House Edge and Noobs

[Kasi]

is that allowed here? :joker:

ok so i been workin on this problem. at this point i'm a lost puppy.
just give me one hint. it doesn't seem as if you've given enough info to know the EV.
i got this one number 6,225 hands and don't know what to do with it or if it's even not bogus.
like maybe you could bet the $40 unit a thousand hands 6.225 times? probably bogus cause i'm not using any known procedure. to lazy to look it up.
and probably a bogus ev=-0.040 which seems ridiculous to my mind
am i gonna have to use that z-stat thing the wiz of odds talks about?
lmao i think i'll go look up the definition of EV.
so maybe i just need to look up the EV for a basic strategy only player? for what game though?

Probably not allowed here lol. But, maybe, even though voodoo, it may show, depending on goals and lenght of time playing, it may be a bit simplistic to just always say "eventually you'll lose". Although true lmao.

You're working too hard - I just meant it as a general thing to indicate how easy it might be to last 20000 hands and at least break even if that's your only goal.

Like in this case I start with $10,000 roll and $1 unit. 10000 units. I play 10000 hands and I'm supposed to be down $40 by then. (HA = 0.4% or .0004)

I have $9960 left. If I now switch my unit to $40, I now have only a 299 unit roll, down from the original 10000 units.

If I get at least 1 $40 unit ahead, flat-betting it, I am now $40 ahead and have made up the $40 I lost. I now have restored my original 10,000 unit roll if I switch back to a $1 unit and do it again lol. Heck, maybe I get lucky and win a $40 double. Good for 20,000 more hands now lol.

I just meant it'd be like if you bought in with $1495 at a $5 table, how often do you think, just flat-betting $5, you would be at least 1 unit ($5) ahead at some point (and that's when you would stop) before losing the entire $1495.

Intiuitively, I'd say pretty often lol.

Come to think of it, I think "What's the chances of finishing at least one unit ahead (pick a game, maybe AC multi) at any point in time flat-betting with a whatever unit roll in a -EV game before losing entire roll" is one of those pesky questions I don't think I've ever been able to answer to my own satisfaction.

Maybe it's more of a sim question lol. Any input on it for BJ appreciated.

See how voodoo is good after all - makes you ask questions

 

[sagefr0g]

Probably not allowed here lol. But, maybe, even though voodoo, it may show, depending on goals and lenght of time playing, it may be a bit simplistic to just always say "eventually you'll lose". Although true lmao.

You're working too hard - I just meant it as a general thing to indicate how easy it might be to last 20000 hands and at least break even if that's your only goal.

Like in this case I start with $10,000 roll and $1 unit. 10000 units. I play 10000 hands and I'm supposed to be down $40 by then. (HA = 0.4% or .0004)

I have $9960 left. If I now switch my unit to $40, I now have only a 299 unit roll, down from the original 10000 units.

If I get at least 1 $40 unit ahead, flat-betting it, I am now $40 ahead and have made up the $40 I lost. I now have restored my original 10,000 unit roll if I switch back to a $1 unit and do it again lol. Heck, maybe I get lucky and win a $40 double. Good for 20,000 more hands now lol.

I just meant it'd be like if you bought in with $1495 at a $5 table, how often do you think, just flat-betting $5, you would be at least 1 unit ($5) ahead at some point (and that's when you would stop) before losing the entire $1495.

Intiuitively, I'd say pretty often lol.

Come to think of it, I think "What's the chances of finishing at least one unit ahead (pick a game, maybe AC multi) at any point in time flat-betting with a whatever unit roll in a -EV game before losing entire roll" is one of those pesky questions I don't think I've ever been able to answer to my own satisfaction.

Maybe it's more of a sim question lol. Any input on it for BJ appreciated.

See how voodoo is good after all - makes you ask questions

omg, lmao. i just re-read the original post.
http://www.blackjackinfo.com/bb/showpost.php?p=95088&postcount=19
saw where you gave the EV and number of hands played. maybe i was a little sleepy when i was fooling around tryin to figure it out. totally missed that you gave the EV of -0.4% and i was workin off of 1,000 hands instead of 10,000 hands as you stipulated. and i was sober to i swear lol. but it was almost three in the morning. so anyway screwing up like that was an easy way to get you to provide the answer.
:joker::whip:
but anyway the really point like your saying if i may paraphrase.....
yeah your gonna lose it all flat betting some negative expectation game.
however that statement ignores what may be some interesting facts when viewed in light of what can happen in the short term. those facts being that it's not so unlikely to be up a unit at some given time before you do lose it all.
and like in your example if you changed your unit to what you lost it wouldn't be all that unlikely to regain what you lost by simply winning one unit which isn't really all that hard to do.
i think i got it right and tonight i am a bit tipsy. hot rice wine yum!

