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

[phil_abuster]

but alas, just plain wrong

says the guy who apparently cant read,
doesnt understand logic/facts,
has trouble w/ very simple 6th grade arithmetic,
and a self-confessed problem expressing himself.

[phil_abuster]

stop betting parlays

easy

AGREED!!!

the overwhelming majority of bettors do NOT have a winning expectation (inspite of their belief to the contrary) and as i stated repeatedly those (average) bettors r doing themselves harm betting parlays!! couldnt agree w/ U more!

[phil_abuster]

Fictitious example - with exaggerations, to illustrate both "risk of ruin" (ROR) and "value"

my apartment building has 6 luxury suites
2 on each floor
as a promotional bonus/reward to us for our leases,
our very wealthy landlord is offering a fun "draw" every week

we can enter a draw on our floor each week for $100 entry
there are only 2 of us competing in the draw so the odds r 50/50
if my neighbor wins the draw he gets his 100 back plus MY 100
plus an additional $10 kicked in by the wealthy landlord. (ie. $110 profit from 100 risked)
sometimes ya win 110, and sometimes ya lose 100.

any 6th grader can agree this is good value, albeit a very small payoff.

OR,

our wealthy landlord invites us to participate in a draw from all 6 residences!
but $1000 entry fee. runs 20 weeks
do NOT have to participate every week
with 6 competing i only have a 1/6 chance of winning any weekly draw
(ie. 16.6%, similar to 3 team parlays)
this is worrisome because of the oft cited "Risk of Ruin"
$1000 every week adds up very quickly, especially if there a long drought between wins!

theres no guarantee i would win for certain during every 6 draws.
i could go 7, 11, or 20 draws before ever winning just once!
the likelyhood of course is once every 6 draws -- but anything can happen!
i may never win, or i may win 3 or 4 times in 20 weeks.
THE RISK OF RUIN IS VERY HIGH since the win rate is only 16.6%!!!
of course i can alleviate most of the risk by participating in only a few draws.

then i discovered what the prize was if/when i win.
$25,000 (each weekly draw for 20 weeks)
so, $1000 entry fee, for a WEEKLY 1 in 6 chance of winning $25,000
even if i only participate occasionally, i still have a 1/6 chance of winning $25,000 from an investment of just $1000.

some might shout "ROR!, ROR!, ROR!!!!"
others might say, "No intelligent person would regularly risk $1000 on a 16.6% chance" of winning
but our wise 6th grader can see the clear "value" (high risk, but exceptionally high payoff)
Good cappers understand "value"

[Hareeba!]

You're just totally missing the point I'm afraid Phil.

[tsty]

at least make the edge same lol

[phil_abuster]

You're just totally missing the point I'm afraid Phil.

actually it is U who is totally missing the point.
i simply cant help anyone who clearly doesnt get simple arithmetic nor the logic of "value"

[phil_abuster]

Now turning to the application of Kelly:

If betting at +600 on 3 team parlays where you rate each leg at 55%, you should be staking 2.7% of your bank.
If betting the teams individually (flat-bet) at -110 you should be staking 5.5% of your bank on each.

That's at "full" Kelly, which of course is generally not regarded as suitable for gambling as the likelihood of total loss is too great. Better to work to no more than one-fifth Kelly.

i will accept that full Kelly has an increased "risk of ruin"
and so a reduced version of Kelly (one-fifth) is recommended, as U state.

and, according to U, given a 55% win expectation, the 3 team parlay should stake no more than HALF the amount compared to flat-betting the same 3 picks (ie. 2.7% BR vs 5.5% BR)

if we accept this premise, then logically,
your "HALF the amount" relationship would hold whether betting full Kelly or your recommended "one-fifth Kelly", as U recommend.
and logically, your "HALF the amount" relationship would also hold when NOT betting any variation of Kelly at all - ie. just adhering to typical flat-betting.

to all my fellow forum members, review PART A in my original essay,
where the simple arithmetic shows clearly a 9.8% greater profit and significantly higher ROI when the consistent WINNING capper (55%) forms 3 team parlays but wagers only ONE-THIRD the amount of flat-betting those same picks!!

well if we follow this suggestion of a "HALF the amount" relationship between parlays and flat betting, then staking $300 on 3 picks would now be compared with a parlay stake of $150 (not just $100) -- and the parlay profit INCREASES even more from +$1,648 more than flat-betting up to +$2,469 - which is 64.6% better profit!

