betting bank of £1000, from which you may withdraw profit every time the bank reaches £1500 or instead whatever profit has been made each three months.

Unit Stakes

Assigning unit stakes to bets can be useful as it makes the punter more disciplined and less likely to over bet an event. Sometimes a maximum and minimum unit stake is used, from one unit to twenty units for example. Depending on the seriousness of the punter a unit may be £1, £10, £100 or even more. These units are usually referred to as points.

The more disciplined a punter the smaller band of units they will probably use. This makes them even less likely to over or under bet an outcome as the difference in confidence between units will be even more clearly defined in their mind.


You may have stakes varying from 1 to 5 points. Each point is worth £20. A minimum bet for you would be £20 and a maximum bet would be £100.

Kelly's Criterion

Kelly's Criterion is a formula that is used to determine how much of a bank should be risked on a given bet. The formula takes into account the odds of the bet and the probability that it will win and the probability that it will lose. This does have the advantage of ensuring the whole bank is never lost on a bet and helps to steadily increase the bank. A disadvantage of this is that there is no way of guaranteeing that money won't be lost. In fact, there is a 1/3 chance of halving the bankroll before it is doubled.

Stake Calculator
Your Betting Bank
Chance of Winning
Best Odds (decimal)
Suggested Stake

Garethp1981 has written a blog about this staking plan, he asks Does Kelly Have The Answer.

Martingale System

The Martingale System is often associated with Roulette but can be applied to any sort of betting. It is essentially doubling up after every losing bet. If the odds are evens or over then in theory the punter cannot lose as they will receive all of their losses back every time they back a winner. However, in this theory the punter has an unlimited betting bank to cover each bet. In reality betting banks can run out and maximum stakes that bookies will allow can be exceeded so the Martingale system is not recommended as the stakes can soon get out of control.


You bet £1 and lose So you bet £2 but that loses So you bet £4 and that loses too You bet £8 and that wins. You make £8 profit on your final winning bet which after the previous 3 losses of 1,2 & 4 = £7, leaves you with £1 profit

Other Staking Systems

Some punters will consistently bet the same percentage of their bank. For example, every bet placed will be 5% of the bank. This means that during bad runs the stakes will be reduced and good runs the stakes will be increased. The problem with this however is that it doesn't take into account how confident the punter is about the bet. It can often be better to change this system slightly to take into account confidence. Three levels of confidence can be assigned, with 7.5% of the bank being a confident bet, 5% being an average bet and 2.5% being a not so confident bet. The exact percentages used are of course up to the punter.

Micko70 has shared a staking plan with followers of his blog and his Stake Back Plan is worth a test if you want to try a new staking system.

Sensible Betting

The key to managing money is to bet sensibly. The main rule every punter should have is to only ever bet what they can afford to lose. Betting should be viewed as fun more than anything else and when it is no longer fun it is probably time to give it up.

Betting an entire bank is never a good idea either. However confident you may be about an outcome, freak results do happen and it can leave you with nothing left with which to try and claw your losses back.

Chasing Losses

The biggest mistake someone can make is to chase their losses. Chasing losses is when someone tries to win back what they have lost in a day or week by either upping their stakes or betting on selections they wouldn't normally gamble on.

The best way to avoid chasing is to set out exactly how much you are going to bet on your fancied selections at the start of a day and preferably what prices you expect (so that you don't over bet to compensate a shorter price than you were expecting). Stick to this rigidly and there shouldn't be any problems.

If you find that you are one of those people that just can't help chasing your losses then place your bets in the morning and don't check the results until that night when all the racing or football matches have finished and there is nothing else to bet on. If you find you have lost money then there is nothing left to chase on.

Chance Versus Odds

A good way of determining how confident you are about a bet and therefore how much you should stake on it is to consider how many times you think that outcome would occur if the event was run or played ten times. The more times it should happen the better the bet is.

Then you have to consider value so consider how many times you think the event will occur and compare it to the odds available. Something that you would expect to happen five times out of ten but is only a 4/5 shot is not as good a bet as something you would only expect to happen one time out of ten but is a 20/1 shot. The better the chances seem compared to the odds available, the more units should be assigned to the bet.


You think a horse would win half of the time if the same race was run ten times.

FAIR ODDS: If the horse is priced at even money. 5 wins @ Evens = 5 profit, but 5 losses =-5 => break even.

BAD ODDS: If the horse is 1/2 5 wins at 1/2 = 2.5 profit, but 5 losses =-5 => -2.5 loss of our 10 staked.

GOOD ODDS: If it was 3/1 5 wins at 3/1 = 15 profit, but 5 losses =-5 => 10 profit from 10 staked.

Know someone who would like this?

Share it with your friends!

Comments from the Forum

  • monkeytennis

    What is the best approach to betting? Sticking to staking plans? Solely staking more on perceived better value bets and less on perceived lesser value bets on a mix of both?

