Ed Tice: Back to simulating hand interaction for insurance

[Ed Tice]

I know I exchanged a few posts earlier with the famous NW, but I want to revisit the idea of how much we can gain by getting others to put up insurance for us when its a smart bet.

I have read that the presence of insurance is worth about 0.40% if you are counting cards. So this would initially appear that just by making proper insurance bets you could break even just flat betting. Of course this isn't right because if you are flat betting those insurance opportunities aren't worthless. Assuming 1:8 spread I imagine that your average bet size when an insurance opportunity turns up is about 4 units. Flat betting one unit, you could only recopu 0.1% because you just don't have the $ on the table.

But what if I could get three other people to put up insurance bets for me? Then I would be able to flat bet and play even just by taking insurance.

But maybe I can do better?

If I used the insurance count defined in PBJ (by SW), I'd have perfect insurance correlation. So maybe I could get 0.15% back. If I could get just two others to put up insurance bets for me, I could play even with flat bets. Pick up a few more wins on 'hand interaction' and things are looking good for me.

So I tried to simulate this using CVData (I try to put it to good use, because its expensive!)

So I plugged in the exact rules for the Showboat in Atlantic City (casino chosen at random), selectd complete basic strategy and flat betting, but no insurance. I ran a simulatin and had an IBA of -0.45%. This is what I expect.

So I went back to the rules screen and enabled insurance. Then I went to the playing screen and enabled the insurance side count. I ran the simulation again. Still a -0.45% IBA!

So I thought that maybe the insurance side count feature was acting funny. So I defined a new playing strategy and keyed in the appropriate count (-2 for 10s +1 for all other cards).

I then selected the "insurance decks strategy" and set it at 4 for one deck... 32 for 8 decks (are these the right parameters)?

But I still have the same -0.45% IBA. With proper insurance this simply can't be true, can it?

Has anybody else tried this simulation? Do I have something wrong?

Does anyody have a statistic on the affect on my IBA if the ONLY count I were to keep were the 10 count for insurance and play it correctly (flat betting no other hand interactions or anything?)

Ed

[Don Schlesinger]

You are vastly overrating the value of insurance.

See BJA, p. 70, where, for the 1-8 spreader, hi-lo insurance was worth only 0.117%. I don't know where you got your 0.40% from, but it's much too high. Flat betting, using all I18 indices, you still can't get an edge over the house!

Obviously, placing three or four insurance bets, instead of just one, might help, but there are surely easier ways to make money.

Don