[Alfa1234]

Following your math, there is even more value and potential profit for a winning capper if he bets 8-leg parlays. Or 12-leg parlays. Think about it. It's nice in theory, however you seem to be forgetting the emotional aspect of losing 84+% of your bets. You also forget the fact that the variance greatly increases this way and your potential for bankrupting your roll increases hugely with such a small win rate.

Read up on variance, you'll find that with a 55% win rate a 10-0 loss streak is not even that exceptional and will occur quite often after a few thousand bets. Calculate what can happen (sorry I can't be bothered) if your win rate decreases to 16% or lower and you'll see you should be betting way less than 1% of your bank to keep the bankruptcy risk down to an acceptable level.

[Cookie Monster]

Tsk, tsk. The 5.5% of bankroll is on EACH leg. Too much for a guy arguing Hareeba "apparently cant read".

So, the math goes (assuming a $10000 bankroll):

Three bets of $550 risk to win $500, 55% WP, (+EV = 5%) = $82.5 average profit.
One parlay risking $274.38 to win $1646.25, 16.64% WP (+EV = 16.46%) = $45.17 average profit.

I suppose now it is your turn to bash Kelly betting as nosense :/

i will accept that full Kelly has an increased "risk of ruin"
and so a reduced version of Kelly (one-fifth) is recommended, as U state.

and, according to U, given a 55% win expectation, the 3 team parlay should stake no more than HALF the amount compared to flat-betting the same 3 picks (ie. 2.7% BR vs 5.5% BR)

if we accept this premise, then logically,
your "HALF the amount" relationship would hold whether betting full Kelly or your recommended "one-fifth Kelly", as U recommend.
and logically, your "HALF the amount" relationship would also hold when NOT betting any variation of Kelly at all - ie. just adhering to typical flat-betting.

to all my fellow forum members, review PART A in my original essay,
where the simple arithmetic shows clearly a 9.8% greater profit and significantly higher ROI when the consistent WINNING capper (55%) forms 3 team parlays but wagers only ONE-THIRD the amount of flat-betting those same picks!!

well if we follow this suggestion of a "HALF the amount" relationship between parlays and flat betting, then staking $300 on 3 picks would now be compared with a parlay stake of $150 (not just $100) -- and the parlay profit INCREASES even more from +$1,648 more than flat-betting up to +$2,469 - which is 64.6% better profit!

[phil_abuster]

Following your math, there is even more value and potential profit for a winning capper if he bets 8-leg parlays. Or 12-leg parlays. Think about it.

true, but irrelevent since the practical application of actually forming 8 team or 12 team parlays (each pick w/a solid 55% win expectancy) is both frivolous and implausible.
U seem to be doing the forgetting here my friend, no disrespect intended.
it took 3 hours to compose my essay. and i took great pains to repeat frequently that this applies ONLY to those with a consistent 55% win expectancy on each and every selection within the parlay! just randomly selecting 8 teams to form a parlay is a recipe for disaster. and frankly i doubt very seriously that a bonafide WINNING capper (they being VERY selective in betting) would EVER have 8 high-quality (>55%) picks available - ATS or totals only - at the same moment, at the same sportsbook! which is why i kept my discussion to OCCASIONAL 3 team parlays.

It's nice in theory, however you seem to be forgetting the emotional aspect of losing 84+% of your bets.

no sir, i am not forgetting.
which is why I pointed out very clearly that in regards ONLY to a bonafide WINNING capper with a long history of winning consistently with ATS or totals ONLY - it is implausible that such a person would be able to form 3 team parlays on a regular basis! only occasionally!!!
respectfully, U think about that.

as for the "emotional aspect of losing 84% of ones bets - 3 things which U r missing from my essay covering all this:

1/ my essay is speaking ONLY of consistent WINNING cappers. (i wish people would remember this before posting!)
sure, the average capper, (non-winner! which is 95% of us here and that now includes me this season) may struggle emotionally if he stupidly throws together scores of 3 team parlays weekly (or daily) - despite NOT having a consistent win rate of 55% on ATS/totals picks - when there is an 84% chance of losing over time. agreed.
but i repeat, my essay is NOT addressing the 95% who r average cappers!
it is intended (and oft stated) ONLY for those with a consistent WINNING expectation of 55% with ATS/totals.

