G'day Punter!

In this Discussion

Who's Online

0 Members & 27 Non Members

The concept of value and the Kelly formula

West Australian Racing
As many more serious punters would know - the best concept in gambling is getting value.

Value is achieved when you can get better odds about an outcome against the actual odds or more importantly in horse racing the preceived or expected odds.

For example

If you reckon a horse had a 50% chance of winning - it expect price is $2 - and you would be looking to back it (for a long term positive outcome) if you got odds better than that

if your pricing is accurate over a long period of time you will finish in front

however another important aspect (often neglected) is bankroll management

even the best punters will have extended losing runs and if you are betting too much of your bank you can go broke

enter the Kelly Formula (copied straight from wiki)

In probability theory, the Kelly criterion, or Kelly strategy or Kelly formula, or Kelly bet, is a formula used to determine the optimal size of a series of bets. In most gambling scenarios, and some investing scenarios under some simplifying assumptions, the Kelly strategy will do better than any essentially different strategy in the long run. It was described by J. L. Kelly, Jr, in a 1956 issue of the Bell System Technical Journal.[1] Edward O. Thorp demonstrated the practical use of the formula in a 1961 address to the American Mathematical Society[2] and later in his books Beat the Dealer[3] (for gambling) and Beat the Market[4] (with Sheen Kassouf, for investing).

Although the Kelly strategy's promise of doing better than any other strategy seems compelling, some economists have argued strenuously against it, mainly because an individual's specific investing constraints override the desire for optimal growth rate.[5] The conventional alternative is utility theory which says bets should be sized to maximize the expected utility of the outcome (to an individual with logarithmic utility, the Kelly bet maximizes utility, so there is no conflict). Even Kelly supporters usually argue for fractional Kelly (betting a fixed fraction of the amount recommended by Kelly) for a variety of practical reasons, such as wishing to reduce volatility, or protecting against non-deterministic errors in their advantage (edge) calculations.[6]

In recent years, Kelly has become a part of mainstream investment theory[7] and the claim has been made that well-known successful investors including Warren Buffett[8] and Bill Gross[9] use Kelly methods. William Poundstone wrote an extensive popular account of the history of Kelly betting in Fortune's Formula.[5] But as Kelly's original paper demonstrates, the criterion is only valid when the investment or "game" is played many times over, with the same probability of winning or losing each time, and the same payout ratio.[1]

the fomula is

A = 1/W - (1 - 1/W )/(D - 1)
Where:
A = Percentage of the total bank to bet
W = Assessed odds of winning (your calculated odds)
D = Decimal odds (actual odds available)

so for example in a race where you price a horse at $2 and can obtain $2.50

doing the calculations tells you to bet 16.67% of your bank

as previously mentioned in the wiki write-up I also recommend using a fractional kelly because in a high-risk game like horse racing there are many variable factors which are hard to account for - and nobody can be the perfect handicapper which is what kelly assumes

I once read where the late and great Alan Woods who had computer systems including over 1500 variables still found there was a small % of price differential that could not be accounted for and would bet accordingly


now this leads me to many popular misconceptions amongst punters

Odds-on look on
If you do form properly on occasions there will be a situation where you will price horses at odds-on

a horse priced at $1.80 and where $1.90 can be obtained gives a Kelly Bet of 6.17%

the comparative bet for a horsed priced at $10 to get the same figure you will need odds of over $24 (6.09%)

stagerring isnt it :shock: :shock: :shock:

how many punters would quite easily back a horse at $24 they believe should be $10 and how many would let a horse slide at $1.90 when they price it $1.80

Comments

  • thefalconthefalcon    20,495 posts
    To much info...from snake oil salesmen....smart ones but still charlatans...
  • AquanitaAquanita    566 posts
    Very interesting information Della. Over the years I have always been confident in my form analysis in the long term but have not had the same confidence in my betting strategies.

    I always look for value or overlays and have tried a number of staking plans with mixed success. The Kelly method is obviously a tried and true concept and I will do some further research as to how it can be applied in practice.

    I still beleive that if your pricing is accurate enough the percentage bet to return a certain amount ensures you maximise overlays.

    For example if you are betting to collect (not win) $1,000 and a horse is rated at $4 your bet is $250 irrespective of the price available as long as it is in excess of $4. If the horse is $5 your collect is $1250 for a profit of $1,000. If the horse is $8 your bet is still $250 for a return of $2,000 and a profit of $1,750. Most punters tend to scale back their investment when the price is well above their assessment denying themselves of the opportunity to maximise returns when they are correct.

    A very interesting topic and there will be some diverse views on the best staking methods from the highly disciplined to the scatter gun approaches.
  • RodentRodent    7,472 posts
    I use the Denis Denuto approach. I just go with the vibe. It has served me well over the years but my vibe radar was off yesterday and I got poleaxed by Cobble Lane.....it's still hurting.
  • TheDivaTheDiva    13,248 posts
    thats what i love about this game. Carey and Rodent, two blokes who have been around the block a few times:

    one thinks Cobble Lane is a special and good value, even at odds on, the other thinks its a huge lay at the price. Its differing opinions like this that give us all our perception of value in a market.
  • mcginty71nzmcginty71nz    195 posts
    said:

    I use the Denis Denuto approach. I just go with the vibe. It has served me well over the years but my vibe radar was off yesterday and I got poleaxed by Cobble Lane.....it's still hurting.

