I actually agree with a lot what you've stated Shaun. But the Kelly Criterion is not the invention of either William T. Ziemba & Donald B. Hausch, being the creation of a John Kelly who was working on a totally different product which had nothing to do with gambling.

He was working for AT&T's Bell Labs in 1956. His original formulas dealt with long-distance telephone transmission signal noise. But the gambling community (think it was Edward Thorp who was the first to write about it in a gambling sense) quickly understood that the same approach may help them to calculate the optimal amount to bet on a horseracing and other gambling options and the best way to take advantage of overlays and underlays, maximizing the growth of your bankroll over the long term.

The Kelly Criterion is a recognized by many to be the best money management system and is often referred to when the question of optimal betting size is being discussed and is mentioned in many non-gambling books on money management such as stock exchange investments.

The question of the emotion that would be required in operating the Kelly is of a major consideration, and is why many use less aggressive alternatives based on the Kelly principles, but that particular aspect is NOT what this topic is about, more that it is MATHEMATICALLY possible to enhance level stake profit through the use of a staking plan such as the Kelly.

More on Edward Thorp and the Kelly Criterion can be found at:

http://www.bjmath.com/bjmath/thorp/paper.htm

La Mer,

The Kelly method of staking flogged itself to death in this forum long ago [R.I.P] and I'm surprised you have resurrected it as the 'shining banner' of successful progression staking suitable for punting.

The Kelly's use is almost entirely restricted to the Stock Market and I bet you don't use it for punting [?].

I fully agree with Shaun's sentiments here regarding the Kelly [or even the 1/2 Kelly], which is a theoretical road that punters would never travel on. It is just far too aggressive and makes unrealistic demands and assumptions of the punter. Price determination errors would become very costly as overlays and underlays don't have large neon signs on them. Practical and theoretical punting are two different worlds and by resurrecting the Kelly I think it is you perhaps holding the mischievous mirrors and creating the smoke.

The thread I supplied has not been challenged on the maths forum site it came from and the fact you [nor I for that matter] don't understand the maths involved, doesn't make it an invalid equation, nor either does the Earth remain flat until it is mathematically proven round in simple maths we all understand.

Next to disproving the maths equation in the link, perhaps you would like to mathematically prove [simple maths will do] that progressive staking is profitable long term in actual punting [not theory] where overlays and underlays are never more than subjective opinion on shifting sand, or at least provide a link to a maths equation that does prove this ?

Crash,

Howya going stranger, have been following this thread with interest, and have to agree with you, if you can't make it flat stakes then give it away.

The only progression is the % of your bank and even in that don't increase your bet from whatever % you are working to until your bank doubles THEN increase your bet by 50%.

I had 1 sojurn into progression betting and that was in the mid 80's.

I was backing 2 horses a race with a $4000 bank. My first bet was $2 per horse and the bet increased by approx 50% each race.

I used to run meeting to meeting as if 1 meeting, in other words if I finished the day on the 8th bet the next meeting I would start at the 9th bet.

I NEVER went bust, with most days running fine but then that dreaded run of outs would come and I was very glad I took a spare pair of undies with me.

I always held my ground and came out of it, but I ask you how many punters would start with a $2 bet ($4) with a $4000 bank and hang in when you were on your second last bet, not many.

Also very noticable you've mellowed in your posts.

Keep Well

If anyone out there is interested in this method let me know and I will post it.

It still works today, you won't win a fortune but it is very consistent, BUT you have to bet the % of bank I did and hang in on the underpants changing days.

Cheers

How would you ever know that Crash? And even if that was the case then doesn't it disprove that what you claimed originally (via your link). If people playing the stock market can see the value of Kelly and noted mathematicians support the Kelly theory then there must be value in what he disclosed back in the mid-50's.

You mentioned some unnamed maths forum, I bet there was not one person on that site that could disprove the Kelly theory with actual prove.

Can you prove what you claim with actual proof? I somehow doubt it.

Hi Kenchar,

Good to see another old dog still kicking and yes this old dog has learnt some new tricks that have tempered the mut in me [a bit anyway]. Approaching old age has mostly done it though !!


La Mer,

I think I understand what you are saying, and you correctly describe the Kelly progression as a 'theory' that can [if conditions are favourable] out perform flat stakes over finite periods.

I'm not disagreeing with you that it can work, but simply 'working' however is no argument that long term it will 'out perform' the returns flat stake % betting will return. If a progression can't, then it has no advantage or indeed relevance, whatsoever. The fact that the Kelly is used by some investors in the Stock Market, only proves that the progression myth is alive and well among those gamblers too, nothing else.

The equation's maths point was not trying to prove that a progression can't return a profit, or in favourable conditions out-perform flat stakes staking during a finite period, it was proving that no progression [and that's all the Kelly is] can 'out perform' or produce greater profit than flat stakes betting long term. Certainly not in the real world anyway.

Maths!!
 

LOL I have a good "working" knowledge of maths. Like when I was in Year 12 I got A+ on all my maths exams, however I did it not by using their theories (b/c I hadn't learned them, I spent my year 12 scribbling horse racing mathematical's in my books, working out theories in my head), but by using my mind and working out the problem, then simply writing the answer in the big gap provided. As such, my maths exams looked like this:

Q14 - Blah blah blah x to the y blah blah
















82.6














Q 15 - Blah blah more blah


SO - I laughed when I followed the link from Crash, b/c I had no earthly idea what was going on, however after reading half a page I came to the comment "Now it is a little easier to see what is going on."

!!!!!!

Thanks for the laugh. That was great. I have no idea if he proved anything on there, but gee it was great being completely mystified, then seeing a comment like "Now it is a little easier to see what is going on." LOL

Duritz.

Duritz,
your reaction to that line was the same as mine :-)


La Mer,
Last night I did a bit of research about the Kelly and I think we have both mistakenly been arguing about a progression method that just is not a progression method at all, as for a start it is correctly named the 'Kelly Method' and not [for good reason] the 'Kelly Progression Method'. Unless you used it to suck me in [ good one:-) ] we both have misinterpreted it's methodology.

The Kelly adjusts staking according to actual odds advantage/disadvantage to maximise profit or minimise potential loss. It is a fractional staking adjustment method [ more like adjusting our bet size in flat stakes % betting, but just more complex and related to odds as well as Bank size], not a progression method as we use the term in horse racing or other types of betting. The flaw that would send anyone broke with this method in racing is that unlike Casino games like blackjack and roulette etc., the probability factor [odds] of a horse winning or loosing can never be more than an educated guess. The Kelly method relies on accurate probability that is impossible to calculate in Horse Racing.

There are several books out about using the Kelly Method in the Stock Market, so I assume it is used there obviously.

I thought the maths equation link had a link to the forum it came from but it doesn't. However, unwittingly you already have it and have supplied it in one of your posts here. The equation itself is written by the site owner :-)