2/ i believe the 5% who r consistent WINNING cappers understand the "value" in high risk for VERY HIGH reward (like i illustrated in my fictitious example above about the landlord draws) they fully appreciate the chances for long losing streaks vs very profitable payoffs. and therefore far less likely to suffer emotionally.

3/ because it is NOT plausible to believe these consistent WINNING cappers (ATS/totals only) could actually form 3 team parlays daily at one book (because winning cappers shop lines at multiple outs and so parlays r NOT frequently possible) then only a small portion of his total number of wagers each week r even available to form into 3 team parlays!! i suggest it may fall into the range of about 10-20% of his total number of weekly wagers MIGHT be formed into 3 team parlays - NOT ALL HIS PICKS, as so many seem to completely misunderstand. therefore, if 80-90% of the weekly wagers of these few WINNING CAPPERS are flat-bet as usual, and only 10-20% of his picks r formed into parlays (16.4% win rate) then LOGICALLY his exposure to significant "risk of ruin" is basically a non-sequitor.
example: he wagers 3% his br on each of his daily picks, day after day. 80-90% of his picks r necessarily wagered that way, but he is able to form 3 team parlays with his last 10-20% of his picks -- whereupon he is only wagering 1% of his BR on each (ie. the one-third relationship to flat-betting i illustrated clearly in my essay) so how does a consistent WINNING CAPPER expose himself to significant danger of "risk of ruin" when, in practical application, less than 20% of his weekly picks can actually be formed into 3 team parlays, and even then only 1% of his BR is wagered on each compared with 3% BR flat betting???

You also forget the fact that the variance greatly increases this way and your potential for bankrupting your roll increases hugely with such a small win rate.

wow!!
from that statement it is OBVIOUS u didnt read my original essay (above)
i took great pains here too to discuss the famous "risk of ruin" (low win rate) and its practical application to betting only a very small percentage of the winning cappers weekly picks, and then at only ONE-THIRD the amount of bankroll compared with his usual flat-bets!

[phil_abuster]

Tsk, tsk. The 5.5% of bankroll is on EACH leg. Too much for a guy arguing Hareeba "apparently cant read".

So, the math goes (assuming a $10000 bankroll):

Three bets of $550 risk to win $500, 55% WP, (+EV = 5%) = $82.5 average profit.
One parlay risking $274.38 to win $1646.25, 16.64% WP (+EV = 16.46%) = $45.17 average profit.

I suppose now it is your turn to bash Kelly betting as nosense :/

no.
U r absolutely correct.
and i have no issues implimenting Kelly (for the WINNING cappers - of which there r very few)

this however doesnt affect the correct math on the significiant value of wagering OCCASIONAL 3 team parlays to the bonafide winning capper (55%)
and that is the whole point of the discussion

[phil_abuster]

Read up on variance, you'll find that with a 55% win rate a 10-0 loss streak is not even that exceptional and will occur quite often after a few thousand bets. Calculate what can happen (sorry I can't be bothered) if your win rate decreases to 16% or lower and you'll see you should be betting way less than 1% of your bank to keep the bankruptcy risk down to an acceptable level.

U r correct.
but unfortunately irrelevent.
U r correct *IF* every single wager the capper plays is somehow formed into 3 team parlays ONLY and at that 16% win expectancy.
bad news!
agreed. if and only if we assume ALL wagers bet in a given week r in the form of 3 team parlays
i submit that is implausible however.
theory is wonderful, but i am speaking only of the practical implimentation and there simply wouldnt be enough 3 team parlays formed by the 55% consistent, HIGHLY SKILLED, HIGHLY SELECTIVE winning capper to b of any real danger to his br.
to an AVERAGE capper (loser), who stupidly throws together 3 team parlays daily, YES.

but my essay is not in any way suggesting that a bonafide consistent winning capper can wager every single one of his picks into 3 team parlays!
the math PROVES the parlay-value, but in practical application, the consistent WINNIING capper (who shops multiple books, and is also VERY selective with betting!!) might be able to form maybe - MAYBE - 10-20% of his weekly picks into 3 team parlays.