    :lol: :lol: :lol: :lol: just laughed out loud in a showroom where there are 2 other people. they think i am a little bit weird now. :lol: :lol: the vibe - a classic from rodent.
  • careycarey    6,424 posts
    said:

    As many more serious punters would know - the best concept in gambling is getting value.

    Value is achieved when you can get better odds about an outcome against the actual odds or more importantly in horse racing the preceived or expected odds.

    For example

    If you reckon a horse had a 50% chance of winning - it expect price is $2 - and you would be looking to back it (for a long term positive outcome) if you got odds better than that

    if your pricing is accurate over a long period of time you will finish in front

    however another important aspect (often neglected) is bankroll management

    even the best punters will have extended losing runs and if you are betting too much of your bank you can go broke

    enter the Kelly Formula (copied straight from wiki)

    In probability theory, the Kelly criterion, or Kelly strategy or Kelly formula, or Kelly bet, is a formula used to determine the optimal size of a series of bets. In most gambling scenarios, and some investing scenarios under some simplifying assumptions, the Kelly strategy will do better than any essentially different strategy in the long run. It was described by J. L. Kelly, Jr, in a 1956 issue of the Bell System Technical Journal.[1] Edward O. Thorp demonstrated the practical use of the formula in a 1961 address to the American Mathematical Society[2] and later in his books Beat the Dealer[3] (for gambling) and Beat the Market[4] (with Sheen Kassouf, for investing).

    Although the Kelly strategy's promise of doing better than any other strategy seems compelling, some economists have argued strenuously against it, mainly because an individual's specific investing constraints override the desire for optimal growth rate.[5] The conventional alternative is utility theory which says bets should be sized to maximize the expected utility of the outcome (to an individual with logarithmic utility, the Kelly bet maximizes utility, so there is no conflict). Even Kelly supporters usually argue for fractional Kelly (betting a fixed fraction of the amount recommended by Kelly) for a variety of practical reasons, such as wishing to reduce volatility, or protecting against non-deterministic errors in their advantage (edge) calculations.[6]

    In recent years, Kelly has become a part of mainstream investment theory[7] and the claim has been made that well-known successful investors including Warren Buffett[8] and Bill Gross[9] use Kelly methods. William Poundstone wrote an extensive popular account of the history of Kelly betting in Fortune's Formula.[5] But as Kelly's original paper demonstrates, the criterion is only valid when the investment or "game" is played many times over, with the same probability of winning or losing each time, and the same payout ratio.[1]

    the fomula is

    A = 1/W - (1 - 1/W )/(D - 1)
    Where:
    A = Percentage of the total bank to bet
    W = Assessed odds of winning (your calculated odds)
    D = Decimal odds (actual odds available)

    so for example in a race where you price a horse at $2 and can obtain $2.50

    doing the calculations tells you to bet 16.67% of your bank

    as previously mentioned in the wiki write-up I also recommend using a fractional kelly because in a high-risk game like horse racing there are many variable factors which are hard to account for - and nobody can be the perfect handicapper which is what kelly assumes

    I once read where the late and great Alan Woods who had computer systems including over 1500 variables still found there was a small % of price differential that could not be accounted for and would bet accordingly


    now this leads me to many popular misconceptions amongst punters

    Odds-on look on
    If you do form properly on occasions there will be a situation where you will price horses at odds-on

    a horse priced at $1.80 and where $1.90 can be obtained gives a Kelly Bet of 6.17%

    the comparative bet for a horsed priced at $10 to get the same figure you will need odds of over $24 (6.09%)

    stagerring isnt it :shock: :shock: :shock:

    how many punters would quite easily back a horse at $24 they believe should be $10 and how many would let a horse slide at $1.90 when they price it $1.80

    you have not given credit to whoever wrote that article nor a link to it.
    and if aw had 1500 factors then i have never had a bet in my life.
    now let me see you apply the formula for multiple horses in a race?

    kelly(fractional) may be fine in theory, but it's impractical, for the great majority of punters.
    bookmakers have limits.
    corporates may not let you on.
    pari-mutuel you can't know the final dividend before you have bet.
    if your method is sound then i suppose you should just bet the amount that gives you the greatest profit in dollar terms, but few people are in a position to do that
  • HitmanHitman    146 posts
    said:

    said:

    I use the Denis Denuto approach. I just go with the vibe. It has served me well over the years but my vibe radar was off yesterday and I got poleaxed by Cobble Lane.....it's still hurting.

    :lol: :lol: :lol: :lol: just laughed out loud in a showroom where there are 2 other people. they think i am a little bit weird now. :lol: :lol: the vibe - a classic from rodent.
    :lol: :lol: Brought tears to my eyes that one Rodent :lol: :lol:
    It`s the vibe of the thing.

