Kelly Criterion; Understood - Arbitrage; Understood - Application; WHAT!?

[Optional]

Hey Phil

If you have not been saying parlaying is a better alternative to Kelly and want to act like I am some sort of senile lunatic for thinking as much, can you explain this statement??

there is an even better method to "maximize potential profits" than kelly

[phil_abuster]

I've been wondering for the last page and a half how this relates to the thread topic.
I don't get where you are coming from saying "dont use Kelly" but play parlays.

quote and end quote

[Optional]

Also Phil, if you need to add big fonts, different fonts, colors, underlines, italics and other garbage to your explanations it merely means that you don't think your words can stand on their own. And you're correct in thinking that.

Parlays can be good in spots but are not comparable to Kelly as a staking strategy.

You have to write essays to try and explain your position and restate it over and over the same way when it is questioned as you don't really understand what you are saying. You are just selectively working backwards to try to prove a flawed theory.

And honestly. now you are resorting to total bullshit and contradicting yourself. Sometimes the intelligent position is to admit you may be wrong, go away and think about it, and stop digging

[Optional]

quote and end quote

Ok, take a point big guy.

Now explain this statement, for the third time of asking!

"there is an even better method to "maximize potential profits" than kelly"

[phil_abuster]

Hey Phil

If you have not been saying parlaying is a better alternative to Kelly and want to act like I am some sort of senile lunatic for thinking as much, can you explain this statement??

some reader might well think u r "some sort of lunatic" for falsely claiming that somewhere phil is alleged to have stated "dont use kelly"

I repeat for the third time: I did not say "dont use kelly" nor would i.
in fact I made a supportive comment FOR KELLY!!!!!!!
your lack of reading ability in this particular thread is causing unnecessary posts/responses.

[phil_abuster]

Also Phil, if you need to add big fonts, different fonts, colors, underlines, italics and other garbage to your explanations it merely means that you don't think your words can stand on their own.

no, it means i understand that people like, um, YOU, r likely to miss or misread key points in my discussion and attention needs to be drawn to it to prevent that from occuring.
but u r correct in thinking im wrong about that, as nevertheless, despite my efforts, YOU still missed it.

Parlays can be good in spots but are not comparable to Kelly as a staking strategy.

notwithstanding risk of ruin.
nevertheless 3 team parlays r better than flat betting and THAT was my point.

And honestly. now you are resorting to total bullshit and contradicting yourself.

another false statement.

Sometimes the intelligent position is to admit you may be wrong, go away and think about it, and stop digging

then do that if u choose

meanwhile i stand by my statement that for the consistent winning capper 3 team parlays are a better value than flat-betting and the math proves it conclusively.
no proof has been offered to the contrary.
period.

this is beating a dead horse now and im done with it

[ace7550]

I'm not a moderator. but let me play moderator.
First of all let me state: I don't care who is right, I just want the best betting strategy possible.
Phil, You are saying that in certain unique situations a 3-team parlay makes mathematical sense. Correct?
Optional, Are parlays a bad bet? Does Phil's logic make sense to you assuming we are still using the kelly principal?

[Optional]

I'm not a moderator. but let me play moderator.
First of all let me state: I don't care who is right, I just want the best betting strategy possible.
Phil, You are saying that in certain unique situations a 3-team parlay makes mathematical sense. Correct?
Optional, Are parlays a bad bet? Does Phil's logic make sense to you assuming we are still using the kelly principal?

Yeah the basics of what he says is correct. A parlay with 3 legs where the true odds are better than the offered odds is a mathematically good option.

In real life it may take you 20 years to find enough bets to meet his criteria before you catch up to the actual growth of your bankroll after 1 year using single bets though. It's kind of a dumb 'technical' argument he is trying to make that isn;t useful in the real world.

But what parlaying has to do with the Kelly Criterion is still lost on me anyway. One is a betting strategy and the other is a staking plan.

