gambler's fallacy?

sagefr0g

sagefr0g

Well-Known Member
i think an idea of the gambler's fallacy is that it is wrong to think the past affects what is going to happen now, or what is going to happen in the future.

this is confusing to me when thinking about risk of ruin.
reason being is that say you have some starting bankroll, well all kinds of things can happen that affects the size of your bankroll. what happened in the past does affect your bankroll, unless it only adds or subtracts zero to or from it. but whatever, at any given time the size of your bankroll does affect your risk of ruin with respect to a given amount of time that you expect to play.

say i just play basic strategy against a csm machine. so negative expectation, and long term 100% risk of ruin.
but say i have only some given amount of time i plan to play and some given bankroll. so now my risk of ruin can be lower than 100%.

so i start out first time ever to play with some bankroll, hence some risk of ruin if i'm gonna play for some given amount of time.
so say i have some good luck but only played a partial amount of the planned upon play time. so now i've got more ammuniton, less play time to be exposed to risk and actually a lower risk of ruin for reaching my alloted play time.
hasn't my initial risk of ruin changed as a result of what happened in the past?
or
so i start out first time ever to play with some bankroll, hence some risk of ruin if i'm gonna play for some given amount of time.
so say i have some bad luck but only played a partial amount of the planned upon play time. so now i've got less ammuniton, less play time to be exposed to risk but possibly a higher risk of ruin for reaching my alloted play time.
hasn't my initial risk of ruin changed as a result of what happened in the past?

so now say i have some positive expectation value (a promotion) that is secondary and not totaly dependent upon the csm game, but because i've chosen to play the csm game it is such that the value of that positive expected value hinges upon playing against this csm for the aforementioned given amount of time. but there are other 'moderately' negative expectation games than the csm game that i could choose to play.

but say i've started off with csm and things go well part way through the alloted time. so wouldn't it be wise to stick with playing the csm against a now lower ROR as opposed to switching to the other possible game to play?
or
say i've started off with csm and things do not go well part way through the alloted time. so wouldn't it be wise to stop playing the csm against a now higher ROR and switch to the other possible game to play?

well in either case, isn't the past affecting things and my decisions on what to do affecting things with respect to the prospects of this overall positive expectation?

even more perplexing to me is the quesiton of if i want to start this process over again at some future date.
ie. to me the gambler's fallacy tells me the past doesn't penalize you for now or the future, but the law of large numbers tells me that if i keep on playing some negative expectation game then i face a 100% risk of ruin, sorta thing.

so i guess the answer is anything can happen in the short term but the long term can eventually catch up with you, sorta thing, but the nature of this stuff seems to me more short term than long term or even paradoxical with respect to what it really is.:confused:
 
K

k_c

Well-Known Member
sagefr0g said:
i think an idea of the gambler's fallacy is that it is wrong to think the past affects what is going to happen now, or what is going to happen in the future.

this is confusing to me when thinking about risk of ruin.
reason being is that say you have some starting bankroll, well all kinds of things can happen that affects the size of your bankroll. what happened in the past does affect your bankroll, unless it only adds or subtracts zero to or from it. but whatever, at any given time the size of your bankroll does affect your risk of ruin with respect to a given amount of time that you expect to play.

say i just play basic strategy against a csm machine. so negative expectation, and long term 100% risk of ruin.
but say i have only some given amount of time i plan to play and some given bankroll. so now my risk of ruin can be lower than 100%.

so i start out first time ever to play with some bankroll, hence some risk of ruin if i'm gonna play for some given amount of time.
so say i have some good luck but only played a partial amount of the planned upon play time. so now i've got more ammuniton, less play time to be exposed to risk and actually a lower risk of ruin for reaching my alloted play time.
hasn't my initial risk of ruin changed as a result of what happened in the past?
or
so i start out first time ever to play with some bankroll, hence some risk of ruin if i'm gonna play for some given amount of time.
so say i have some bad luck but only played a partial amount of the planned upon play time. so now i've got less ammuniton, less play time to be exposed to risk but possibly a higher risk of ruin for reaching my alloted play time.
hasn't my initial risk of ruin changed as a result of what happened in the past?

so now say i have some positive expectation value (a promotion) that is secondary and not totaly dependent upon the csm game, but because i've chosen to play the csm game it is such that the value of that positive expected value hinges upon playing against this csm for the aforementioned given amount of time. but there are other 'moderately' negative expectation games than the csm game that i could choose to play.

but say i've started off with csm and things go well part way through the alloted time. so wouldn't it be wise to stick with playing the csm against a now lower ROR as opposed to switching to the other possible game to play?
or
say i've started off with csm and things do not go well part way through the alloted time. so wouldn't it be wise to stop playing the csm against a now higher ROR and switch to the other possible game to play?

well in either case, isn't the past affecting things and my decisions on what to do affecting things with respect to the prospects of this overall positive expectation?

even more perplexing to me is the quesiton of if i want to start this process over again at some future date.
ie. to me the gambler's fallacy tells me the past doesn't penalize you for now or the future, but the law of large numbers tells me that if i keep on playing some negative expectation game then i face a 100% risk of ruin, sorta thing.

