PDA

View Full Version : mathematical demonstration of the advantage of keep betting in a lucky strike



raaaid
12-18-2010, 07:46 AM
folklore knows if you are in a winning streak you have to keep betting while if you are losing is better to retire

ill prove this is mathematically true and logic:

lets analize the odds of following this startegy in a single game

my startegy is bet in pairs, on draw(one win one lose) i keep gambling, on losing i retire (one lose one lose), i only keep gambling on win(one win one win) or on draw

in this universe the first two posible outcomes are:

00=win or 11=lose

01 or 10 doesnt exist in this universe cause if this was the case i would keep gambling, i bet 0.5 euro in ech game

so the first two posible games that exist in the universe of this staregy are:

win or lose 00 or 11 for this two theres a chance of 1/4 each profit=0

the next posible outcomes in this universe are:

ww or wl with a chance 1/16 profit 1 euro

then www or wwl chance 1/64 profit 2 euro

then wwww or wwwl chance 1/256 profit 3 euro

then wwwww or wwwwl chance 1/1024 profit 4 euro

...


these are the only posible outcomes of a single game with this strategy the rest just dont exist

so now its just a matter of adding the odds:

1/16+ 2/64 +3/256 +4/ 1024 ...=0.109375

so following the staregy of keeping in a lucky strike and retiring on a bad one gives you a 10% edge

popular folklore is right on this one

thefruitbat
12-18-2010, 08:52 AM
are you incapable of keeping any random thought you have to yourself.

NOBODY cares.

how much spam is enough?

http://www.todayifoundout.com/wp-content/uploads/2010/09/spam1.jpg

raaaid
12-18-2010, 09:03 AM
man this is not spam this might be my best thread ever

and id ont complain for the ten pager on wikileaks i just dont bother with it

thefruitbat
12-18-2010, 09:35 AM
Originally posted by raaaid:

and id ont complain for the ten pager on wikileaks i just dont bother with it

that's just one thread, as opposed to the 15 you have on the front page.

here's an idea, instead of opening a new thread for every inane idea that you think is great and revolutionary, DON'T. Just append one of the others.

Unfortunately, in spite of the fact that your on my ignore list, it doesn't hide your millions of threads on the front page, so i have to scroll through all your cr*p to see everybody else's stuff.

happy christmas.

raaaid
12-18-2010, 09:46 AM
oh i see you re annoued cause you have to use your mouse wheel not because you would like not mainstream ideas censored

oh what a torment

myslef not scrolling down i admitt it

raaaid
12-18-2010, 09:59 AM
with this like with quantum entanglement is imposible to know whats gonna be a randon number

still nonlocality as with quantum entanglement is big enough as to can have a winning startegy

Zeus-cat
12-18-2010, 10:25 AM
Where exactly do you live raaaid? I really need to open a casino near you!!! And you can have all the beer and wine you can drink for free!

raaaid
12-18-2010, 12:00 PM
actually i got it reversed lets start all over with the simetric strategy:

i gamble 3 couples of games then stop playing unless my startegy makes me stop before

if i draw getting 01 or 10 i keep gambling if i win i stop if i lose i keep at it untill i win or the 3 couple of games are over

this is the universe of posible end games:

w i retire
lw i retire
llw or lll or ll draw or ll draw

there are no other elements in this universe

for w the chance is 1 / 4 so theres a chance 1 /4 i win one
lw doesnt count is a draw 0 chance 1/(4*4)
for llw or lll or lldraw or lldraw theres a chance 1/(4*4*4) each

for llw i lose -1 with lll i lose -3 with lldraw i lose -2 with ll draw i lose -2

so lets add the odds:

win chance 1* 1/4=16/64
lose chance -8* 1/64= -8/64


edit

i neglected ldl and ld"l so actually i can lose - 12/64 but win 16/64

dll doesnt exist in this universe because if i draw on first i stop

raaaid
12-18-2010, 12:49 PM
oh first ill make sure ive really found the non locality of randomness with my toy roulete and buttons

if it works i may think about gambling for real though probably not cause i dont mind about money

raaaid
12-18-2010, 03:14 PM
the key is im applying a tetravvalent value w,l,d or d" to a binary tree

imagine i follow the startegy of completly ignore draw just dont count them an i applied this system

the universe of this startegy would be only composed of

w +1 chance 1/2(now its 1/2 not 1/4)
lw 0 chance 1/(2*2)
llw -1 chance 1/(2*2*2)
lll -3 chance 1/(2*2*2)
now with this startegy i win 1*4/8

and lose -4 * 1/8

0 gain 0 lose chance

http://forums.ubi.com/groupee_common/emoticons/icon_smile.gif

but if you apply a tetravalent tree to a binary one locality of randomness is broken http://forums.ubi.com/groupee_common/emoticons/icon_smile.gif still you cant know the future but you can have an strategy to have future to be favorable

raaaid
12-18-2010, 03:58 PM
well ive done the probelm over and over and i get an edge of 4/64, at this i have to substract the house edge of when the ball falls on 0 which is 1/37

with this i get a 3% edge

ill have pretty much fun testing it at my toy roulette http://forums.ubi.com/groupee_common/emoticons/icon_smile.gif

raaaid
12-19-2010, 06:33 AM
i realized of my mistake:


i neglected ldd", ld"d , ldd and ld"d" with with the 4/64 edge is lost http://forums.ubi.com/groupee_common/emoticons/icon_frown.gif

GoToAway
12-19-2010, 06:41 AM
House has $200 riding on a hand.
The cards total 18.
Should he double down?