Reverse Kelly Concepts

I mentioned about the reverse Kelly concept in a recent webinar that I held and there has been a lot of interest in going into more detail on it. I have to say straight up that this wasn’t my invention. I came across the concept a while ago from an American author, David Schwarz (you may have heard me mention him before as I have gained many good ideas from him), and I have played and investigated with it for many years and it holds true and is very powerful.

The theory is based around your advantage and ROI. Using a percentage of your betting bank on every bet will automatically lower these figures and every bettor wants to raise these numbers not lower them!

Take a minute and think about what you are doing when you bet a Kelly or any percentage of your bankroll. You are placing high stakes on the losing bets and lower stakes on the winning bets.

For example you have a £100 bankroll and you bet £10 on the first bet and win £150 and then you bet £15 and lose and then £14 and lose and then £12 and win. You have placed a smaller bet on the winner and bigger bets on the losers.

This scenario means that you can come out as a loser even if you would have made a profit flat betting. What all this adds up to is that to bet a percentage of your bankroll you need to have a greater advantage to profit than if you were flat betting.

Okay so now for the theory. It is really a very simple principle. If it is decreasing our advantage when we follow a Kelly bet or a percentage of the bank bet then if we play the opposite of it, it should increase our advantage.

Think about that for a little bit and let it sink in.

Now unfortunately when I started writing this article I had all sorts of ideas that I wanted to demonstrate but then I realized I couldn’t because this is published information and if I explain much more then I am going to be breaking the copyright law. Sucks I know.

However if there is some serious interest in this money management method then I can get in touch with Dave and see if I can arrange to get the method printed up over here. If you would like to see this made available in the UK then just post a comment below and if we get enough then I shall see what I can do.

Michael Wilding

Michael started the Race Advisor in 2009 to help bettors become long-term profitable. After writing hundreds of articles I started to build software that contained my personal ratings. The Race Advisor has more factors for UK horse racing than any other site, and we pride ourselves on creating tools and strategies that are unique, and allow you to make a long-term profit without the need for tipsters. You can also check out my personal blog or my personal Instagram account.


  1. Hi,
    Very interesting, i like to look at staking plans as it is an important part of my betting once i get my selections, punters often overlook the inportance of the correct staking, so this theory has got me thinking which has to be good!, i look forward & welcome any new staking plans.

  2. I would be interested. Incidently is David Schwartz the same person who is known as the stock market historian and writes in the Times etc?

  3. Hi Michael

    Yes I would like to know more about this method, it would seem that what is being stated makes sense.


  4. Very interested. I am from Australia and this has been always been a concern to me – investing high on losses and low on wins. Bring it on.

  5. I would certainly be interested in seeing the kelly or percentage money management method.

  6. My heads a bit frazzled think of it 🙂

    More complete info on the concept would be good.

    Loss recovery style staking for example you are effectively increasing the % of bankroll used on a bet following a loser or string of losers. The very rough core priciple of that is matched above. The danger there is total blow out at a faster point than would be hit by levels , which in turn is faster than staking which is a proporation of bankroll size. Total blow out is one thing traditional % bank style betting theoretically protects you against. ( in reality of course turning £1000 into £1 is effective blowout. )

    Staking plan optimisation in of those things easier to do if you have a lot of past
    history to examine.

    The choice of shoe must fit the foot etc 🙂

    Further expansion a good idea however.


  7. Correct me if I am wrong, but surely using a % of Bank staking plan is not the same as using Kelly.

    Also the reasons for using a % of bank staking plan rather than flat stakes is two fold:

    1. Over a long series of bets with longer winning sequences than losing sequences, the Bank will grow faster than if using flat stakes at the same initial size. So it is good for many Laying systems – especially when used in a Liability rather than stake manner.

    2. Using a % of Bank staking plan also protects you from being wiped out by an unusually long losing sequence. This is because as you continue to lose, you are progressively making smaller bets. This makes it suitable for Backing Systems where the odds are substantially greater than evens.

