Giving up ev for reduced variance - from Gambling With an Edge

Do you

  • I'd take $1,500

    Votes: 9 37.5%
  • I'd take the variance.

    Votes: 15 62.5%

  • Total voters
    24
  • Poll closed .

sagefr0g

Well-Known Member
#21
not enough options in the poll, for me.
if there was a load of these plays and i wasn't ready to call it a night, i'd risk the variance,,,,,,,,
if it was the last play of the night and i had a long drive home, i'd probably take the sure thing.:rolleyes:
 
#22
sagefr0g said:
not enough options in the poll, for me.
if there was a load of these plays and i wasn't ready to call it a night, i'd risk the variance,,,,,,,,
if it was the last play of the night and i had a long drive home, i'd probably take the sure thing.:rolleyes:
Sage,

It is the last hand of the night, you know an ace was coming even though the count is zero. You shove out $1500, and bang you get that ace and a lovely paint with it. Oops, now the dealer has an ace up and asks, "Insurance?" You would insure that blackjack?
 

sagefr0g

Well-Known Member
#23
Richard Munchkin said:
Sage,

It is the last hand of the night, you know an ace was coming even though the count is zero. You shove out $1500, and bang you get that ace and a lovely paint with it. Oops, now the dealer has an ace up and asks, "Insurance?" You would insure that blackjack?
no.
 

sagefr0g

Well-Known Member
#25
Richard Munchkin said:
Why would this be different?
to be honest i'm not entirely sure i understand fully the original post.
but anyway the blackjack scenario is a scenario where one can expect 'many' plays over ones playing career, this envelope promotion is likely relatively 'few' plays, no? that doesn't necessarily mean to not go for the highest ev with out regard for variance, in the case of the envelope promotion, but the two scenarios are different in that regard. so but maybe ev is ev, regardless the scenario so really nothing wrong with going after the higher ev regardless the variance. still, maybe a distinction still holds, in that if this is a 'rare' play and you have this certainty of a given payoff as opposed to the possibility of a lower payoff, for this rare event.
so, i don't really understand certainty equivalent very well, but those arguing CE seem to have a point. imagine you are wandering, lost in the desert, you have a chance for a life sustaining drink of water, or more water than you really need, or alas, not enough water. but of all those choices the life sustaining choice is the only sure one, which would you take, while at the same time you know that if you live you have a decent chance of finding gold nuggets out there in the desert and more water later?:rolleyes:
 
#26
sure thing a good thing?

1666 gamble win on average?
1500 sure thing?
The CE of the sure thing seems really high compared to the gamble.
If this was bj one would take the sure thing.
 

Percy

Well-Known Member
#27
sagefr0g said:
to be honest i'm not entirely sure i understand fully the original post.
but anyway the blackjack scenario is a scenario where one can expect 'many' plays over ones playing career, this envelope promotion is likely relatively 'few' plays, no? that doesn't necessarily mean to not go for the highest ev with out regard for variance, in the case of the envelope promotion, but the two scenarios are different in that regard. so but maybe ev is ev, regardless the scenario so really nothing wrong with going after the higher ev regardless the variance. still, maybe a distinction still holds, in that if this is a 'rare' play and you have this certainty of a given payoff as opposed to the possibility of a lower payoff, for this rare event.
so, i don't really understand certainty equivalent very well, but those arguing CE seem to have a point. imagine you are wandering, lost in the desert, you have a chance for a life sustaining drink of water, or more water than you really need, or alas, not enough water. but of all those choices the life sustaining choice is the only sure one, which would you take, while at the same time you know that if you live you have a decent chance of finding gold nuggets out there in the desert and more water later?:rolleyes:
When deciding what to bet, all I consider is what is within my risk of ruin (usually determined by half Kelly and ignoring heat). Even if I am only going to play this game for a short time (to a level way below my N0 say) this will not stop me from playing the game or betting below half Kelly. My bankroll does not care which game I am playing, it is just really concerned with my Risk of Ruin (ie. The relationship between EV and variance).

For example, if the situation calls for it, I am still very happy to count cards even though I don’t put in nearly enough hours to reach my N0 within the next few years. I welcome the extra EV without putting undue stress on my bankroll.

For the purposes of the envelope example, the same principle applies.

In the case of Sage’s desert example, a potential outcome is 0 (ie. You die). Kelly would not allow you to take this risk, as it never allows to risk your entire bankroll on one bet (unless it’s a lock bet, when it’s not really a risk).
 

sagefr0g

Well-Known Member
#28
Percy said:
When deciding what to bet, all I consider is what is within my risk of ruin (usually determined by half Kelly and ignoring heat). Even if I am only going to play this game for a short time (to a level way below my N0 say) this will not stop me from playing the game or betting below half Kelly. My bankroll does not care which game I am playing, it is just really concerned with my Risk of Ruin (ie. The relationship between EV and variance).

For example, if the situation calls for it, I am still very happy to count cards even though I don’t put in nearly enough hours to reach my N0 within the next few years. I welcome the extra EV without putting undue stress on my bankroll.

For the purposes of the envelope example, the same principle applies.