 

[sagefr0g]

....

Intiuitively, I'd say pretty often lol.

Come to think of it, I think "What's the chances of finishing at least one unit ahead (pick a game, maybe AC multi) at any point in time flat-betting with a whatever unit roll in a -EV game before losing entire roll" is one of those pesky questions I don't think I've ever been able to answer to my own satisfaction.

Maybe it's more of a sim question lol. Any input on it for BJ appreciated.

See how voodoo is good after all - makes you ask questions

i think maybe your intuition is for the good.
well i fiddled with it for a flat betting basic strategy player for the game shown below using the time constraint goal probablility calculator in cvcx.
also included below risk assesment of that 12 units for 100 hands

 

[Kasi]

i think maybe your intuition is for the good.
well i fiddled with it for a flat betting basic strategy player for the game shown below using the time constraint goal probablility calculator in cvcx.
also included below risk assesment of that 12 units for 100 hands

Thanks O Wise One - that's really interesting.

First of all I didn't think his calculators would work with -EV's but I guess they maybe must have been the "no time constraint calculators?". I have to check that out!

Anyway interesting it flattens out after 50 hands. Maybe playing 5000 hands wouldn't increase your chances that much? Would having a 300 unit roll and goal of 301 chnage the 90%?

And apparently you still could have have some your roll left anyway.

Like with single-zero roulette I think flat-betting red you have about a 94% chance of winning one unit before losing everything with a 40 unit or 400 unit roll - it doesn't make much difference. I'm not too good with time constraints.

But, getting back to our example, I guess you're saying I could take my $10K, split it into 12 $833 units, play at most 50 hands and have a 90% chance of winning at least $833? In effect, if i did that right off the top, and suceeded, I could now play 200,000 $1 hands with my 10,833 unit roll before EV would say I'm about even?

In effect, this voodoo has a 90% chance of surviving 200,000 hands by playing only 50 hands?

If something happens 10% of the time, going broke, even though here you wouldn't be broke that often after 50 hands, how many times does it take before you have a 50-50 chance of it happening at least once? I'm a favorite to last at least 1,000,000 hands aren't I? Is it maybe more like 12,000,000 hands?

I'd call this a high-risk voodoo system with a goal of at least breaking even over 1,000,000 hands

I suppose I now have to add the cover-all "I'll lose in the long-run" stuff so I can make fun of anybody betting a -EV game without knowing his roll, his goal, his game, his system, and his time lol.

 

[sagefr0g]

Thanks O Wise One - that's really interesting.

First of all I didn't think his calculators would work with -EV's but I guess they maybe must have been the "no time constraint calculators?". I have to check that out!

well, i used his probability of reaching a goal given a time constraint calculator. lmao maybe i wasn't sposed to do that. maybe the results are bogus?

Anyway interesting it flattens out after 50 hands. Maybe playing 5000 hands wouldn't increase your chances that much? Would having a 300 unit roll and goal of 301 chnage the 90%?

i didn't go beyond 100 hands. i guess it would rise ever slower?
i did try another higher roll amount. just 60 units. what it did was give virtually the same results up to i think it was 33 hands or so and then it 'improved' over the 12 unit roll. by improve i mean the probability increments widened giving a little better chance of making the one unit win over the chances for the 12 unit roll. i didn't finish that chart yet. see below.

And apparently you still could have have some your roll left anyway.

Like with single-zero roulette I think flat-betting red you have about a 94% chance of winning one unit before losing everything with a 40 unit or 400 unit roll - it doesn't make much difference. I'm not too good with time constraints.

well i guess you'd have some roll left and could look at it and say ok it's going bad time to bail or maybe it was going sorta ok so hang in for a while and hope for the unit.
yeah, i'm not good with time constraints or with out lol. so that's why i dunno if what i tryed to do was bogus or not. i just worked it off of a basic strategy sim that i created in cvcx. the calculators load the needed parameters from the sim and then you just tell it what roll you have and what roll you want and how many hands you want to play.