    1. You are right using a % is not the same as using a Kelly. You are exactly right with point 1 and 2. The main thing is that due to the structure of using a % you will need a bigger edge to make a continuous profit than you do with flat betting because long-term for back betting you place more winning bets at smaller stakes and more losing bets at bigger stakes. It is possible to change this around so that you see a much higher ROI.

  8. could this be the breakthrough we are all looking for ?
    very interesting – ……..cheers ………Peter

  9. Would like to hear more.
    I compromise using % of bank betting.
    I increase stakes on a winner but hold the latest level of staking when I lose

  10. I shall see what I can do, because the method is copyrighted I don’t know how much I am allowed to give away. I shall investigate though.

  11. Always interested in any new idea and wouldn’t mind knowing (even roughly) how Kelly works – never heard of it. I suspect won’t suit me, but would like to hear more. Thanks


    1. Sheila I have made a note to write an article about the Kelly Criterion. Keep an eye out on the blog for it soon.

  12. I have decided to change my betting from personal project to portfolio based following
    a selections of tipster services so weighting and decreasing stakes is a very important thing to learn so I am keen for you to be able to continue educating me on this subject matter.
    regards Shane

  13. Its important to have a positive expectation of your bets. This then leads into the EDGE you have over the market. Kelly says that if you have 100% edge(Complete Certainty) then you can use 100% of your bank. It therefore follows that with 5% edge you can risk 5% of your bank. Racing being what it is. The half of your edge% stakes errs on the side of safety which , hopefully, allows for your errors during the creation of the odds line.

    To me, the above is a universal truth in betting that I have adopted. Michael, Are you now saying that its a load of bunkum?

    1. Absolutely not, edge and value is everything. However it is impossible to know before a race what your edge is, you can only estimate it, and Kelly staking has high penalties for over/under estimating. It also means that as your bankroll grows your staking more, by default a losing run will follow a winning run so you are always going to be betting more during a losing run than a winning run. It is more about finding a way to counteract those issues with the staking.

  14. Ok so theres always one guy who comes in from left field, would this reverse staking method work with a laying system that has so far in its short run showed a strike rate on lays of almost 83%, I thought it would be fun to be the odd one out lol

  15. I haven’t had major disasters as yet but, yeah, a more scientific approach to staking is at the back of my mind and this would be most welcome…

  16. A lot has been written about Kelly betting since John Kelly’s original paper was published about 60 years ago.
    Mostly by people who never used the method in practice with real money or who have little understanding of the maths or problems associated with real-time betting.

    For me the most significant part of Kelly’s paper was a mere footnote referring to his supervisor, the legendary Claude Shannon. There is no Nobel prize for maths: if there was Shannon would probably have picked up three!

    Kelly does require accurate calculation of “edge” which few punters consistently are able to do. Kelly attempts to maximize not the bankroll size but the LOGARITHM of the bankroll size, if you understand the maths of Kelly you will then know why. Incidentally, this concept was around before Kelly.

    We know from a paper published long ago (maybe Leo Breiman?) that in a closed betting pool using the same ratings a Kelly bettor competing against all the non-Kelly bettors will eventually end up with all the money.

    There are some practical issues with Kelly: infinite divisibility of bets ( can I get set for 0.5 cents or $10 million?), how long do you live? (being told you will have the biggest bankroll in 1000 years time is somewhat academic!), utility of money ie does the loss of say $10000 have the same utility to you as to Bill Gates? eg you may be happy to ascribe the utility value of a litre of water to be $1 normally but stranded in the desert it’s value could jump to $100!
    A lot more could be written, just my $0.02 worth.

  17. Michael, very interested in the concept of reverse Kelly. Is it published in a book? Be great to get more details. Thanks.

Leave a Reply

Your email address will not be published. Required fields are marked *

Back to top button