[Optional]

some reader might well think u r "some sort of lunatic" for falsely claiming that somewhere phil is alleged to have stated "dont use kelly"

Now you are just being obtuse. Actually you have been at every objection.

How about I try for a fourth time and state it this way instead then.

You said "there is an even better method to "maximize potential profits" than kelly"... I sincerely apologize if I stupidly interpreted that to mean to you thought there was a better option than Kelly and ergo, you might be suggesting to use it instead of kelly.

But maybe I'm just a simpleton, do you think you could explain what you did mean then?

[phil_abuster]

Ok, take a point big guy.

Now explain this statement, for the third time of asking!
"there is an even better method to "maximize potential profits" than kelly"

Thought:
i claim that a tuna sandwich has better overall nutritional value than a hamburger.

oddly, there r argumentative people (perhaps in this forum) who might read that statement and "infer" that phil is stating "DONT EAT HAMBURGERS!"

and i really hate when people do that BS.

so, heres the deal:
ill fully answer your Q above, with detailed info, (you know the detailed type of info/analysis which U have not yet shown), only AFTER you first provide proof that i stated "dont use kelly" IN THOSE WORDS (and leaving the BS inferences out of it!)
you do that and i answer with my detailed analysis
you dont, then i wont.
simple.

readers: i DO NOT intend anyone here any disrespect.
its fortunate im on my days off, because ive spent about 8 hours just on this thread the past 24.
the math is there and its correct. yes, as one stated it may be "marginal." according to him. but nevertheless it is there. bottom line.
given that the math was PROVEN, rather than merely criticizing AND being argumentative, all anyone had to say was something like "Thanks for all the effort u put into that 3 hr essay phil. I'll look into that" or "Interesting thoughts, phil! Good luck w/your picks this week" and be done with it.
this experience would lead anyone to think that trying to help, trying to contribute to the forum, trying to offer something more than the usual one liners like "How 'bout dem cowboys" IS JUST NOT WORTH THE EFFORT.

anyway to all the polite cappers out there, i sincerely DO wish u good luck w/your pics this week
goodbye everyone, and good luck

phil

[tsty]

You claimed that you should bet the same amount on a 3 leg parlay as you do on a single leg bet.

You have been trying to argue that you should be betting more on your parlays than kelly recommends.

[trytrytry]

phil these guys are trying to help you, stop, slow down, forget what you think you know is true and start over. trust me

[TheMoneyShot]

Phil_Abuster

I'm not trying to bust your balls here at all. Trying to read this with an open mind. I have one question....

What makes you think you can win 3 team parlays at 16.64% (+600 odds?) To even give you this edge?

Everyone thinks they can hit em. Realistically it's not that easy.

[Waterstpub87]

Phil_Abuster

I'm not trying to bust your balls here at all. Trying to read this with an open mind. I have one question....

What makes you think you can win 3 team parlays at 16.64% (+600 odds?) To even give you this edge?

Everyone thinks they can hit em. Realistically it's not that easy.

Key to the thread. Everyone wants to always argue about optimal bet size and such, like they definitely will hit 55% forever. Hit the 55% first for 10,000 plays, then come back.

[TheMoneyShot]

Key to the thread. Everyone wants to always argue about optimal bet size and such, like they definitely will hit 55% forever. Hit the 55% first for 10,000 plays, then come back.

That is true. It is very hard to do. And I know I couldn't.

But by the same token... I'd like to see someone hit 17 out of a 100 3 team parlays at +600.

[ace7550]

I think on paper the math is correct but when you put it in practice it's not as viable. Basically, the higher your win % the more you would want to parlay. If you won 100% of your bets than obviously you would want to parlay every time.
If you win 55% (I agree that no one actually does this over the long run) than I think it does make sense to 3-team parlay if the odds are all good on one site.
Like Waterstpub said, hit 55% over 10,000 plays. Good F'n luck! You're gonna need it...

[HeeeHAWWWW]

If you win 55% (I agree that no one actually does this over the long run) than I think it does make sense to 3-team parlay if the odds are all good on one site.
Like Waterstpub said, hit 55% over 10,000 plays. Good F'n luck! You're gonna need it...

