Myths in the Movie 21
- Thread starter Guynoire
- Start date
Yes, No and Maybe
Yes, I understand the kelly theory requires constant resizing.
I stated 1/6th Kelly (though using the Kelly name was perhaps incorrect) fixed bets, it is terminology some of the viewers may be familar with. The initial ROR I spoke of .000631% was fixed betting with no resizing. If you are willing to cut to half stakes if you lose half the bank then it became the .0000015% ROR Sonny spoke of.
It's not hard to bet with a .000631% ROR and cut to half stakes if necessary and not fall below table minimums ($5). Again, one can get close to the theory of 0% ROR in the real world.
Yes, I understand the kelly theory requires constant resizing.
I stated 1/6th Kelly (though using the Kelly name was perhaps incorrect) fixed bets, it is terminology some of the viewers may be familar with. The initial ROR I spoke of .000631% was fixed betting with no resizing. If you are willing to cut to half stakes if you lose half the bank then it became the .0000015% ROR Sonny spoke of.
It's not hard to bet with a .000631% ROR and cut to half stakes if necessary and not fall below table minimums ($5). Again, one can get close to the theory of 0% ROR in the real world.
Well, here's my thoughts lol.
If you have a $10K roll with a 10% ROR (I'm gonna assume "lifetime"), then,yes, you have almost a 90% chance of doubling your roll if you play "forever". At that point, since you have now doubled the units in your original and you keep the same dollar spread, you now, from that point forward, have a 99% chance of doubling your original 10K roll again. (.1*.1=.01).
If you kept the same dollar spread, say $5-$500, whatever, your original ROR of 10% would now be .1^4 = 0.01% with a $40K roll since your original
roll has now quadruopled in units.
You can bet $10K or $100K with the same ROR as long as you *10 your original $ spread since you will have same # of units in roll (like $50-$5000 now for example now).
You want to get really close to 0% ROR, then keep the same $5-$500, for example, spread with a $100K roll instead of an $10K roll. You now have a 20000 unit roll instead of a 2000 unit roll.
It's impossible to bet the same absolute dollar spread and have the same ROR with different rolls since the units in the roll you are betting to will have changed.
The risk of halving original roll playing forever at some point is square root of original risk. So with an original lifetime ROR 0f 10%, the risk of halving your roll at some point is 31.6%.
Practically speaking, since none of us can play for infinity, pick a number of hands.
With an original low lifetime ROR, chances are if you ever double your original roll, you will win money forever since, after that point, doubling your original or roll or 100* your original roll are, practically speaking, about the same keeping original $ spread the same. After all, one only makes x units per hand with the same fixed $ spread, whether you have 1 or 1000000 units in roll.
Hope that helps a little.
If you have a $10K roll with a 10% ROR (I'm gonna assume "lifetime"), then,yes, you have almost a 90% chance of doubling your roll if you play "forever". At that point, since you have now doubled the units in your original and you keep the same dollar spread, you now, from that point forward, have a 99% chance of doubling your original 10K roll again. (.1*.1=.01).
If you kept the same dollar spread, say $5-$500, whatever, your original ROR of 10% would now be .1^4 = 0.01% with a $40K roll since your original
roll has now quadruopled in units.
You can bet $10K or $100K with the same ROR as long as you *10 your original $ spread since you will have same # of units in roll (like $50-$5000 now for example now).
You want to get really close to 0% ROR, then keep the same $5-$500, for example, spread with a $100K roll instead of an $10K roll. You now have a 20000 unit roll instead of a 2000 unit roll.
It's impossible to bet the same absolute dollar spread and have the same ROR with different rolls since the units in the roll you are betting to will have changed.
The risk of halving original roll playing forever at some point is square root of original risk. So with an original lifetime ROR 0f 10%, the risk of halving your roll at some point is 31.6%.
Practically speaking, since none of us can play for infinity, pick a number of hands.
With an original low lifetime ROR, chances are if you ever double your original roll, you will win money forever since, after that point, doubling your original or roll or 100* your original roll are, practically speaking, about the same keeping original $ spread the same. After all, one only makes x units per hand with the same fixed $ spread, whether you have 1 or 1000000 units in roll.