[phil_abuster]

Tsk, tsk. The 5.5% of bankroll is on EACH leg. Too much for a guy arguing Hareeba "apparently cant read".

So, the math goes (assuming a $10000 bankroll):

Three bets of $550 risk to win $500, 55% WP, (+EV = 5%) = $82.5 average profit.
One parlay risking $274.38 to win $1646.25, 16.64% WP (+EV = 16.46%) = $45.17 average profit.

I suppose now it is your turn to bash Kelly betting as nosense :/

btw, in your example above
u r comparing wagering 3 bets for a total of $550 risk
to the same 3 picks formed into a parlay but only risking a total of $274.38
this is an improper comparison because the wager amounts r not equated.

nevertheless, because your stated bottom line results in $82.5 profit compared to just $45.17 profit your implication is the former (risking more is better)
thus, your conclusion appears to b that wagering 200% MORE but gaining only 182.6% greater return is better?!

Kelly: the ROI on $550 risked for $82.5 return is a good 15%
3 team parlays: the ROI on $274.38 risked for $45.17 return is 16.46%
and if the amounts were equated - as per proper comparison -
then the 3 team parlay return is just over $90

perhaps then, as it applies ONLY to the few consistent WINNING cappers,
a good strategy might be to apply a version of Kelly to 80-90% of his weekly picks and then
form occasional 3 team parlays whenever he can (ie. about 10-20% of his weekly picks)

[Cookie Monster]

For non winning cappers, implementing Kelly is very easy: Bet exactly zero. Fortunately, most of humanity consist of non-bettors, so they follow Kelly strictly

You are doing your homework, but still have much to learn. You three hour essay could be summed up as: The parlay has 16.46%EV, it is better than three 5% straight bets. If someone does not agree that 16.46 > 3*5, he should go to school. But if you risk the same on the parlay than in each straight, the advantage goes higher, as 16.46%>>5%!

Trust me, all of us once thought the same as you. We learned about advantage betting with the +EV. We wanted the biggest, best +EV, a yuuuuge +EV. And it kind of works, when you are comparing sides/totals close to +100. But +EV does not work when comparing bets with longer or shorter odds. Eventually we learned about money utility functions, the base of Kelly betting. And then we understand why +EV is not the number to maximize, but +EG (expected bankroll growth). And then bankroll management makes more sense (previously one could assume your "risk the same on a parlay than in a straight and win more").

You may not know Hareeba!, but he is a very sharp player. His posts are usually short and heavy on sarcasm, but he is wrong very rarely. So, instead of calling names, try to understand what are we saying. For example his "No intelligent punter would regularly risk the same stake on a 16.6375% chance as he would on a 55% chance." is a big truth, but you took it as " "No intelligent person would regularly risk $1000 on a 16.6% chance of winning". Doh.

no.
U r absolutely correct.
and i have no issues implimenting Kelly (for the WINNING cappers - of which there r very few)

this however doesnt affect the correct math on the significiant value of wagering OCCASIONAL 3 team parlays to the bonafide winning capper (55%)
and that is the whole point of the discussion

[phil_abuster]

You are doing your homework, but still have much to learn. You three hour essay could be summed up as: The parlay has 16.46%EV, it is better than three 5% straight bets. If someone does not agree that 16.46 > 3*5, he should go to school. But if you risk the same on the parlay than in each straight, the advantage goes higher, as 16.46%>>5%!

true.

Trust me, all of us once thought the same as you.

i think not.
i will repeat my platform again:
for the consistent, proven, ATS&totals capper, with an >53% win expectation on ALL his picks, he can increase his profit more by implimenting SOME 3 team parlays into his capping - when possible - as opposed to only flat betting.
the math is there and it is proven.
sadly, some people reading my essay seem to believe wrongly that in practice the consistent WINNING capper would actually be able to form multiple 3 team parlays daily AT ONLY ONE BOOK.
i repeat, the "value" is obvious, but the real life practical implimentation of 3 team parlays - for the consistent and highly selective WINNING capper is simply reduced to MAYBE just a small percentage of his total bets per week!
the whole discussion basically changes when we confine it (as I have) ONLY to those with a positive win expectation.
for the overwhelming majority of cappers forming parlays is NOT a good idea.
i have stated that repeatedly.