so i guess the answer is anything can happen in the short term but the long term can eventually catch up with you, sorta thing, but the nature of this stuff seems to me more short term than long term or even paradoxical with respect to what it really is.:confused:
Click to expand...
Risk of ruin is dynamic. If bankroll goes up then risk of ruin goes down if same betting strategy is used. If bankroll decreases then risk of ruin goes up if same betting strategy is used. At each separate point whatever has occurred in the past has no bearing on current risk of ruin.
 
sagefr0g

sagefr0g

Well-Known Member
k_c said:
Risk of ruin is dynamic. If bankroll goes up then risk of ruin goes down if same betting strategy is used. If bankroll decreases then risk of ruin goes up if same betting strategy is used. At each separate point whatever has occurred in the past has no bearing on current risk of ruin.
Click to expand...
ahh ok that helps quite a bit.

still i'm confused over the issue for the scenario i put forth in the original post.

i'll try and simplify it a bit:

say i play basic strategy against a csm negative expectation game for some set amount of time.
for that set amount of time the action has some risk of ruin but not 100%
but long run action it would be 100% risk of ruin
so say i 'get away with it" for lack of better terms and don't lose any money against the csm or don't lose any significant money relative to some expected value represented by some promotion, sorta thing.

so the question becomes if i try and attack this promotion in the same manner again, not just again, but again and again ......
errrhh well at some point i'd be flirting with that 100% ROR, no?
but when? errhh well i guess that's the thing one can't know.

so but ok here is what confuses me, well, each of these seperate 'attacks' against the promotion are for some limited amount of time, which represents some amount of hands played and consequentially some risk of ruin less than 100% .
so do i have to for each 'attack' no matter how successful or unsuccessful, do i have to consider the total number of hands played for the aggregate number of 'attacks', in order to determine my risk of ruin for a current or future 'attack'?
or can i just consider each 'attack' that was already made as inconsequential to any present or future 'attack' and only have to account for the number of hands for the present or future 'attack' when figuring my risk of ruin?
 
K

k_c

Well-Known Member
sagefr0g said:
ahh ok that helps quite a bit.

still i'm confused over the issue for the scenario i put forth in the original post.

i'll try and simplify it a bit:

say i play basic strategy against a csm negative expectation game for some set amount of time.
for that set amount of time the action has some risk of ruin but not 100%
but long run action it would be 100% risk of ruin
so say i 'get away with it" for lack of better terms and don't lose any money against the csm or don't lose any significant money relative to some expected value represented by some promotion, sorta thing.

so the question becomes if i try and attack this promotion in the same manner again, not just again, but again and again ......
errrhh well at some point i'd be flirting with that 100% ROR, no?
but when? errhh well i guess that's the thing one can't know.

so but ok here is what confuses me, well, each of these seperate 'attacks' against the promotion are for some limited amount of time, which represents some amount of hands played and consequentially some risk of ruin less than 100% .
so do i have to for each 'attack' no matter how successful or unsuccessful, do i have to consider the total number of hands played for the aggregate number of 'attacks', in order to determine my risk of ruin for a current or future 'attack'?
or can i just consider each 'attack' that was already made as inconsequential to any present or future 'attack' and only have to account for the number of hands for the present or future 'attack' when figuring my risk of ruin?
Click to expand...
Whatever the aggregate of your assets is at any point determines present risk of ruin. A promotion could be an asset, maybe allowing a bigger win than normal upon a winning trial, lowering risk of ruin. If promotion expires by the time the next trial arrives then risk of ruin could go back up to where it would be without the promotion due to loss of an asset.
 
sagefr0g

sagefr0g

Well-Known Member
k_c said:
Whatever the aggregate of your assets is at any point determines present risk of ruin. A promotion could be an asset, maybe allowing a bigger win than normal upon a winning trial, lowering risk of ruin. If promotion expires by the time the next trial arrives then risk of ruin could go back up to where it would be without the promotion due to loss of an asset.
Click to expand...
yup ok, i got that part.
but like for the aggregate of my assets, do i have to consider past hands played or just future?

edit:
like ok in the images below
the first image is for a five hour session circa 660 hands played
where the risk of ruin is 5.39%

the second image is for a ten hour session circa 1320 hands played
where the risk of ruin is 19.16%

ok but here is the thing, i can understand that for my first five hour session i would have a risk of ruin of 5.39%.
but ok say i perform that five hour 660 hands played session and quit having lost nothing.

so now i want to play another five hour session of 660 hands.
is my risk of ruin again 5.39% since i'm just playing 660 hands or is my risk of ruin 19.16% since i've already played 660 hands and i'm now going to play 660 more hands for a sum total of 1320 hands?:confused:

well i think the answer is that i'm not penalized by whatever happened in past sessions, so i think my second session of 660 hands the risk of ruin would be 5.39%.
but it seems strange, errh i mean the long run risk of ruin is 100%, so really at some point i'd expect to be penalized with respect to risk just for playing more and more hands, sorta thing.:rolleyes:
 

Attachments

C

Canceler

Well-Known Member
You've mostly got it, but...