Just for perspective, "even" 54% at -105, 10k bets at conservative (1/3 Kelly) staking, and you'll turn $100 into $550k median.

Most serious professionals are looking at considerably smaller edges than that, but with bigger starting bankrolls.

[tsty]

I think on paper the math is correct but when you put it in practice it's not as viable. Basically, the higher your win % the more you would want to parlay. If you won 100% of your bets than obviously you would want to parlay every time.
If you win 55% (I agree that no one actually does this over the long run) than I think it does make sense to 3-team parlay if the odds are all good on one site.
Like Waterstpub said, hit 55% over 10,000 plays. Good F'n luck! You're gonna need it...

This doesn't make any sense

It's either viable or not

If it's viable through the math then it should be viable

Also it's not hard hitting 55% over 10000 players if you only play 55% bets for this scenario...

[ace7550]

Right, Of course if you only bet on favorites you are going to win most of your bets. I think Waters is saying 55% at even odds.
The reason why parlays get more and more attractive the higher your win % is because you win a lot more money and the higher your win % the higher your chance of hitting them.
Let's say you win 75% of your bets. You are going to hit a 3-team parlay pretty often. But if you win 50% that 3-team parlay rarely hits. Anyone know how to calculate how often you will hit a 3-team parlay if your win % is 75? 50? etc...
This is all hypothetical anyways. No one wins 75% of their bets and I don't think anyone wins 55% either. It's more of a math discussion since it doesn't apply to realistic betting situations.

[Optional]

Right, Of course if you only bet on favorites you are going to win most of your bets. I think Waters is saying 55% at even odds.
The reason why parlays get more and more attractive the higher your win % is because you win a lot more money and the higher your win % the higher your chance of hitting them.
Let's say you win 75% of your bets. You are going to hit a 3-team parlay pretty often. But if you win 50% that 3-team parlay rarely hits. Anyone know how to calculate how often you will hit a 3-team parlay if your win % is 75? 50? etc...
This is all hypothetical anyways. No one wins 75% of their bets and I don't think anyone wins 55% either. It's more of a math discussion since it doesn't apply to realistic betting situations.

You just multiply the probabiliites.

55% we have been talking about is .55 x .55 x .55 = 16.63% probability of cashing.

75% is .75 x .75 x .75 = 42.19%

[ace7550]

That is so simple! Always helpful Opti

[tsty]

you missed my point and the threads

[HeeeHAWWWW]

Oh, missed most of this thread. With regard to parlays, the problem is that although you do multiply your edge, you give most of that straight back due to lower volume. More importantly: variance goes nuts, and the net effect is much lower EG.

For example, 600 bets, half-Kelly, 5% edge at evens = median bankroll growth 76%. Now with trebles, 200 bets, but at an edge of 15.7625%, and betting at 8.0 (+700) = median bankroll growth 29%.

[Martinr]

The Shawshank Redemption has a lot to answer for (#obtuse)

Phil I've done the math. You are correct that odds of +600 on a 3 team parlay > odds of +595 you'll get by going all-in on 3 separate singles at -110.

But if you shop around for better odds on each single you can do better than +600 for a 3 teamer (for instance -105 * 3 = +644!!)


Assuming a SR of 55% on each leg (so a SR 16.64% on each 3-teamer), Kelly says you should bet:
2.74% on the +600 3 leg parlay
2.62% on the all-in single @-110 *3 (combined odds of +595)
3.69% on the all-in single @-105 *3 (combined odds of +644)

each scenario has different risk of ruin:expected growth ratios and if you are uncertain of your edge you MUST bet fractional Kelly.