    Staking Plans

    Think this Betting School lesson needs anything else added to it? Disagree with anything written above? Let us know your thoughts and your insight could be included in the OLBG Betting School!
  • nors

    I was giving this some thought today, lets presume many OLBG members can find winning bets fairly regularly, i think we can all find winners on whatever sport.

    Having a staking plan for your betting should surely be a must. We all know days when have 20 units on a loser and 2 units on a 5/1 winner and end the day down, even though we found a 5/1 winner!! A staking plan would solve many of these issues.

    The Shark wrote his staking plan

    The plan that works best for me and my betting style is the take out plan but with a bit of flexibility. So for a standard bet I stake to win 5% of my betting bank. However I allow myself very rare bets to win 10% - maybe only 4-5 a year, perhaps something I think is great value and can't believe what a good price it is (Djakadam at 10/1 for the Gold cup for example). Also rare bets to win 2.5%, again not very often - I don't like to go this low as I believe a winner when you bet as infrequently as I do should have a decent impact on your betting bank.

    The reason this plan works particularly well for me is it can take the impact of lots of losers at bigger prices and discourages bets at shorter odds. I will rarely bet shorter than around 5/2 unless I am really convinced its a great value bet as a loser at shorter odds takes out a lot of my bank - I prefer longer odds losers!

    We have a busy weekend of sport and it would be interesting to see members staking plans and see if you would have made a profit had you bet with a staking plan as opposed to our sometimes scattergun approach.
  • nors

    You can bet a percentage of your bank and as your bank increases your stake goes up. So for example 5% of a 100 bank would see bets of 5 and once you had increased your bank to 200 your stake would be 10 but still 5%.

    it does take discipline something many of us need to get to grips with.
  • davidg3907

    This little post turned into something a bit longer. I shall convert to a more detailed blog when time permits. You have been warned. <!-- s:D --><img src="{SMILIES_PATH}/icon_biggrin.gif" alt=":D" title="Very Happy" /><!-- s:D -->

    Skip to the bottom if simply wanting an insight into one of my current methods.

    To many people, a staking plan is a means by which a losing system can be turned into a potentially winning one. While it is possible for a plan to do this, the time would be better spent trying to find a better system in the first place!

    All staking plans can be shown to make a profit given infinite resources and at least one winner, but selecting the correct staking plan for any method of selection, whether systematic or not, is as important as the staking plan itself.

    Staking plans designed purely to convert any significant level stake loss into a profit, must be avoided. It is possible to turn around a system that makes a slight loss, but the strike rate should be both good and consistent, essentially allowing control of the stakes to be maintained.

    I always feel that Strike Rate should be paramount in deciding whether a staking plan should be employed at all, and only then, which one.

    The primary objects of any staking plan must be to safeguard the bank, and to make a winning system more efficient. That doesn't mean some losing systems can't be turned around by a spot of tinkering with the stakes, but using plans on losing systems with long losing runs will inevitably fast track the journey to the workhouse.

    Staking methods such as Kelly, for instance, are out of commission for most, as they rely on an accurate assessment of value (don't worry, I'm not going down that route here).

    Any reader of my posts will know how I feel about chasing losses, so you may be surprised to learn that in certain circumstances, I do use staking plans - even loss retrieval ones.

    Staking plans that I currently use involve laying to fixed stake, laying to fixed liability, Fibonacci, d'Alembert, Retirement, % of the bank, and of course backing to level stakes.

    Getting laying out of the way first, if using a wide range of odds, I tend to use the two methods in conjunction to lessen the fluctuations in the bank that either can cause individually. One method can have adverse swings when short odds winners are laid, the other has adverse swings when the winners are at longer odds.

    d'Alembert was a system I employed playing Blackjack, without realising that it had a name, so basically I 'reinvented' it! Many feel that counting cards and varying stakes accordingly is easy, and a sure way to win. It is a sure way to get banned, but few are good enough, or have the concentration required to do it properly. Playing accurately and in a way to win the greatest number of hands can be achieved slightly more easily.

    Any form of % of bank staking has one 'flaw' in the eyes of many. It will have the biggest stakes on losers 'by design'. However, with a winning system at the helm, this will continue to increase the bank with most selection processes that return a good strike rate along with a level stake profit. I currently employ this for place only bets. Taking profits at various stages is advised as this will safeguard against a bad run and avoid stakes getting out of hand if you are uncomfortable with higher stakes, even when winning. There will always be room for the 'I don't want to give it them back' approach.

    The Retirement staking plan is too complicated for me to explain here. It is a 'slow burner' but should turn a profitable system with a good strike rate into a better one. I use this to small stakes myself, but as I just let a bot get on with it, using a % of a notional bank that includes Betfair exposure, it will be some time before I can accurately assess the performance.

