Insurance is the shiiit

[ThodorisK]

I just saw Griffin's "The theory of blackjack" page 61 table. At this table it says (for 1 deck):

(unseen cards) (insurance gain) (non-insurance gain)
(8-12) (0.44%) (3.11%)
..............
(43-47) (0.4%) (0.6%)

This means that the lower the true counts, the higher the fraction: (insurance gain/overall gain)?
Thus for 6 and 8 decks the fraction (insurance gain / overall gain) is close to 40%? I start doubting the effeciency of the halves counting system now But on the other hand, the true count of +3 or higher where insurance is favourable, happens much less frequently than the true counts of +1 up to +3, and this should reduce the fraction (insurance gain/overall gain).

 

[ThodorisK]

Hm, Griffin's values do not apply for wonging, lol.

 

[creeping panther]

Zen

Without regard for count or cards held? zg

I was just as surprised as you I have not had the pleasure of playing SD in several years....I will have to go through my reference library to gather more info but I know (the anomaly) had stuck with me through the years.

Actually there is also a few other things that differ from what BJ celebritys are saying these days

CP

 

[Warlord]

Voodoo. zg

vodoo.... or the reason why it is said to take insurance when the dealer has an Ace up first hand in SD. (at least according to CreepinPanther)

 

[Automatic Monkey]

Come on guys just use the insurance index for your count when making an insurance decision.

If you want to give yourself some cover, taking insurance whenever you have more than a minimum bet up only costs you a little bit, and you don't have to stop and think about the count.

 

[ThodorisK]

Coverplay is a stupid waste of money. Soon management knows you are counting, no matter how brilliant the coverplay is. Not only I ONLY wong (bet zero money when there is no edge), but also have quarels with other players because I open and close boxes.

 

[Warlord]

Coverplay is a stupid waste of money. Soon management knows you are counting, no matter how brilliant the coverplay is.

I do not think he meant "cover" in that way.

He meant cover from losing to a dealers natural; not covering your play with wrong moves as to camo your AP'ing.

 

[rukus]

I do not think he meant "cover" in that way.

He meant cover from losing to a dealers natural; not covering your play with wrong moves as to camo your AP'ing.

the term "cover" includes using both incorrect playing decisions as well as betting decisions to allay some of the heat from the pit.

what ThodorisK seems to not grasp in his infinite desire to disrespect some of the more knowledgeable/senior posters here is that for some of the posters here, cover is indeed worth its cost.

now obviously for a red chipper, not only is cover an easy way to negate any positive EV, it is not needed at that level.

there are several cover plays that when summed up cost a HIGH betting counter maybe $10 an hour in EV. if this "buys" a high limit counter even a single hour more of playing time then these cover plays are worth it. one hour EV could be $100 or $200 or $500 or who knows, of which that cost of cover is peanuts.

so Thodorisk, maybe it's time to actually start reading posts (a reference to another thread) and to think before insulting someone like Auto Monk by saying his suggestion is a "stupid waste of money".

respectfully,
rukus

 

[ThodorisK]

lol, it is that more likely that the pitbosses will let you stay longer if you make a few mistakes? I dont think so. First of all, they do not care how profecient counter you are. They just care whether you are counting or not. They will ban any sort of counters. I dont get the logic of giving them away money for the hope to stay longer. On the contrary, you should make the most out of your time because they will ban you soon. Thus you should not give them even one penny, and bet as much as possible. If the "heat" from pitbosses makes you giving away money, hey, that's their purspose. Actually, it is them who are on heat because they want to ban you but for some reason they are hesitating to do so yet. Who told you I am not experienced? I have been banned too.

 

[Warlord]

lol, it is that more likely that the pitbosses will let you stay longer if you make a few mistakes? I dont think so. First of all, they do not care how profecient counter you are. They just care whether you are counting or not. They will ban any sort of counters. I dont get the logic of giving them away money for the hope to stay longer. On the contrary, you should make the most out of your time because they will ban you soon. Thus you should not give them even one penny, and bet as much as possible. If the "heat" from pitbosses makes you giving away money, hey, that's their purspose. Actually, it is them who are on heat because they want to ban you but for some reason they are hesitating to do so yet. Who told you I am not experienced? I have been banned too.

WTF are u talking about? the context of insurance covering here is not as a misplayed hand but a pure EV move.

Now if you take insurance always regardless of the count (and you are a counter) then you are using the play as a camo/cover, and I think it would be the most costly "cover" play u could make ; especially if you are taking even money.

If anything taking insurance at different times is the opposite of making cover moves. Taking insurance increases your EV at the cost of possibly being noticed that you take insurance at seemingly random times (when count is +) that happen to be when your bet is large.

So if anything, being someone (a counter) that takes insurance at different times can lead to increased scrutiny. So I would say stuff like; "last time the dealer had it so I am insuring this time" when the count dictates you take insurance.

As opposed to some ploppies that take it always.

