JR Miller's record

[wintermute]

What happens if you wager 100 times instead of 50 times?

Results now are:

Kelly - 715 wins, 285 losses, average finishing capital 35130 dollars.

Flat bets - 678 wins, 322 losses of which 298 are wipeouts, average finishing capital 29316 dollars.

The more wagers you make the more Kelly is favoured. This is to be expected with any system based on "compound interest". Compound interest will eventually surpass fixed interest no matter what the advantage fixed interest holds over compound interest in absolute terms.

In an earlier post I addressed the problem of having multiple bets - the Kelly ratio has to be adjusted but it still gives superior results.

I'm not necessarily pushing Kelly - it takes work to make it work properly and it takes guts to put up with the wild variations in your capital. But from a mathematical point of view it can not be surpassed with any other wagering system.

In closing:

I have to wonder whether JR's contribution to this forum is a net positive. He writes well and is informed but is also very arrogant. His point of view it seems is the only point of view and he refuses to address valid points made criticizing his arguments. zippy and I are willing to investigate what he says. It would be nice if he returned the courtesy.

Saying 1000 times that 2 + 2 makes 5 doesn't make it true.

wintermute

[AussieVamp2]

Yeah, zippy, 'mute, good stuff.

[Scott Landau]

Zippy, Congratulations! You have just passed Aussie Vamp for the 5th longest post in the history of BW! Of course, in 1st place is Aussie Vamp, 2nd place Aussie Vamp, 3rd place Aussie Vamp.....4th, 6th,7th,etc. Aussie Vamp...AV, I know you will take that only in the humourous nature that it was intended.

[AussieVamp2]

A general comment

J.R's advice would be more useful in general for most people, I would think.

A bunch of analytical and punting genii such as yourselves can be a bit more sophisticated than that.

I even do multiple bet type levels myself.

For instance, realised last season that was getting over 60% on ice hockey, so that was a certain amount at the time.

Add a new league that I haven't worked out yet sufficiently, that will be a lower level.

So I guess I do that myself, but I think J.R.'s advice is solid as a general case.

[wintermute]

Aussie

Doesn't matter - if you have money you're still in the game. And eventually over enough time, Kelly will prevail.

zippy

Kelly doesn't get wiped unless your capital falls so low that you can't scrounge up enough money for a minimum bet which with a starting capital of 10000 and an advantageuos game is going to be a very rare occurrence.

Flat bettors will get wiped a lot because to compete with Kelly they have to wager a greater and greater proportion of their capital per bet and hence run greater and greater risk of gambler's ruin.

See Epstein - The Theory of Gambling and Statistical Logic.

[zippy]

Thank you mls and AussieVamp2. Scott Landau, I feel like I’m in pretty good company to be compared to AussieVamp2, albeit for verbosity. (Appreciated wintermute’s earlier comparison of me to reno as well. Heady company indeed. I have to be careful there, though. I don’t want to wear reno’s crown and have to put up with manic attacks from the likes of REALITY and noholdsbarred.)

wintermute and AussieVamp2: Thank you for the clarification on the Kelly system (not) going broke. I was thinking in more generic terms of betting strategies that alter the number of units one bets based on the likelihood of winning (like I used in my post about Miller’s playing card experiment). Clearly the Kelly system—-which I’m really not familiar with in any detail—-also varies the size of the units themselves based on how much you have in your ever-changing bankroll. Now I understand why you don’t go broke using it.

Does Miller himself understand this last point of yours, by the way? In his article, he summarizes the results of the playing card experiment by saying: “The Kelly loses, and it loses every time. In fact, using most forms of the Kelly criterion, I would be surprised if after 70 or 80 ‘bets’ you are not—-for all intents and purposes—-broke.” This seems flatly contradicted by wintermute’s analysis.

[Scott Landau]

Thanks for the laugh, Zippy. Brains, writing ability, and a sense of humor. How about we annoint you with a life-sized replica of Reno's crown! Also available at your local 99 cent store.

[sportshobby]

'Mute, as I've said before, there are a lot of interesting and contributing writers here that have egos the size of the Milky Way. I'd like to keep them all, let them post with arrogance as long as it isn't personal attacks, and use my own common sense to filter through the rubbish.

Interesting contribution - I hope it is well taken here.

[AussieVamp2]

Scott,

I would think Reno would get the Bettorsworld James A. Michener Trophy

[AussieVamp2]

ok, 'mute

so Kelly betting is good if you are an unperturbable android who can calculate his general edge highly accurately

I would certainly be guilty of being too conservative in general.

[J. R. Miller]

The man or woman going by the name 'Wintermute' says, "I have to wonder whether JR's contribution to this forum is a net positive."

Mr. (or Ms.) 'Wintermute', I don't think it is. You're right. Even though i was notified that you and others were attacking me here, and so I came only to clarify my position, I'm sorry I came here to defend myself. I am reminded of an old high-school saying: "You can buy 'em books, but they'll just eat the pages."

I've been accused of being evasive. That's a curious accusation when I am the only one here using his real name. Where I come from, grown men don't even use the word "zippy," much less call themselves "Zippy." Why don't you guys have the balls to use your real names? That's about the chicken-****est thing I can think of.

Anyway, I have tried to explain how sports betting "works." Here's the books;- you boys go ahead and eat the pages. I give up.

[AussieVamp2]

'mute

Have you simulated the plateau theory as well - or was it just straight flat bets?

i.e. Go up by whatever factor at a 50% increase, or down - that would be an interesting intermediate comparison if you had not done so.

[AussieVamp2]

and another wrinkle, this all might be very well for pointspreads, etc., but how about situations like this

Betting on different returns
Such as the even money bet where you have a 0.55 chance to win.

Or the 8.00 bet where you have a 0.175 chance to win - both, for example, 0.05 higher in probabilistic terms - but how do you handle them differently saying they are based near different numbers, distributions of sequences will be different, etc.?

[reno]

'Mute, I concur. Zippy is a superb writer. Furthermore, the overall caliber of writing throughout this post thread is very high.

I also appreciate the fact that Zippy emphasized the difference between the perceived contentiousness of Slam Dunk and the childish obnoxiousness of Noholdsbarred. Slam Dunk offers this forum much-appreciated logical insights. Noholdsbarred offers this forum...well, nothing anymore. Jeff, bless his heart, kicked the curmudgeon out of Bettorsworld.

After considering J.R.'s rationale, is there anyone in Bettorsworld who now plans on implementing his money-management system? Is there anyone, other than J. R., who finds fault with Huey Mahl's defense of the Kelly system?

[zippy]

I am only one of many, many Bettorsworld posters—-including some of our most highly regarded members—-who have raised questions about or have expressed disagreement with one or more of J.R. Miller’s views. We have attempted to do so constructively, by making substantive points, not simply by “attacking” him. His comical, blubbery non-response in his final post tells us all we need to know about whether he is capable of civilly and cogently defending those views.

Don’t let the door hit you in the butt on your way out.