Hope that helps a little.
zero percent 
did you see this post? thought maybe you missed it. http://www.blackjackinfo.com/bb/showpost.php?p=90479&postcount=65
by the way these sims were for basicly the same game the warrior's are up against. so different games and all different situation.
right Kasi we were thinkin lifetime ror at least starting out on the discussion.
to get better than 50% chance you gotta go maybe 35,000 hands.
did you see this post? thought maybe you missed it. http://www.blackjackinfo.com/bb/showpost.php?p=90479&postcount=65by the way these sims were for basicly the same game the warrior's are up against. so different games and all different situation.
right Kasi we were thinkin lifetime ror at least starting out on the discussion.
to get better than 50% chance you gotta go maybe 35,000 hands.
Attachments
Therein lies the problem. Even if computing power did increase perfectly exponentially (ala Moore's Law), the casino would have to be in a constant state of upgrading the hardware to keep up with the data.
There are some interesting real world limitations, though. There's only so many people who will visit Las Vegas, so the growth of the data pool will level out after a while. So the "gather" stage will take less and less computing power.
It would also be possible to optimize the "search" stage by de-prioritizing or even excluding certain records. Visitors who haven't been seen in years. Known non-cheaters ("trusted" customers). Etc.
I guess it comes down to how sensitive the computer is to changes. If something as simple as a pair of glasses will fool it, then you have a nigh-unlimited number of disguises. Just having 4 pairs of glasses, and 4 fake mustaches gives you 16 new faces. If you have 10 aspects that you can change, and 10 variations on each, you probably have a lifetime of disguises. (You may want a log book to make sure you don't duplicate, though =) )
Interestingly enough, the laptop I bought last month comes with a built in webcam and facial recognition "login" software. I disabled it because it was a resource hog, but that should show you how common its getting.
I guess it'll all come down to how much they'll spend on the equipment + software + running/maintenance costs vs. how much they'll potentially save catching cheaters vs. how much Laurence Fishburne's eyeballs cost. =)
There are some interesting real world limitations, though. There's only so many people who will visit Las Vegas, so the growth of the data pool will level out after a while. So the "gather" stage will take less and less computing power.
It would also be possible to optimize the "search" stage by de-prioritizing or even excluding certain records. Visitors who haven't been seen in years. Known non-cheaters ("trusted" customers). Etc.
I guess it comes down to how sensitive the computer is to changes. If something as simple as a pair of glasses will fool it, then you have a nigh-unlimited number of disguises. Just having 4 pairs of glasses, and 4 fake mustaches gives you 16 new faces. If you have 10 aspects that you can change, and 10 variations on each, you probably have a lifetime of disguises. (You may want a log book to make sure you don't duplicate, though =) )
Interestingly enough, the laptop I bought last month comes with a built in webcam and facial recognition "login" software. I disabled it because it was a resource hog, but that should show you how common its getting.
I guess it'll all come down to how much they'll spend on the equipment + software + running/maintenance costs vs. how much they'll potentially save catching cheaters vs. how much Laurence Fishburne's eyeballs cost. =)
on the issue of gamble or investment
even though the maths of blackjack advantage play is based upon the blessed normal distribution of true counts and the associated advantages we could still complicate the matter and ask if there may be any Black Swans involved:
http://www.fooledbyrandomness.com/GIF.pdf
even though the maths of blackjack advantage play is based upon the blessed normal distribution of true counts and the associated advantages we could still complicate the matter and ask if there may be any Black Swans involved:
http://www.fooledbyrandomness.com/GIF.pdf
sagefr0g said:
zero percent did you see this post? thought maybe you missed it. http://www.blackjackinfo.com/bb/showpost.php?p=90479&postcount=65
by the way these sims were for basicly the same game the warrior's are up against. so different games and all different situation.
right Kasi we were thinkin lifetime ror at least starting out on the discussion.
to get better than 50% chance you gotta go maybe 35,000 hands.
did you see this post? thought maybe you missed it. http://www.blackjackinfo.com/bb/showpost.php?p=90479&postcount=65by the way these sims were for basicly the same game the warrior's are up against. so different games and all different situation.
right Kasi we were thinkin lifetime ror at least starting out on the discussion.
to get better than 50% chance you gotta go maybe 35,000 hands.
well lemme see if i can talk it where it sounds like sense lol.
the game basicly same one the warriors are doing. a 1:24 spread.
sim shown in this link: http://www.blackjackinfo.com/bb/showpost.php?p=90861&postcount=88
now to get that 10% lifetime ROR we were shooting for on the ten grand bankroll called for a $7 unit. so with all that the sim shows you make about $20.37 per hour or circa $20.37 per hundred hands say.