But, getting back to our example, I guess you're saying I could take my $10K, split it into 12 $833 units, play at most 50 hands and have a 90% chance of winning at least $833? In effect, if i did that right off the top, and suceeded, I could now play 200,000 $1 hands with my 10,833 unit roll before EV would say I'm about even?

In effect, this voodoo has a 90% chance of surviving 200,000 hands by playing only 50 hands?

well i was just foolin around with it with what your idea in general was but i wasn't trying to relate it to the specific numbers you used. i just chose 12 units cause the sim said the standard deviation was 11.274 units per hand.
thought i'd start there and see what happens sort of thing.
so but i guess your analysis is in line. but wouldn't your chance of making the one unit be lower than 90% if all you played was 50 hands? my results didn't reach 90% till 55 hands and were worse up until then. not sure lol gotta think about it.

If something happens 10% of the time, going broke, even though here you wouldn't be broke that often after 50 hands, how many times does it take before you have a 50-50 chance of it happening at least once? I'm a favorite to last at least 1,000,000 hands aren't I? Is it maybe more like 12,000,000 hands?

lmao heck i dunno. i'm getting lost again. i'm young and easily confused. but i'm gonna think about all that.

I'd call this a high-risk voodoo system with a goal of at least breaking even over 1,000,000 hands

I suppose I now have to add the cover-all "I'll lose in the long-run" stuff so I can make fun of anybody betting a -EV game without knowing his roll, his goal, his game, his system, and his time lol.

yeah ok. just hope this doesn't lead to a three day 86. maybe Sonny won't see it lmao. at least all this is in the general section. lmao, i'd like to try something along this line with fuzzy counting. :cat:

 

[sagefr0g]

Thanks O Wise One - that's really interesting.

First of all I didn't think his calculators would work with -EV's but I guess they maybe must have been the "no time constraint calculators?". I have to check that out!

.....

again like i said i used the "with a time constraint calculator". so your statement above prompted me to check in cvcx and see if any stipulation was made about +EV or what ever. there wasn't that i could find but there was a referance to Blackjack Attack page 136 for the formula. wow what a beaut that formula is! got my most dreaded things in it. trigonometry and infinite series and natural log's. Yuck!
and off hand i can't tell just from eyeballing the formula if it would be valid for a negative ev or not. i think that could be ascertained though. i just right now haven't the patience and fortitude to check the math rules.
then it tells how the formula comes from the field of finance and options theories. and they talk about stocks having some expected return or "drift" just as in blackjack we have a positive e.v.
so i dunno if because he talks about stocks having some expected return and us in blackjack having a positive e.v. it should mean that the formula is only good for +ev's and not -ev's. it would seem nonsense to mean that though since people lose money on stocks and options lots of times. unless they have this assumption that over the long haul all stock and options are going to appreciate positive returns. but thats simply baloney.
it kind of gives one pause to even wonder about the validity of the formulas in the first place when you hear they come from the field of finance.
financial prognastications are known to fail often enough. just one failior scientifically makes it invalid. but i can't really criticize the validity of the formula cause i can't even begin to understand it and i don't know if infact it always works in the financial field or not. just have reason for doubt. you'd have to think that if it was a valid formula it might be more likely to work for blackjack than finance. one thing i know is i wouldn't begin to cast any doubt on the likes of Schlesinger or QFIT as far as their confidence in the formula.
but i dunno i guess it seems reasonable the formula would work for a negative expectation situation simply because there is standard deviation involved as well. standard deviation that at times can be positive over and above the negative expectation. well it would be interesting to know how those finance guys derived that formula. i wouldn't be suprised if it started out from some sort of gambling related stuff. thats where statistics and probability originated from anyway. so maybe it is right for blackjack lmao.
i sure don't know i'm just as confused as ever.
so i'm gonna try and sim this basic strategy negative ev scenerio maybe with sage software and hold the results up against the goal probability with time constraint results of a similar cvcx sim. i guess thats possible but maybe it wouldn't be to meaningful since i guess you'd be hard pressed to get enough trials. maybe i'll get around to it in a few months or less or what ever. might be interesting but a bit tedious maybe can't even do it.
maybe the best defense for your voodoo scenerio is like you said 'intuitively' you have a good sense that it's not that hard to make up one unit when just playing basic strategy. lmao.