And I think Creeping panther is right in regards to taking insurance on if the dealers first hand off the top is an Ace in Single deck.
I have played hundreds of hands and more than 50% (probably been ~70%) of the time the dealer has had it.

 

[Kasi]

I was just as surprised as you I have not had the pleasure of playing SD in several years....I will have to go through my reference library to gather more info but I know (the anomaly) had stuck with me through the years.

Actually there is also a few other things that differ from what BJ celebritys are saying these daysCP

Well, maybe I'm not understanding exactly what you're saying but I'm just saying playing SD heads-up off the top of a new deck versus a dealer Ace never take insurance regardless of any count and no matter what your 2-card is comprised of.

I mean only 3 cards are seen and 49 are unseen. The ratio of 10's to remaining cards can never be 33% or more.

 

[EasyRhino]

You know, I'm not that big a fan of insurance, because to me, it increases variance, in two ways:

1) It is, itself a 2:1 bet, so the winning hands are only about 34% of the time.
2) When done in a high count, it ends up putting EVEN MORE money out on the table for one hand, all of which could be lost if things don't turn out well (like, the dealer has no natural, but you're dealt two stiffs).

That being said, it's still a +EV move, so you gotta take it when the count calls for it, regardless of how you feel about it.

 

[Automatic Monkey]

You know, I'm not that big a fan of insurance, because to me, it increases variance, in two ways:

1) It is, itself a 2:1 bet, so the winning hands are only about 34% of the time.
2) When done in a high count, it ends up putting EVEN MORE money out on the table for one hand, all of which could be lost if things don't turn out well (like, the dealer has no natural, but you're dealt two stiffs).

That being said, it's still a +EV move, so you gotta take it when the count calls for it, regardless of how you feel about it.

Yes, certainly when it's that kind of +EV. In a SD game with an ace neutral count we're talking +15%. It's like selling the casino keno tickets.

You just have to bite the bullet and accept the variance.

 

[ThodorisK]

I too at first thought it increases variance, because it's a ~1/3 to win bet. But is it? It can also be considered as a bet which increases the chance to win the hand much over 50%, as if you dont win the insurance, you may win your normal hand. So which of the 2 views is the correct one? I dont know!

 

[Automatic Monkey]

I too at first thought it increases variance, because it's a ~1/3 to win bet. But is it? It can also be considered as a bet which increases the chance to win the hand much over 50%, as if you dont win the insurance, you may win your normal hand. So which of the 2 views is the correct one? I dont know!

Don't think of it as having anything to do with your hand. It's a sidebet, and either it has +EV or it doesn't. because it's not an even money bet it has higher variance involved than a mostly even money bet like a blackjack hand, but it's really not that bad at 2 to 1, much better than most sidebets.

 

[ThodorisK]

The overall win/loss probability of the hand depends on both these 2 bets (normal bet and insurance), thus it is not clear that these 2 bets are independent regarding the variance of bankroll, and therefore it is not clear whether insurance increases or decreases the variance of bankroll. If you can, PROOVE that these two bets are independent regarding the variance of bankroll. I don't see such a self-evident proof.

 

[k_c]

The overall win/loss probability of the hand depends on both these 2 bets (normal bet and insurance), thus it is not clear that these 2 bets are independent regarding the variance of bankroll, and therefore it is not clear whether insurance increases or decreases the variance of bankroll. If you can, PROOVE that these two bets are independent regarding the variance of bankroll. I don't see such a self-evident proof.

let p10 = probability of drawing a 10
let EV = expected value of hand for bet of 1 unit if no insurance is taken
let roundEV = expected value of round for units bet if no insurance is taken
let roundBet = amount bet before hand is dealt
let insEV = expected value of taking insurance
let insBet = amount bet on insurance (max = roundbet / 2)

letTotEV = expected value of round including insurance

insEV = (3 * p10 - 1) * insBet
roundEV = EV * roundBet

if insurance is taken:
totEV = roundEV + insEV

if insurance isn't taken:
insBet = 0 and insEV = (3 * p10 - 1) * 0 = 0
totEV = roundEV + 0 = roundEV

It should be obvious that insEV is completely independent of roundEV (i.e. it is a side bet.) totEV is changed only by what is bet on the insurance side bet. roundEV does not change whether or not insurance is taken.

 

[ThodorisK]

Who said about EV? I was talking about the variance of the bankroll.

 

[ThodorisK]

I am pretty certain now that I have thought on it. Placing insurance is decreasing variance instead of increasing it.

 

[Kasi]

Who said about EV? I was talking about the variance of the bankroll.

It seemed like you did when you say "The overall win/loss probability of the hand depends on both these 2 bets (normal bet and insurance), thus it is not clear that these 2 bets are independent regarding the variance of bankroll," and link the probability of winning to variance.

Almost anything that increases EV increases variance. Better rules, better pen, higher spreads, wonging.

Variance has nothing to do with edge per se as far as I know. It's just bet size squared times the frequency of that bet?

If you're putting out more money at +3, generally speaking, you've increased variance compared to if you haven't I'd say.