Insurance

[moses]

Delete

[RS]

If you insure when there are less than 33.333% TENS remaining, it's -EV.

If you insure when there are exactly 33.333% TENS remaining, the EV = 0 (breakeven proposition, no edge either way).

If you insure when there are more than 33.333% TENS remaining, it's +EV.


Remember, insurance is simply a side bet that is strictly betting the dealer's hole card is a T. EV wise, it has nothing to do with what your hand is. If you have a big bet out, catch a 20 and dealer has an Ace, you might insure anyway to lock in a win (except the rare occasion when the dealer hits to 21 beating your 20)....even if the count doesn't call for insurance.....but making the insurance bet (in this case) is still -EV.

[Shoofly]

If you insure when there are less than 33.333% TENS remaining, it's -EV.

If you insure when there are exactly 33.333% TENS remaining, the EV = 0 (breakeven proposition, no edge either way).

If you insure when there are more than 33.333% TENS remaining, it's +EV.


Remember, insurance is simply a side bet that is strictly betting the dealer's hole card is a T. EV wise, it has nothing to do with what your hand is. If you have a big bet out, catch a 20 and dealer has an Ace, you might insure anyway to lock in a win (except the rare occasion when the dealer hits to 21 beating your 20)....even if the count doesn't call for insurance.....but making the insurance bet (in this case) is still -EV.

Your first three examples are exactly right. You should have stopped there. In the example outlined in the following paragraph, you are not locking in anything. As you said yourself, it is a completely separate bet. Take insurance if the count calls for it. Don't take it if it doesn't.

[SiMi]

Hi,

Not sure if this helps but...

Everything I've read states that insurance is +EV when the HiLo TC is >=+3. That's all there is to it. The name "insurance" is not very apt to me. You're not insuring anything. You're betting on whether the Dealer has a 10 down when she has an Ace up. That's it. It's just a side bet that has nothing to do with your hand.

For me, with a TC of +3, I have 4 units out. As the TC gets higher, I have more units out. So, my "insurance" side bets only occur when I have 4 units or more out. I don't know if you consider those "large" or not. I might do something else for cover once in a while but it's not the correct play from an EV stand point.

It is irrelevant to me what MY CARDS might be (except it is called "taking even money" if I ALSO have a Blackjack) since the bet has nothing to do with MY CARDS. The bet is about the DEALER'S CARDS. Whenever I put out a bet that is 4 units or larger, I KNOW if I see the Dealer draw an Ace up, to start getting my insurance bet out there since they often don't give you enough time. (They nearly all assume no one is going to take insurance until you've done it a couple of times and then most of them will slow down.) And, if I have a "large" bet out, it means there are more ACES and 10's coming out (HiLo) so I know the odds go up that someone, including the Dealer, is going to get a Blackjack.

This is the point in my game where I'm most alert because of the greater stress from having higher bets out, calculating the amount of the "insurance" side bet, getting the bet out fast enough and the greater likelihood of the Dealer getting Blackjack. It means I have to be ready with an "insurance" side bet and it means I also have to remember to count her 10 card (if the Dealer has it) because the hand is cut short abruptly, everyone whines and groans, and the cards will be scooped up more quickly than normal. It's easy to forget to count her 10 down card with all the commotion.

A lot is going on when high bets are out! The real bummer is when the TC is high, the Dealer has a 10 card up and checks for Blackjack and SHE HAS IT! Then, you don't even get a chance to place the side bet and your bigger bets are just gone (unless you also have a BJ). Too bad they don't let you place the side bet if the Dealer has an Ace OR A 10 CARD up...

[Three]

Thanks for your input. The 10s represent 30.77% of the deck. I'm wondering where you come up with 33.333%? I insure when 10s increase to 40% of the deck composition.

Insurance pays 2:1 if there is a ten down. If there are twice as many other cards as there are tens (66.66 to 33.33% the insurance bet is 0 EV) if there are more tens it is plus EV. If there are less tens it is negative EV.

[DSchles]

You're completely misunderstanding this and are looking at it the wrong way. You've received excellent answers, but you're still telling us when YOU insure, when your question was: when do WE insure, and what criteria should go into the decision?

In fact, there are nuanced composition-dependent insurance indices that vary from values somewhat below 3.0 to somewhat above, depending on the actual two cards in your hand. But, that has absolutely nothing to do with your method for deciding which hands to insure, which is quite erroneous.

The 33.33% proportion of tens comes not from how many there are in a full pack (4/13 = 30.77%), but rather what proportion there needs to be, after cards have been dealt out, to make insurance a profitable wager. Since the bet pays 2:1, you would need at least 1/3 of the remaining cards to be tens to have an advantage on the insurance wager. And, yet, again, what you hold in your hand at the moment makes virtually no difference whatsoever.

Clear?

Don

[DSchles]

You're being silly. You have index numbers for a variety of plays that you make: stand, double, split, surrender. And, you have one for insurance: >=+3, you insure; <+3, you don't. What's to have a weak point in your game? Makes no sense. You're waaaay overthinking this.

Don

[mofungoo]

You're being silly. You have index numbers for a variety of plays that you make: stand, double, split, surrender. And, you have one for insurance: >=+3, you insure; <+3, you don't. What's to have a weak point in your game? Makes no sense. You're waaaay overthinking this.

Don

Grosjean covered this subject in a chapter of The Book titled Expectation Isn't Everything. What the basic number for insurance does is to maximize expectation. There is also variance to consider. One can give up a small bit of expectation for a large reduction in variance. This can be done by insuring "good hands" sooner than the number says to insure, and waiting for a slightly higher count than the insurance number to insure weaker hands. See The Book for details.

[Sage]

I have been reading "Risk & Reward, The Science of Casino Blackjack" By Werthammer,
He states the Hi-Lo insurance positive threshold is around +4, but only near the start of the shoe and rises as the depth increases. At deep penetration it is a -EV at +4. He suggests a player use +4.5 to +5. Then sums it up: "Overall insurance does not represent much incremental value and is lot less than claimed: The weekend Hi-Lo counter, who plays about 10 hours on a trip, can disregard it with very little loss or accept the offer at +5. The BJ pro, who plays every day, gets some value from the insurance bet.

[SiMi]

Sage
I just ran 7 sims in BJ678 for my normal game and took insurance at various times. Here's the returns with all other factors held the same:

TAKE INS AT RETURN
Never--------------->1.11%
Only when
I had BJ------------>1.11%
TC = +3----------->1.21%
TC = +4----------->1.22
TC = +5----------->1.23
TC = +6----------->1.20
TC = +7----------->1.18

Maybe I screwed something up but this seems to agree with what you read...

[DSchles]

Please do us all a favor and NEVER quote Werthamer here again. We'll all be better off for it.

Don

[mofungoo]

Why not?

[ohbehave]

This I didn't consider. Perhaps I was tripping on the word "insurance". Like a new vehicle (good hand) is worth insurance. But a clunker (bad hand) is not.

Exactly why they named it insurance, its what the industry wants the average gambler to believe. And why it looks like an advantage play to insure a bad hand.