You played 660 hands in a –EV game and broke even. Lucky you! Do you expect that luck to hold? (Hint: You should be expecting to lose 3.82 per 100 hands.)

You are going to be “penalized” in the future, not by the weight of the hands you’ve previously played (which have no relevance), but by the weight of the house advantage on the hands you’re going to play.
 
sagefr0g

sagefr0g

Well-Known Member
yup now i got it

Canceler said:
You played 660 hands in a –EV game and broke even. Lucky you! Do you expect that luck to hold? (Hint: You should be expecting to lose 3.82 per 100 hands.)

You are going to be “penalized” in the future, not by the weight of the hands you’ve previously played (which have no relevance), but by the weight of the house advantage on the hands you’re going to play.
Click to expand...
there ya go Canceler......... now i got it, believe it or not.:)
 
M

Machinist

Well-Known Member
Sir Froggo,,,,,,,,,,,you have a little to much time on your hands it appears to me :p
-EV is minus ev.....no matter how you try to figure it.....Its gonna get ya.....
Lets worry about +EV.....its alot more funner and you know dam well your gonna increase your bankroll ......NO??????
Get these thoughts out of your head.......Sh!t happens in life then we die......-EV..Things that we do along the way...that's +EV.......yuck yuck!!:laugh::laugh:

Machinist
 
sagefr0g

sagefr0g

Well-Known Member
Machinist said:
Sir Froggo,,,,,,,,,,,you have a little to much time on your hands it appears to me :p
-EV is minus ev.....no matter how you try to figure it.....Its gonna get ya.....
Lets worry about +EV.....its alot more funner and you know dam well your gonna increase your bankroll ......NO??????
Get these thoughts out of your head.......Sh!t happens in life then we die......-EV..Things that we do along the way...that's +EV.......yuck yuck!!:laugh::laugh:

Machinist
Click to expand...
no worry Mach, it's part of a rope a dope, sort of a ploy.
just i wanna make sure i don't get one too many concussions like Ali did.:eek:
nuthing worth writing home about though.:(
 

Attachments

sagefr0g

sagefr0g

Well-Known Member
hmm lol i may have too much time on my hands

considering these two images...... further text below




would it be proper to conclude that the difference in risk illustrated by the images is mainly caused by the number of hands or time exposed to the negative expectation game 'fueled' by the negative aspects of variance?

well whatever, it appears that from the stand point of session length, that shorter sessions have a potential for greater safety with respect to risk of ruin.
as such it would appear that playing two sessions as opposed to one long session where the aggregate hands were equal has at least the potential of being safer.

is that a fact or an illusion?
 
L

London Colin

Well-Known Member
sagefr0g said:
as such it would appear that playing two sessions as opposed to one long session where the aggregate hands were equal has at least the potential of being safer.

is that a fact or an illusion?
Click to expand...
The whole concept of sessions is illusory. If you have a fixed bankroll, then ruin occurs if you exhaust it. Splitting your play into sessions does not change anything.

You are not going to start your second session with a fresh $400, or if you do then the concept of 'ruin' does not apply, since you are able to bring more money to the table than your initial $400.

In other words, two short trips, each starting with $400, may have a lower probability that you will bust out on either trip, compared to one long trip, but that's because you have to provide more cash if you make a loss on the first trip.
 
sagefr0g

sagefr0g

Well-Known Member
London Colin said:
The whole concept of sessions is illusory. If you have a fixed bankroll, then ruin occurs if you exhaust it. Splitting your play into sessions does not change anything.

You are not going to start your second session with a fresh $400, or if you do then the concept of 'ruin' does not apply, since you are able to bring more money to the table than your initial $400.

In other words, two short trips, each starting with $400, may have a lower probability that you will bust out on either trip, compared to one long trip, but that's because you have to provide more cash if you make a loss on the first trip.
Click to expand...
yes i see your point, what i was saying about the potential for less risk would only be true in the short term case where one was lucky enough to either not lose any of the $400 or maybe even gain a bit of money, sorta thing. that is the point i was missing when comparing the two images.
and the point Mach and Canceler made with respect to what we expect for negative ev games solidifies your point even more.
so like say you start out with $400 you'd only have a 5.33% ROR but if say you had lost $200 in that session then you'd be faced with a much greater risk of 34.9% for the next similar session as in the image below........ pick an amount one might have lost and whatever the risk just gets worse, sorta thing.
got it!
 

Attachments

M

Machinist

Well-Known Member
I knew a guy that gambled every friday.......He would cash his paycheck....at the casino..go to the blackjack table bet 1 black.........................win or lose he walked........Dont know weather he won or lost in the long run....
But the question to be asked is.............is that one hand a session???? :rolleyes:
Just got back from a month long working trip...getting caught up on reading posts...

Machinist
 
You must log in or register to reply here.
Top