in effect, for an easy example, a consistent WINNING capper may average about 36 high quality ATS & totals picks per week (all with 55% expectancy based on his many years of carefully tracking all his wagers) of those 36, which r shopped among a few different books, only 6 of them would he actually find they all line up at the same book at the same time. thus 83.3% of his wagers r flat bet, let say at 3% his br ($100 ea.) from the start of the month. with only his last 6 wagers (16.7%) he will form just 2 parlays from 6 picks.
so, instead of flat-betting all 36 picks, he will flat bet 30 and only be able to form 2 parlays each week at one book.
this same scenario will repeat itself every week for a year (minus his 2 week holidays)
he is flat-betting 83.3% of his picks.
so each week the question becomes those last 6 picks.
should he continue to flat bet them, or form those into two 3 team parlays?
6 picks in question every week for 50 weeks = 300 in question
spread over an entire year, and remember this guy is a solid, consistent WINNING capper, so he is building his br as he rolls along
300 picks in question is the same number as my essay illustration, so we can refer directly to the math in my illustration for the answer.
yes, it would be a different discussion if this WINNING CAPPER actually could form his high quality, multiple-book, 300 picks into parlays within a month. now we r talking serious "risk of ruin" as i discussed before. AGREED!
but NOT when so few picks r actually capable of being formed into quality 3 team parlays - ATS & totals ONLY - by a consistent winning capper who by virtue of being a consistent winning capper is both very selective about his bets and uses multiple outs!
we must think about all this in real life practical terms, not theory!

try to understand what are we saying. For example his "No intelligent punter would regularly risk the same stake on a 16.6375% chance as he would on a 55% chance." is a big truth, but you took it as " "No intelligent person would regularly risk $1000 on a 16.6% chance of winning". Doh.

well it is important to put my statement into proper context, dont u agree?
it was speaking about the 1 in 6 chance of winning $25,000 by risking only $1000
though a fictitious and exaggerated example, it was intended to illustrate that some payoffs r definitely worth the risk.
forming parlays by the 95% of cappers (average/loser <52.38%) is definitely NOT worth the risk!
but for the consistent winning capper who would, in practice, be able to form very few 3 team parlays per week, it is DEFINITELY worth the small increase in risk as the payoffs r significantly higher than flat betting those same FEW picks.

from what ive read the past 24 hrs nobody seems to understand this, but oh well........

[yak merchant]

btw, in your example above
u r comparing wagering 3 bets for a total of $550 risk
to the same 3 picks formed into a parlay but only risking a total of $274.38
this is an improper comparison because the wager amounts r not equated.

nevertheless, because your stated bottom line results in $82.5 profit compared to just $45.17 profit your implication is the former (risking more is better)
thus, your conclusion appears to b that wagering 200% MORE but gaining only 182.6% greater return is better?!

Kelly: the ROI on $550 risked for $82.5 return is a good 15%
3 team parlays: the ROI on $274.38 risked for $45.17 return is 16.46%
and if the amounts were equated - as per proper comparison -
then the 3 team parlay return is just over $90

perhaps then, as it applies ONLY to the few consistent WINNING cappers,
a good strategy might be to apply a version of Kelly to 80-90% of his weekly picks and then
form occasional 3 team parlays whenever he can (ie. about 10-20% of his weekly picks)

You can make yourself feel better with your ROI calcs all you want. But all you are really doing is trading a large amount of variance and risk for the benefit of 6 to 1 payout instead of the true odds of 5.96 to 1 (assuming -110). And trust me all the extra ROI isn't coming from the difference between 6-1 and 5.96-1. It's because the denominator in the ROI calc is what you want it to be.

You can say you are only betting 100 on the parlay and 300 on the straights, but if you step back and look at it, parlays are really just a way to make a "bigger" bet on the come. In reality, you are making 3 if bets with a parlay. 100 to win 90.9, 190.9 to win 173.55 and 364.45 to win 331.32. And because most people who bet parlays are dopes the casino gives you a bonus of $4.

