When Insurance is a "good" bet?

anyone know how to determine what percentage of shoes would yield a deck in which Insurance becomes advantageous to the player? thanks for the help.

Percentage of shoes?

The bet is never in the player's favor unless the player is counting cards, and there are a lot of ten value cards remaining. How many is a lot? At least 1/3 of the cards must be tens.


Unless you're counting cards, never take insurance, except:

If you're playing a single deck game, and the dealer has an ace on the second hand, and the table is full (or near full), and it occurs to you that you didn't see any ten value cards in anyone's hand in the first hand, then that is a time to take insurance.

Of course, the observance of those conditions is kinda like counting.

The general rule of thumb is a true count of +3 right?

i guess to rephrase the question. what is the likelyhood that a true count becomes excessive enough to give the player an edge?

The true count represents the ratio of big cards (10s) to small counts (3-6s). Theoretically, insurance becomes profitable when the ratio of 10s to other cards in the deck is > 1/3

At zero, there are 4/13 tens, or 30.7982%
-- At a true count of +1, there is one more ten per deck than small card. That is: 16.5/15.5 with 20 unknowns. 16.5/52 <1/3. Do not take insurance.
-- At a true count of +2, there is two more tens per deck than small cards. That is 17/15 with 20 unknowns. 17/52 < 1/3. Do not take insurance.
-- At a true count of +3, there are three more tens per deck than small cards. That is 17.5/15 with 20 unknowns. 17.5/52 > 1/3: take insurance.

Most websites concur that you should take insurance at a count of +3 or more.

OP, the true count will be +3 or higher about 10 percent of the time on a 6-deck game with 75 percent penetration. If penetration is poorer, the result will be considerably less. Most high true counts occur at the end of a shoe.

There are a couple sources I checked on this that showed slighly conflicting percentages, but both of their simulations showed something very close to 10 percent in a game with decent penetration.

are these sources available online?

Yep, the first is from blackjack in color, the go-to resource on these matters.
http://www.blackjackincolor.com/truecount1.htm

You should explore the rest of the site, the charts and explanations on the subject of blackjack and card counting are truly excellent.

I also remembered that the blackjack school had a chart of the exact info you requested. Scroll down to the 'effect of penetration' section:
http://www.bjrnet.com/GameMasters/GameMasterClassics11.html