[Martinr]

yes.
but.....as i mentioned, if i begin a 100 bet sample, with assumed 55% win expectation on each and every one of those 100 bets, but say i get off to an 0-10 start using the kelly criterion my bet sizes are now significantly smaller and i still only have a 55% chance of winning all succeeding bets - though now at a lower wager sum.
by comparison with flat betting of the same initial size i would come out behind.
otoh, if betting kelly i get off to a rousing 10-0 start i would come out ahead of the flat bettor because those winners have effectively elevated my wager size (as my br increases) and as a result of betting LARGER it follows i would make more $$ even though the odds of winning any succeeding bet remains at only 55%.
bet more, win more
bet less, win less.
thus the START could determine whether or not i continue by either betting more or betting less on all the remaining 55% matches.
or at least this is my understanding of the kelly principle.
i could be wrong.
open to learning improved methods *IF* not complicated

A 0-10 start using Kelly staking shouldn't really make a difference. It's just a downswing that is eventually overcome.
Using the mythical 55% SR and betting full Kelly (about 5% *BR each bet) a 0-10 start would reduce a 100 point bankroll to approx 59 points. If that is unacceptable then Kelly *0.20 (the amount that someone here with experience has already mentioned), would reduce the BR to somewhere around 82 points. At a SR of 55% it wouldn't take long to get out of that slump.
The probability of going 0-10 is about 0.03%.
On the other hand the probability of going 10-0 is about 8* more likely.
When you look at it like that you realise that a SR of 55% on 50-50 propositions is fantasy land for most everyone bar the totally dedicated few, and the delusional many.

[Hareeba!]

When you look at it like that you realise that a SR of 55% on 50-50 propositions is fantasy land for most everyone bar the totally dedicated few, and the delusional many.

To be fair, he did say that he was talking about genuine 55% propositions (55-45), not 50-50 propositions.

[Optional]

To be fair, he did say that he was talking about genuine 55% propositions (55-45), not 50-50 propositions.

I thought he was talking about hitting 50/50 sides at a 55% clip.

[Hareeba!]

I thought he was talking about hitting 50/50 sides at a 55% clip.

Isn't that the same thing?
The market has them at 50/50 but he's a smart handicapper who rates them accurately at 55%

[Optional]

I don't get what your to be fair comment meant in that case.

[Hareeba!]

I don't get what your to be fair comment meant in that case.

Just that I interpreted Martin's comment to assume that they were only 50-50 chances not genuine 55% chances as was the basis of Phil's whole discussion.

[Miz]

No time to read all this mess... I am a winning player over multiple years and thousands of bets...
Couple things... calculating your edge is an educated guess. You'd better build in some padding or use 1/2 kelly etc.

Parlays are a complete pain in the ass and don't really boost your growth that much even when bet optimally, ... so even if they are +EV... I don't bother with them. It is hard enough to get the straight bets in while decent +EV spots exist.
The variance is maddening on anything with a low expected win pct.

Hopefully this has been mentioned ... but I believe the SBR kelly calculator provides the optimal amount to bet on individual games and parlays by hitting the "ALL" option.

[Buffalo Nickle]

Did not know there were enough people on SBR to carry on this conversation for more than a half dozen posts. This should be in the drunktank.

[Buffalo Nickle]

For the original poster, Kelly betting can be considered to be equal to suicide if used improperly which will be almost all of the time.

[Optional]

Hopefully this has been mentioned ... but I believe the SBR kelly calculator provides the optimal amount to bet on individual games and parlays by hitting the "ALL" option.

Good point. https://www.sportsbookreview.com/pic...ly-calculator/

[Martinr]

Just that I interpreted Martin's comment to assume that they were only 50-50 chances not genuine 55% chances as was the basis of Phil's whole discussion.

I guess that's a fair way to interpret it. If he is hitting at 55% then I suppose the market's assumed 50-50 probability is wrong to start with and they are really 55-45 propositions but I think Opti took it in the context I meant.

Hitting sides and totals (50/50 props) (there I go again) at 55%-
Going 0-10 = 0.03%
Going 10-0 = 0.25%

This was in response to Phil's post on the first page where he was suggesting Kelly was at a disadvantage to Flat if starting 0-10. The conversation then headed in another direction.