So like I said before, if you want to make yourself feel better about betting parlays with your jacked up ROI calcs, be my guest, but in an apples to apples world you are really committing to bet 655.35 with every parlay. So if you are sure you are hitting 55 percent, forget about waiting for all 3 plays to line up on the same day with the same book, and just roll your own 3 team parlays by betting leg 2 and leg 3 bigger on straight bets, and you'll still be rich even though you are only getting paid 5.96-1

[phil_abuster]

You can make yourself feel better with your ROI calcs all you want. But all you are really doing is trading a large amount of variance and risk for the benefit of 6 to 1 payout instead of the true odds of 5.96 to 1 (assuming -110). And trust me all the extra ROI isn't coming from the difference between 6-1 and 5.96-1. It's because the denominator in the ROI calc is what you want it to be.

ok.
ill just have to rest on the FACT that 3 teams parlays post a solid 9.8% greater profit (at 55% expectation) than does flat betting the same picks (as per my factual illustration, and given my oft-repeated caveats)
thank u

[phil_abuster]

You can make yourself feel better with your ROI calcs all you want. But all you are really doing is trading a large amount of variance and risk for the benefit of 6 to 1 payout instead of the true odds of 5.96 to 1 (assuming -110). And trust me all the extra ROI isn't coming from the difference between 6-1 and 5.96-1. It's because the denominator in the ROI calc is what you want it to be.

to b fair my ROI quotes to which u r referring were the comparison of Kelly to parlays

You can say you are only betting 100 on the parlay and 300 on the straights, but if you step back and look at it, parlays are really just a way to make a "bigger" bet on the come. In reality, you are making 3 if bets with a parlay. 100 to win 90.9, 190.9 to win 173.55 and 364.45 to win 331.32. And because most people who bet parlays are dopes the casino gives you a bonus of $4.

well, an extra $4.22 free money, actually
and i'll be as happy to take that as i would finding a 1/2 pt on a line at different book.

So like I said before, if you want to make yourself feel better about betting parlays with your jacked up ROI calcs, be my guest, but in an apples to apples world you are really committing to bet 655.35 with every parlay. So if you are sure you are hitting 55 percent, forget about waiting for all 3 plays to line up on the same day with the same book, and just roll your own 3 team parlays by betting leg 2 and leg 3 bigger on straight bets, and you'll still be rich even though you are only getting paid 5.96-1

and since if bets are virtually the same as 3 team parlays - less the $4.22 bonus for using parlays - then sure, go ahead and if bet to your hearts content - with the understanding that the ROR would similarly b the same albeit w/a slightly smaller payoff

[yak merchant]

ok.
ill just have to rest on the FACT that 3 teams parlays post a solid 9.8% greater profit (at 55% expectation) than does flat betting the same picks (as per my factual illustration, and given my oft-repeated caveats)
thank u

So your rebuttal to my position is "see above" bad math? Okay, let's make this simple. You have a magic football computer that picks 3 games each Saturday, one in the morning, one in the afternoon and one in the evening. Your magic computer has an exact win percentage of 55%. You so nice that you give me the exact three teams your computer picks so we can both bet them. We both borrow $100 from your mom, and head to a casino Saturday morning bright and early for Football Saturday with only the $100 in our pocket. You take your $100 and bet a 3 team parlay when we first get there. I take my 100 and bet it all to win on a 9:00am PST game. The nice people at the window give me 190.90 when it wins. Then I bet the 190.90 to win 173.55 on the afternoon game. When that wins I take the ticket for 364.45 and bet it all on the single game. You win your parlay and I win all 3 of my bets. You cash your parlay ticket, and you now have $700 in your pocket for a profit of $600. I cash my ticket and I have $696.00 in my pocket for a profit of $596.00, and we then head back to your mom's house. We repeat this every football Saturday for the next 10,000 football Saturdays. Each Saturday we have an expectation of coming home with that amount money 16.6375% of the time. The other Saturdays we come home with zero dollars. Therefore, over the long years of our friendship, we left your Mom's house with $1,000,000 and you returned home with $1,164,625.00 and I returned home with $1,157,970. You made $6,655.00 dollars more than me. Your ROI is 16.4625% and Mine is 15.797%, Your ROI is 0.6655% better than mine. So yes "technically" I put more money through the casino window, but we left the house with the same amount of money and the casino gave you a .6655% bonus for betting parlays. My factual illustration shows that your factual illustration of choosing your denominator and getting 9.8% increase in ROI is a crock.

[Optional]

to b fair my ROI quotes to which u r referring were the comparison of Kelly to parlays

I've been wondering for the last page and a half how this relates to the thread topic.

