# Cover + Risk Aversion w/ Insurance Betting?

#### zengrifter

##### Banned
Cover + Risk Aversion (variance reduction) w/ Insurance Betting

When you are slightly below the insurance threshold feel free to take
partial to full insurance on 8,9,10,11,19,20,21.

When you are at the threshold or slightly above it insure
5,6,7,14,15,16,17,18 for less.

Well below the threshold always
insure 21 and 'occaisionally' insure 10,11,19,20.

Can someone refine or otherwise dismiss this
approach altogether?

zg

#### The Mayor

##### Well-Known Member
On RA insurance bets

I agree that the insurance bet is one that sticks out the most for the pit and other "eyes". When you insure your 16, but don't insure your T-T, they know what you're doing.

That being said, I saw a guy put down the table max at single deck and insure his 16, and a few hands later put down the table min. I think act goes a lot further than RA indices and careful planning.

If you always take insurance, then the times you are making the wrong play you have a small bet anyways, and when you are making the right play, you have a big bet. This is called "counter basic strategy" and is often used for cover.

There are many plays in "counter basic strategy", like doubling 8 vs. 6, 11 vs. A, standing on 12 vs. 2 and 3, that if you always do them, it costs you very little, because the time you shouldn't be doing them you have a small bet anyway. The money you give back can be considered the cost of cover.

RA strategies are key. For those who don't know, RA (risk adverse) play means to make plays that greatly reduce variance with only a minor decrease in EV. Or conversely, have a large increase in EV with only a small gain in Variance.

However, I prefer to think of insurance as a separate side bet, wholly unassociated with the bet on the table. In this situation, if the count is very close (but just above) the index, then it may not be "kelly" justified to make the entire wager, and insure for less is correct. We would need to get a separate bet ramp for the insurance bets associated with the Crush count to get it pefect. I am sure this has all been done, maybe some reader knows much more and can say something.

--Mayor

#### zengrifter

##### Banned
Re: On RA insurance TWO INSTANCES

TWO INSTANCES - where I would be inclined to use some of this arcane RA-insurance tactics - 1) If I'm fighting for survival (tournament, shortstakes, etc.) or 2) If I have an excellent game/pene and I'm winning with "lite scrutiny" from the pit. zg
--------
Here is MathProfs reaponse to a similar original post last year -

Risk Aversion and
Insurance

Posted By: MathProf.
in response to: Cover +
Risk Aversion w/ Insurance Betting (The Grifter&reg

Partial insurance always reduces your variance. Now, if
you knew that insurance were a break-even proposition
(because 1/3 of the unseen cards are 10s), you would want
to try to reduce your variance.

Half-Insurance (which can be implemented by insuring one of two
spots) will reduce your Variance even on stiffs.
(Actually, about one-third insurance is optimal.)

If your choice on Insurance is all or nothing, then you
would take full insurance unless you have a stiff.
Over-insuring weak hands will actually raise your variance.

Clearly, with a big enough bet, there is some value in
trading off EV from insurance with risk aversion. You
could create different risk-averse indexes for
different types of hands.

But there is one danger.
If you are not keeping a 10-count, then there would
be slightly different insurance indexes for
different hand compositions. The HiLo index for 10-10
(which is the "best" hand to insure for risk-aversion)
is a actually somewhat higher, especially in
single-deck. The insurance index for 9-7 is actually lower.

I think cover considerations outweigh any fine
mathematical points in these insurance decisions, especially