also $10,000/$20.37 = 490.92
and 490.92 x 100hand = 49,092 hands to make circa $10,000.00
so this is all just a rough way of looking at it but it gives you an idea that you need way more than 1,000 hands to double your $10,000.00 .
but using the goal with time constraint probability calculators in cvcx they show you need about 35,000 hands to have a better than 50% probability of doubling the bank.
thats for this particular case. the ten grand bankroll at 10% ror case with a 1:24 spread.
the game basicly same one the warriors are doing. a 1:24 spread.
sim shown in this link: http://www.blackjackinfo.com/bb/showpost.php?p=90861&postcount=88
now to get that 10% lifetime ROR we were shooting for on the ten grand bankroll called for a $7 unit. so with all that the sim shows you make about $20.37 per hour or circa $20.37 per hundred hands say.
also $10,000/$20.37 = 490.92
and 490.92 x 100hand = 49,092 hands to make circa $10,000.00
so this is all just a rough way of looking at it but it gives you an idea that you need way more than 1,000 hands to double your $10,000.00 .
but using the goal with time constraint probability calculators in cvcx they show you need about 35,000 hands to have a better than 50% probability of doubling the bank.
thats for this particular case. the ten grand bankroll at 10% ror case with a 1:24 spread.
aslan said:
Heck, I've been using a mere 1 to 10 spread in the weekend warriors and have doubled my $5,000 br several times over, so I'm trusting what you're saying but I'm wondering why I'm doing so well when the RoR says I don't have a chance of doing as well as I'm doing or something like that.
you've made $41,076.00 in 8,588 hands.
so your nothin unexpected just a mere 4 high standard deviations from the norm.
just like i was lol before my great fall from stalkoidism.
now if you was on the other side of the bell curve. yeah your ror woulda beat you down pretty bad.
Attachments
Ah yes - into the breach again lol.
"Their's not to make reply,
Their's not to reason why,
Their's but to do and die:
Into the valley of Death
Rode the six hundred...
When can their glory fade?
O the wild charge they made!"
But, even in the real world, I'd think one could make one's ROR as low as one wants, as close to zero as one wants. Just not zero.
If you tell me what "lifetime" (playing forever) real-world ROR you desire and I tell you how to do it or, even if you do put a time limit on it, make it as close to zero as you like, and none of that sissy re-sizing stuff either, will you still do that :whip: stuff for me? :grin:
"Their's not to make reply,
Their's not to reason why,
Their's but to do and die:
Into the valley of Death
Rode the six hundred...
When can their glory fade?
O the wild charge they made!"
But, even in the real world, I'd think one could make one's ROR as low as one wants, as close to zero as one wants. Just not zero.
If you tell me what "lifetime" (playing forever) real-world ROR you desire and I tell you how to do it or, even if you do put a time limit on it, make it as close to zero as you like, and none of that sissy re-sizing stuff either, will you still do that :whip: stuff for me? :grin:
aslan said:
Heck, I've been using a mere 1 to 10 spread in the weekend warriors and have doubled my $5,000 br several times over, so I'm trusting what you're saying but I'm wondering why I'm doing so well when the RoR says I don't have a chance of doing as well as I'm doing or something like that.
Basically, I've always been a little unclear exactly which sim you may be betting to lol. Not to mention I'm not real clear with KO preferred or whether you are using indexes.
On the other hand, whatever roll you may be betting to, once you have doubled that roll that you are betting to, whether you actually ever had it or not in the first place, chances that you will money forever greatly increase.
So, how have you been betting?
sigh. we've been through this before guys.
http://www.blackjackinfo.com/bb/showpost.php?p=48010&postcount=7
http://www.blackjackinfo.com/bb/showpost.php?p=48010&postcount=7
Lifetime (real world) is not playing forever. Real world has a finite number of hands you are playing. So one can get so far ahead that given their lifetime they will not go broke! A very extreme case but it will prove my point. You have won 100,000 playing BJ, you are down to the last hour of your playing career. You only backcount and place a top bet of $25. Now this is an insane, perverse and extreme example, but I hope it shows a 0% ROR attainable in the real world.
sissy resizing?:whip:
resizing is where the magic is:joker:
most return/least risk:devil:
If real Kelly betting is continuous resizing and that is optimal, then the real question is how to mimic Kelly betting in the real world.
sissy resizing?:whip:
resizing is where the magic is:joker:
most return/least risk:devil:
If real Kelly betting is continuous resizing and that is optimal, then the real question is how to mimic Kelly betting in the real world.