How does telling people to bet parlays translate as an alternative strategy to Kelly bet sizing?


Surely if your parlay plan is workable in real life then you still have to choose the stake size to bet on it. Why wouldn't you use Kelly theory to maximize those profits?

I don't get where you are coming from saying dont use Kelly but play parlays. Seems a bit like telling a person to ignore the doctors advice to go on a diet and practice tiddly winks each day instead. I don't see the either/or relationship.

[Hareeba!]

I've been wondering for the last page and a half how this relates to the thread topic.

How does telling people to bet parlays translate as an alternative strategy to Kelly bet sizing?


Surely if your parlay plan is workable in real life then you still have to choose the stake size to bet on it. Why wouldn't you use Kelly theory to maximize those profits?

I don't get where you are coming from saying dont use Kelly but play parlays. Seems a bit like telling a person to ignore the doctors advice to go on a diet and practice tiddly winks each day instead. I don't see the either/or relationship.

I'd been wondering about that from the start of this discussion but thought it might be stretching the grey matter a tad far to introduce that contradiction whilst he was still grappling with the comparison of straight betting to parlays.

[Hareeba!]

Apart from me there have been 6 contributors to this discussion Phil, and every one of them has disagreed with you.

What do you think that says?

[Optional]

BTW... the reason that three +EV legs in a parlay calculate out slightly better than straight betting is because if those bets really do have an edge, then parlaying them multiplies that edge.

But are you really good enough at capping to rely on your calculated edge being real 100% of the time?

I'm guessing for 99% of bettors inaccuracy there more than cancels out any advantage multiplying the edge %s in a parlay gives.

[ace7550]

Interesting topic.
I don't think anyone is arguing against Kelly here. The conversation got away from the original post topic rather quickly.
As far as the parlay subject goes, I think the higher your win % is the more it would make sense to use parlays. It's a simple way of putting it but I think it is the most direct.
You've obviously put a lot of thought into this Phil. If it's working for you then that's awesome.
PS John Wayne rules

[phil_abuster]

You made $6,655.00 dollars more than me.

Bottom line.
and my whole point
thank u.

[phil_abuster]

I've been wondering for the last page and a half how this relates to the thread topic.
How does telling people to bet parlays translate as an alternative strategy to Kelly bet sizing?

have no idea in regards to your specific statement since i am NOT telling people to bet parlays!
in fact i am saying the opposite as it relates to ~95% of cappers (including here)

Surely if your parlay plan is workable in real life then you still have to choose the stake size to bet on it. Why wouldn't you use Kelly theory to maximize those profits?

and if u read my posts u would have seen that i DID state i have NO ISSUES about implimenting Kelly (or a watered down version of it, like one-fifth, as one poster suggested)
if , for the benefit of those 5% of cappers who r proven consistent ATS & totals winners, they wish to also use Kelly in some form they r free to include that. i believe, in some variant, we all do.

I don't get where you are coming from saying dont use Kelly but play parlays.

non-sequitor.
i never stated dont use kelly nor would i. so your confusion stems from a false premise.
nor would i tell people how to bet.
i simply offered the math/tip.

[phil_abuster]

Apart from me there have been 6 contributors to this discussion Phil, and every one of them has disagreed with you.

What do you think that says?

since u ask ill answer:
it tells me several people have difficulty reading and understanding my point, despite my efforts to the contrary.
the common aversion to parlays is understandable - it accelerates losing - AS IT APPLIES TO THE 95% OF CAPPERS WHO DO NOT WIN CONSISTENTLY.
it is however a profitable tool for the few consistent ATS&totals winners, as is a version of Kelly.

by wrongly attributing a tip intended for consistent winners to themselves, the ~95% will react w/derision.
that i understand.
thats what i think it says to answer your Q

[Optional]

I think my "confusion" that you are arguing Parlays are a better method than Kelly came from your response to me here;

Quote Originally Posted by Optional View Post
Risk of ruin is there using full Kelly but there is no better way to maximize potential profits.
---
thanks. i appreciate your comments.

however, regarding your last statement i will submit there is an even better method to "maximize potential profits" than kelly

[Hareeba!]