    Still banging on about value Della :lol:
    Just let it go buddy. Let it go :)
  • chocchoc    789 posts
    Value IS important.
    If you like something strongly enough, take 10/1-on.
    If you dont perceive that value exists at that price (and who would) then dont have a bet in the event.

    You can back 7 winners out of a card of 8 and still lose if you're STAKING is wrong - though you shouldn't.
    Conversely, you can back one winner on a program and win, if you're staking (and acceptance of a price of value) is adequate & accurate.

    There is no precise science when it comes to punting, because there are so many variable and intangibles which come into force. Ultimately, the punter makes his own assessment of the market and enters in where he perceives value to be.

    There's a guy in the weekend paper who writes a stockmarket column, signing off by saying "tread your own path".
    Never a truer word spoken.
  • WyongiWyongi    152 posts
    the fomula is

    A = 1/W - (1 - 1/W )/(D - 1)
    Where:
    A = Percentage of the total bank to bet
    W = Assessed odds of winning (your calculated odds)
    D = Decimal odds (actual odds available)

    so for example in a race where you price a horse at $2 and can obtain $2.50

    doing the calculations tells you to bet 16.67% of your bank


    There is no known way that I would bet 16.67% of my betting bank on one horse, let alone a number of horses in any single race. I have always stuck to a hard and fast rule to limit my total bets on any one race to a maximum of 2% of my betting bank. I would rather make a small profit on any race with little chance of going broke, than making a larger profit with a greater chance of going broke. BTW I have survived successfully on the punt going back to when Borehead won the Caulfield Cup.
  • edited April 2011
    said:

    the fomula is

    A = 1/W - (1 - 1/W )/(D - 1)
    Where:
    A = Percentage of the total bank to bet
    W = Assessed odds of winning (your calculated odds)
    D = Decimal odds (actual odds available)

    so for example in a race where you price a horse at $2 and can obtain $2.50

    doing the calculations tells you to bet 16.67% of your bank


    There is no known way that I would bet 16.67% of my betting bank on one horse, let alone a number of horses in any single race. I have always stuck to a hard and fast rule to limit my total bets on any one race to a maximum of 2% of my betting bank. I would rather make a small profit on any race with little chance of going broke, than making a larger profit with a greater chance of going broke. BTW I have survived successfully on the punt going back to when Borehead won the Caulfield Cup.


    If you had read thro I dont recommend using the full Kelly either

    Like carey has said its hard to use this system accurately in real life but when you get familar with similar numbers often enough you get to know the approx amount to be betting

    I tend to go with a 20% Kelly - so in that example you are looking at a 3% bet - which is more realistic

    the majority of bets will fall way below this mark if youre handicapping is relatively accurate
  • something else I had found over my many years of punting and recording and reviewing the results is that the further away your price is of that available the greater it is that you are wrong - and the accurate price is somewhere in the middle

    so what I have found is that I will bet a less kelly % when the overs appear to be massive

    for example - I really like a horse and price it at $2

    in scenario 1 I can get $2.50 - here I go with my normal 20% system and bet 3% of bank

    now in scenario 2 I can get $5 (in these circumstances I always tend to second guess myself and go thro possible reasons why I may be wrong)

    if there is nothing obviously I will still bet - 20% kelly is about 7.5% but I am more likely to go with a 10% kelly and bet 3.5-4% as a safety measure

    I am still winning a shitload more in scenario 2 if the horse gets home

    it is extremely rare I will bet more than 5% on an individual horse - but I can bet up to 10% on a race if I am very confident on an outcome covering several horses
  • DarkDark    909 posts
    The only problem with the Kelly formula is it assumes you actually know what your advantage is.

    For 99.9% of punters they simply dont. They might THINK they now i.e pricing something at $5 when $6 is on offer, but the fact remains they dont.

    AW and WB actually KNEW what their advantage was (is) hence using 1/4 Kelly.
  • pajonespajones    669 posts
    said:

    the fomula is

    A = 1/W - (1 - 1/W )/(D - 1)
    Where:
    A = Percentage of the total bank to bet
    W = Assessed odds of winning (your calculated odds)
    D = Decimal odds (actual odds available)

    so for example in a race where you price a horse at $2 and can obtain $2.50

    doing the calculations tells you to bet 16.67% of your bank


    There is no known way that I would bet 16.67% of my betting bank on one horse, let alone a number of horses in any single race. I have always stuck to a hard and fast rule to limit my total bets on any one race to a maximum of 2% of my betting bank. I would rather make a small profit on any race with little chance of going broke, than making a larger profit with a greater chance of going broke. BTW I have survived successfully on the punt going back to when Borehead won the Caulfield Cup.

    Borehead sounds like a rather insulting name for a horse Wyongi ( probably better than MY RICHARD though ).
  • The Kelly method def has its place - I have been using it successfully on the Jockeys Challenges for a long time
  • jigajiga    15 posts
    any of you experts
    ever reslise that you dont need to pick the winner to win!!!!

    if your in it to make money,why would you try to pick the winner of a race
    when you can give yourself 3 chances at winning,by lookin for placegetter
    :lol: :lol: :lol:
Sign In or Register to comment.