In the case of Sage’s desert example, a potential outcome is 0 (ie. You die). Kelly would not allow you to take this risk, as it never allows to risk your entire bankroll on one bet (unless it’s a lock bet, when it’s not really a risk).
just me maybe, my guess is you have the most optimal point of view on the matter. like you said your bankroll doesn't care what game it is. so lol, i guess the bankroll is just saying "feed me, feed me, dammitt, gimme plays!"
so at the very least, i imagine your tact yields a greater rate of bankroll growth over time.
still and maybe this is meaningless, there are caveats that come to mind here. well, me, i don't fully understand Kelly stuff, but i do know, that at least as far as the stories around here go, not everyone plays full Kelly.
so i wonder why, lol, i don't really know, lmao.... is it cause full Kelly is just too much for a salient being to bear, or is it a real danger for real world circumstances that Kelly theory doesn't protect us from?
whatever, i think Kelly is a utility theory, and as little as i know about the significance of that, i think it means there is an element of subjective wiggle room implied with respect to risk tolerance.
now, just me maybe, but risk isn't so simple a matter as just maybe losing one's bankroll or a matter of one's long term prospects. i mean heck, does the short term mean anything? is yielding $166 instead of $1,500 certain a risk or you can expect long term $1,650 on maybe a one time shot. probably a lot of AP's would say no, but i wonder how they know what outside influences on their bankroll, tomorrow will bring.
or also there is a question of value. like, consider the classic sweat shop factory compared to some factory that's had union influence or some kind of humanitarian influence. which factory produces the most value, how do you calculate the humanitarian value? which scenario has the greatest long term risk?..... or say you have that long, long, long drive home in the middle of the night, you ended up with errhh whatever it was $166, $1,650 or $1,500, what's the value of one's psychological state on the long drive home?
another caveat, like i said, i dunno if i even fully understand the original post, heck i dunno am making this promo play instead of some other play of greater or lesser value?
but the really point is, real world, just me maybe, i make some plays that i don't really understand fully what the heck is going on. plays that have assumptions, sorta thing, assumptions one hopes is correct. or maybe you aren't sure the ev and you have two opposing plays having such uncertainty, or varying 'environmental' conditions, problems to over come. plays of varying size of ev, plays of unknown potential fluctuation or variance.
so yeah that certainty equivalent idea, whatever that is, there is a point to it, there is some significance as to who one is, what ones financial position is, too like hey, bojack, or machinist, or creeping panther or mazz or whom ever, i'm not. some plays just me, i won't make, lol.
but from a practical point of view Percy, imho you are absolutely correct.:laugh:
 
#29
its not all EV

A casino has a little party/contest
It takes an hr & you get $20.
If you play bj for an hr your EV is $25
According to CE you take the party, all other things being equal.

So in the OP the sure $1500 seems high compared to $1666.
 

Lonesome Gambler

Well-Known Member
#33
Who cares about CE and small sample sizes in instances like this? As discussed on the show, the promo is only available to players putting up a certain amount of action, so risk shouldn't be much of an issue here. With risk out of the picture, I just can't imagine why anyone would care about a "sure thing?"

I play several games that are unlikely to net me over a few hundred trials each over the course of a lifetime. These are high-edge, one-off opportunities. The variance is heavily skewed by several factors. However, I don't feel that I'm wrong to want to get the money in any time that I have an edge and risk is not an issue.

Know you're getting an Ace? Bet as much as you can afford (keeping risk into account)! Gamble with a higher EV, or take the "sure thing?" Gamble! Someone will give you 2:1 on a fair coin toss, but only for one toss? Take the bet!

I'm sure that I'm oversimplifying the issue quite a bit, but really, you've got to take all the edges you can get these days.
 
#34
Alter of EV

Lonesome Gambler said:
Who cares about CE and small sample sizes in instances like this? As discussed on the show, the promo is only available to players putting up a certain amount of action, so risk shouldn't be much of an issue here. With risk out of the picture, I just can't imagine why anyone would care about a "sure thing?"

I play several games that are unlikely to net me over a few hundred trials each over the course of a lifetime. These are high-edge, one-off opportunities. The variance is heavily skewed by several factors. However, I don't feel that I'm wrong to want to get the money in any time that I have an edge and risk is not an issue.

Know you're getting an Ace? Bet as much as you can afford (keeping risk into account)! Gamble with a higher EV, or take the "sure thing?" Gamble! Someone will give you 2:1 on a fair coin toss, but only for one toss? Take the bet!

I'm sure that I'm oversimplifying the issue quite a bit, but really, you've got to take all the edges you can get these days.
So SCORE, CE & N0 do not matter?
 

Gamblor

Well-Known Member
#36
I'm hitting some -var right now so I'd take the $1500! Ask me another time I might not :grin:

I don't think its so straightforward as just purely EV considerations. Variance is just as important aspect to APs as EV. Why do some people take insurance on a BJ at a +1 or +2 TC?
 

Machinist

Well-Known Member
#37
I guess it's all about bankroll and time issues.
First thing i would be thinking about is how the heck do i get to partake in more than one of these drawings. How many shots can i potentially get at this promo. I like lots of shots....lots!!!!
If its a one time deal for only 1500 or so,...eh take the cash and move on. I dont like to play games, jump through hoops, dont like people getting a good look at me for any length of time. Are they gonna parade me around on a stage??? Forget it...
The numbers are kinda crappy for a promotion.........1500 is so close to the average of 1666, just grab the 15 and run. Now if i can do this alot of times well sure , play the game.

Machinist
 
#38
i resemble that remark?

WRX said:
We should just name this the Ploppy Detection Thread.
Deciding which play to take does not determine ploppy or AP status. There is a subjective answer to the problem.

A very dangerous AP may be clueless on how to determine the EV answer.
 
Top