since u ask ill answer:
it tells me several people have difficulty reading and understanding my point, despite my efforts to the contrary.
the common aversion to parlays is understandable - it accelerates losing - AS IT APPLIES TO THE 95% OF CAPPERS WHO DO NOT WIN CONSISTENTLY.
it is however a profitable tool for the few consistent ATS&totals winners, as is a version of Kelly.

by wrongly attributing a tip intended for consistent winners to themselves, the ~95% will react w/derision.
that i understand.
thats what i think it says to answer your Q

It doesn't.
All along I've accepted that we are talking about genuine 55% chances.
+600 parlays only advantage is marginal and then only if the alternative is straight betting at worse than -108 odds.
Your illusory advantage comes from the fact that you fail to recognise that a $100 bet on a +600 parlay is in effect a total bet of $657.21 on 3 events to return $700.

[phil_abuster]

I think my "confusion" that you are arguing Parlays are a better method than Kelly came from your response to me here;

huh??
you just claimed previously that I am saying "dont use kelly"

I don't get where you are coming from saying dont use Kelly but play parlays.

I NEVER SAID "dont use kelly"

u r falsely attributing statements to me which i never made.

in fact in response to one polite poster previously i formed the argument that it may be a good strategy FOR THE WINNING CAPPER ONLY to use kelly for his typical 80-90 of weekly pics which he simply cant form into parlays and then 3 team parlay the few which he can.

[yak merchant]

Bottom line.
and my whole point
thank u.

Whatever. If 0.66% and 9.8% are equivalent in your alternate dimension then that is awesome.

[phil_abuster]

Interesting topic.
I don't think anyone is arguing against Kelly here.
The conversation got away from the original post topic rather quickly.
As far as the parlay subject goes, I think the higher your win % is the more it would make sense to use parlays. It's a simple way of putting it but I think it is the most direct.
You've obviously put a lot of thought into this Phil. If it's working for you then that's awesome.
PS John Wayne rules

finally!
someone who can both read and understand my point.
U r correct --> nobody (including me) is arguing AGAINST using Kelly (or variations)
please DO NOT infer that because i am pointing out the advantage of 3 team parlays over flat-betting that this means Kelly should be ignored. far from it!

As far as the parlay subject goes, I think the higher your win % is the more it would make sense to use parlays.

YES!!
even though actually being able to form 3 team parlays AT ONE BOOK regularly, will, in practical application, result in very few weekly parlays compared with the vast majority of other wagers (and thus an insignificant increase in ROR) - the point is the math is there that proves the better "value" during those few occasions when 3 team parlays can be formed vs flat-betting the same FEW pics!

You've obviously put a lot of thought into this Phil.

somebody finally noticed. i appreciate/respect that!
i am simply trying to offer information of which some may not be fully aware
they r free to use or disregard as they wish.

Thank YOU, dude!

[phil_abuster]

It doesn't.
All along I've accepted that we are talking about genuine 55% chances.
+600 parlays only advantage is marginal and then only if the alternative is straight betting at worse than -108 odds.

correct!
and that was my point!
so whats your problem ??

Your illusory advantage comes from the fact that you fail to recognise that a $100 bet on a +600 parlay is in effect a total bet of $657.21 on 3 events to return $700.

another false premise.
i DO recognize this!
and inherent within the resulting math is an advantage, perhaps marginal as u state, but nevertheless an advantage!
bottom line!
and yes my example used the typical odds at the overwhelming majority of sportsbooks! and in that regard theres the advantage.
IF one can regularly get better odds then terrific!

[Hareeba!]

correct!
and that was my point!
so whats your problem ??




another false premise.
i DO recognize this!
and inherent within the resulting math is an advantage, perhaps marginal as u state, but nevertheless an advantage!
bottom line!
and yes my example used the typical odds at the overwhelming majority of sportsbooks! and in that regard theres the advantage.
IF one can regularly get better odds then terrific!

In previous posts you'd suggested a rather greater than marginal advantage for parlays.
And I'm sure you told me I was wrong to say that straight betting at -108 or better would produce a superior result.
Now you appear to be accepting my arithmetic?
Yes, I can ALWAYS get better than -110 and almost always better than -108 for spreads and totals.

[phil_abuster]

In previous posts you'd suggested a rather greater than marginal advantage for parlays.

obviously we view the term "marginal" differently
fine.

Yes, I can ALWAYS get better than -110 and almost always better than -108 for spreads and totals.

good for u, dude!