    A spreadsheet monitors results for the same selections using Fibonacci, d'Alembert and % of bank - along with an incorrect form of Retirement, which doesn't help! <!-- s:lol: --><img src="{SMILIES_PATH}/lol.gif" alt=":lol:" title="Laughing" /><!-- s:lol: -->

    As mentioned, Fibonacci is monitored, but no longer used for these selections. I am now backing at slightly higher odds, so with the risk of a longer losing sequence on the horizon and the winner arriving at shorter odds, that is the recipe for a potential disaster.

    I do still use Fibonacci when laying short priced selections, but there are fairly strict rules and initial stakes by default should be that small so as not to be inconvenienced by laying a string of winners. Only a relatively small profit can be made (or even attempted) of course, but a relatively small profit is better than the relatively small loss that would probably be the case at level stakes. I doubt that the level stakes loss when laying, could be converted to any sort of profit if backing the same selections, due to the dangers of a disastrous losing run.

    The simplest staking system, and of course the most accurate way to judge any method, is that of level stakes. It hardly needs any explanation, but if anyone doesn't understand the importance of a profit at level stakes, they should not contemplate a staking system.

    Returning to the % of bank methods, it should be clear that using this on a losing method will actually increase the losses due to having higher stakes on the losers - unless one can somehow organise the winners to be at much greater odds that the losers.

    Heavy loss retrieval systems must be non-starters in all but expert hands - and in general, most experts will not contemplate using them. That rather ends the discussion on that aspect of staking.

    I currently use an increased staking method on a system of backing certain 'favourites'. This is a 'daily slightly modified d'Alembert', whereby stakes are set for the day and amended for the following day if indicated.

    I say favourites, but because it is impossible to back unnamed favs at SP on Betfair, I must let my bot select the market leader a few seconds before the scheduled start time. I therefore occasionally end up backing the second or even third best in some close markets, or when the race is late starting. This should of course not make a significant difference over a long period due to the swings and roundabouts factor.

    This is still in testing mode, but results have been very satisfactory to date.

    Using a bot, it is easy to play around with stakes much smaller than Betfair minimums, so testing and monitoring systems is very easy and potential losses are minimal. I prefer this to simple 'paper trading' as there is a definitive record of actual results.

    I have been following these methods for 137 actual racing days. I am not sure how many single bets that entails, but somewhere in the region of 1200-1500. That is an extensive test and I shall endeavour to find that exact information at some point.

    There have been ups and downs as with any method, but both the LSP and staking plan keep hitting new highs so consistency is to a large extent likely, if far from guaranteed.

    To level stakes, this is showing a profit of 127.45 points after 5% commission, something I still find difficult to comprehend - and the only reason I have not increased the base stake on the system.

    I have been using a base stake with increases after each losing day, and generally dropping by the same amount after a winning day. Small wins or losses are ignored, and should the increased stakes lead to such a profit that warrants a greater decrease in stakes, that is modified accordingly. This prevents stakes getting too high, even when winning.

    The profit to a 1 base with 25p increments is 209.30 ( compared to 127.45 to level stakes), or double that for 2 and 50p increments. Using a 5 base with 1 increments shows a profit of 911.26, better than level stakes but inferior to the 2 version in performance.

    With all increased staking plans, keeping stakes under control is essential if the bank is not to be put in jeopardy. In testing the above method, 2.75, 5.50 and 14 have been the respective top bets so far - all less than 3 times the basic stake. That is comforting in itself, but I remain in disbelief to some degree as I would have been skeptical if someone had told me that making money from backing unnamed favourites could be so easy.

    I shall update progress on this method via a betting blog after the completion of 150 racing days.
  • nors

    Slightly off topic but have the bot selections that end up NOT being favourite and therefore a bigger price played a noticeable significant part in the profits? I suppose not if the bot chooses the unnamed fav and takes the price a few mins ahead of the off.
  • davidg3907

    The bot is actually set at 10 seconds before the scheduled off - I daren't push it any closer.

    Over a period, I would expect the effect to be virtually nothing, due to the relatively small sample of affected races.

    At times, two people following the same system could find themselves backing different horses if the timer was set differently. It is even possible that liquidity could match one person's bet but not everyone's even if the timer was the same. There is just no way to overcome such things. Even backing the unnamed favourite with a bookie has problems when you have joint favourites. The advantage with Betfair in such cases is that you will only be backing one horse.

    I keep an eye on them whenever convenient, and get occasional bonuses as well as occasional disappointments, seemingly in similar amounts.

    Occasionally, the same thing happens with the odds range. One may be inside the range and get backed before the BSP is outside the range - or vice versa.

    Backing a 10/1 chance instead of a 2/1 chance, or the other way round could make a lot of difference, but where small fractions are concerned, it will matter less.

    While the general trend holds, and it has been amazingly consistent, I shall be happy.

    The only really poor run coincided with the bad weather and very bad ground, which of course can adversely affect many aspects of systemic punting, particularly favourites.

Take A Poll

  • Do you have a betting bank separate from your other monies?
  • Yes

Take An Exam

You need to login or register below to take the Exams