Lot of bad advice in this thread IMO.

1. I find it ironic that people here are saying "you want to play these pots because your opponents don't know how to play kill pots". This can sometimes be OK poker reasoning, but for it to be, you need to be darned sure you know what you are doing first.

2. Posting kills are to be avoided where possible. Check your online databases-- what two positions did you lose money in, no matter how good a player you were? You got it- the small and big blinds. Posting blinds is incredibly expensive. A typical good player's winrate is a big bet an hour, or thereabouts. Well, a half kill is 3/4's of a standard big bet. It's a really expensive tax. (A full kill is a full big bet.) Yes, you get some of it back, but you also have to post another one every time you win.

3. Corrallary to 2, but where you really want to tighten up (not discussed in this thread) is when you have the leg up. There is very little benefit to winning small pots with the leg up that force you to post kills. If you can win a medium or big pot with the leg up, of course that is worth it, but bear that in mind. You get a huge non-reciprocal benefit from playing cautiously with the leg up while your opponents play their same loose game when they have a leg up.

4. In position, you probably should defend a lot of hands in the kill multi-way. But trash is still trash. Reverse implied odds are still a thing, even if they cost you less in position. And you still have to subtract the big blind you will be posting next hand (and in slightly worse position) if you win the kill pot.

5. In position, you should probably open wide if it is folded to you with the kill. For the same reason a poster should open wide if it is folded to her.

6. Out of position, you should be looking to fold your kill to raises. The only exception might be under the gun, where posting a kill on your big blind isn't that costly. But let's say you are UTG+1-- you are going to play a hand out of position, with crap cards and reverse implied odds, and if you win you get to post a kill UTG. You should be looking to fold. A lot.

7. People should be careful about saying things like "I'd like to post a kill on the button.". People say the same things about Mississippi straddles. But nobody's actually ever offered any mathematical proof that posting additional blinds on the button is profitable.

Posting blinds forces you to play crap hands that you would otherwise fold. It forces you to pay money for nothing when you are facing raises you can't call. Posting blinds costs money, lots of it. It's just highly unlikely that you will actually make a profit posting a kill on the button or near it. Now, will you lose less than with the UTG kill? Sure. But it's still likely to be a loser. There are just lots of hands you would like to fold, even on the button.

Great post.

Misery loves company, it seems.

We have to prove it by algebra? Accumulated PT/HEM data about posting doesn't count?

The "you have to post the kill again" argument overlooks the fact that the kill becomes a sunk cost the moment you win when you had a leg up. Posting the kill again doesn't mean losing another two small bets; it just means you are posting the same two bets as before. It's like playing in a game with a rock, i.e. a bundle of chips representing a preflop bet.

Where is this data? I've never seen any, or even heard it given as an argument before.

Leaving your snark from the other post aside, this is quite wrong.

Consider the following two situations:

1. A kill pot is like a blind. You have to post it, but you don't have to post it again if you win.

2. A kill pot works the way it actually works.

In scenario 1, you will of course not subtract 1BB from your expected value with respect to pot odds calculations and considerations about playing your hand, right? You will do the same calculations you would do with any other blind that you post.

So in scenario 2, the actual situation with kill pots, your expectation is 1BB (or a little less than 1BB if you are going to get to post the kill in good position next hand) less.

And that's because you got the sunk cost fallacy backwards. The kill is, in fact, a sunk cost, which means that when you post it again next hand, it doesn't matter that you posted it before in the last hand.

Suppose you win ten pots in a row in a kill game, and lose the 11th.

After the first pot you have one a pot at the small limit and have a leg up.

After the second pot you have won a second pot at the small limit and now must post a kill.

After the third pot you have won a pot at the large limit including the kill you have just posted, and must post the kill. . . .

After the tenth pot you have won a pot at the large limit including the kill you have just posted, and must post the kill.

After the eleventh pot, you have lost your kill, and any money you put voluntarily into the pot.

We have killed it nine times in a row. Have we lost nine small bets at the big limit because of this? No. We have only lost one.

The disincentive to win takes place in the pot when we have a leg up. Winning that pot incurs the obligation to kill. It is when we have a leg up that we must include the cost of killing the next pot when we figure pot odds.

Once we kill it the first time, the cost of killing is a sunk cost, and it should make no difference in how we play until the next time we have a leg up.

No, you haven't lost nine kill bets. But that's because you have gotten lucky and won 11 times in a row. Eventually you will lose and that point you lose the kill blind.

I mean, I posted the correct analysis. If you are right about the kill being a sunk cost that you only pay once, then the odds are exactly the same in a game where you don't have to post another kill if you win as they would be in a game where you do. And that is just logically wrong.

What is confusing you is that as long as you are on that winning streak, it feels like you are only paying the kill once and that you already paid it.

But that's not true, mathematically. Each of those hands you are being forced to post a blind, and that blind is costing you EV on the hand.

Here's another analogy. Let's say you are playing a home game and that due to some silly rule, they make you post a round of antes. You have to post on every hand in the orbit. And let's say that you then win all the hands on that orbit. So you get all your money back. Does that mean that the EV of posting those antes is zero? No it doesn't-- it just means you lucked out.