Doing business - Is this enough overlay?

[phydaux]

I play NL cash games, mostly on the internet now but I'm looking to playing live regularly. And I'm thinking about regularly offering to "do business." The two things I'm thinking of doing are offering insurance and "buying the dog."

What I'm wondering is is 10% overlay enough, or should I ask for more?

For example, if someone is all in, they turn their hands over and the favorite is 80% to win. I would offer insurance at 22% of the asked amount. If they were 90% to win I would offer insurance at 11% of the asked amount. So I would be getting 10% more than the correct odds.

I'm also thinking of offering to buy the dog hand. If the had is a 20% dog, then I would offer the player 18% of the pot to buy their hand. If they were a 90% dog, I would offer 9% of the pot. Again, a 10% overlay for me.

I'm thinking with "just" a 10% overlay there would be massive swings and I would need a huge BR in addition to my regular poker BR. But if I try to squeeze out a 15% or 20% overlay then I'll never get any action.

Thoughts?

[Dromar]

[ QUOTE ]

I'm thinking with "just" a 10% overlay there would be massive swings and I would need a huge BR in addition to my regular poker BR. But if I try to squeeze out a 15% or 20% overlay then I'll never get any action.

Thoughts?

[/ QUOTE ]

I think you just answered your question there.

Seriously, I'd suggest not mentioning any %ages at the table. But when the hands flip over, if you can offer the guy an amount of money that's profitable to you, and sounds like an okay deal to him, then go ahead. You could get a lot of business from the Phil Hellmuths at your table. It'd probably be okay to mention %ages to them, since "if there weren't luck involved, they'd be rich" etc.

Sounds okay, and I might actually start doing this myself. But of course, you'd need a stable bankroll as you're already aware.

[binions]

No, it's not enough overlay. You will go on bad streaks and get killed.

In a typical hold'em insurance chart, the house has a much larger edge than you propose.

1 out = 20:1
2 outs = 12:1
3 outs = 8:1
4 outs = 6:1
5 outs = 5:1
6 outs = 4:1
8 outs = 3:1
9 outs = 2.75:1
10 outs = 2.5:1
11 outs = 2.25:1
12 outs = 2:1
and so on until 22 or 24 outs is even money.

If all in on the flop, outs are doubled. 3 outs twice is 6 outs on the insurance chart. If all in on the flop and the turn brings more outs, the player has to reinsure. Otherwise, he could lose the pot and the insurance.

For example, if AK has AQ dominated on a A high flop, AK can insure 6 outs (3 queens twice). Lets say he takes 400 to 100. If the turn brings AQ a flush draw and he does not reinsure, then he owes the house 100 if the flush hits plus he loses the pot.

As you can see, the player insuring his hand is giving up a lot more of his edge than you propose taking. When someone truly only has 3 outs twice, you are ~87% to win. But insuring at 4:1, you only receive 80% of what you insure. You give up ~7% of your edge. You'd be surprised, but people insure their hands all the time in Houston desite the hefty premuim.

If the players tie, insurance is off. When there are a bunch of tie outs, true odds get a lot closer to the insurance chart.

I have never seen it done, but a better way to do insurance with all the poker equity calculators is to input the hand into a computer and figure the true percentages, then take 5% away from the player in the lead and give it to the player that needs to suck out.

For example, let's say you input a hand, and with 2 cards to come, player A is 75% to win, player B is 20% to win, and they are each 2.5% to tie.

Insurance is off in a tie. So the true odds are 75:20. Charge your premium, and you would offer to insure at 70:25.

This way, there would be no need to reinsure the river if the turn brought more outs since that was part of the 75-20-2.5-2.5 calculation.