22: You've just pushed elevnty-bajillion times

[ QUOTE ]
Ah, no, I didn't mean that. I mean a 0.4% +EV is definitely good enough at this spot, I'd usually take +0.2% or perhaps +0.3% (it depends on the table) here. It's just the argument of how much he's pushed lately and that +0.4% is probably the "best" he can hope for here, sorry that I wasn't making myself clear on that point.

[/ QUOTE ]
I think that we all have to remember that the numbers we get with SNGPT are only guesses. Switching calling ranges or blind discounts just a little often change results dramatically. When we're using a minimum edge of +0.5%, it's not because we want to pass up on other positive EV plays, but rather that when they get closer to 0 EV we're less confident that they're actually plus.

Sort of like when someone says they're a winner because of their 4% ROI. Well, we're not really so sure.

[IdiotVig]

[ QUOTE ]
I think that we all have to remember that the numbers we get with SNGPT are only guesses. Switching calling ranges or blind discounts just a little often change results dramatically. When we're using a minimum edge of +0.5%, it's not because we want to pass up on other positive EV plays, but rather that when they get closer to 0 EV we're less confident that they're actually plus.

Sort of like when someone says they're a winner because of their 4% ROI. Well, we're not really so sure.

[/ QUOTE ]

Well, I was quite proud of my 4% ROI. Snob.

This was a really great hand to play around with in SNGPT. I played around using fairly wide calling ranges for the short stacks (~20-60%), and tighter ranges for the big stacks (5-15%), and was really amazed how small tweaks can drastically change the outcome.

But an even more important moral of the story is that even if there were an exact way of knowing this is marginally +EV, it's okay to take a pass on a hand like this based on the image I've established. Which was kinda cool.

In the actual hand, I mucked. w00t.

[ QUOTE ]
[ QUOTE ]
Ah, no, I didn't mean that. I mean a 0.4% +EV is definitely good enough at this spot, I'd usually take +0.2% or perhaps +0.3% (it depends on the table) here. It's just the argument of how much he's pushed lately and that +0.4% is probably the "best" he can hope for here, sorry that I wasn't making myself clear on that point.

[/ QUOTE ]
I think that we all have to remember that the numbers we get with SNGPT are only guesses. Switching calling ranges or blind discounts just a little often change results dramatically. When we're using a minimum edge of +0.5%, it's not because we want to pass up on other positive EV plays, but rather that when they get closer to 0 EV we're less confident that they're actually plus.

Sort of like when someone says they're a winner because of their 4% ROI. Well, we're not really so sure.

[/ QUOTE ]

Agreed. However, tools like SnGPT are really powerful and, basically, we should listen to what they have to say. If SnGPT says a play is +.02%, and you think villain's could be on either wider or looser calling ranges, then that means (very, very roughly) the play's 50% to be slightly better than +.02% and 50% to be slightly worse than +.02%, still making the play correct.

Tiger's point is, I think, that this specific hand was only +.04% with a little bit of optimism. He was arguing that this optimism added more than +.04% to the actual number, so the play was actually -EV.

If, for example, a push is +.01% even if you put your opponents on calling with any two (and >.01% if they call with less), and you expect them to call with any two, you should definitely push because there's no possible way your play could be incorrect.

[tigerite]

Yep, that is my point indeed. The interesting point, I think, is what -0.x% or even -1.x% we need to take in certain situations, when we are desperate, losing FE, blinds about to go up and hit us, against 2nd shorty on the bubble, and also the slightly more obscure ones like a win would let us dominate the bubble, or prevent loss of EV in future because people are less prone to push into our BB, and so forth.

Obviously another would be 'future image' but you can't possibly expect a model to work on that - the others are at least slightly within the realms of reality..

I'm glad I bugged you about this. Turned out really well. Nice posts.

[tigerite]

No problem. I have to say actually though, that even if someone uses a default of wanting 0.4% EV or whatever, they probably aren't losing any discernable huge amount of $EV, if it's just a quick and easy answer that they want. Sure you might miss the odd one where, like you say, an any two call is 0.1%, but for 0.4% that any two would no doubt be 80%+ anyway. So it's not horribly wrong to want a default % if it makes someone feel more "comfortable" with a move.. maybe lose the few odd cents here and there.

[ QUOTE ]
No problem. I have to say actually though, that even if someone uses a default of wanting 0.4% EV or whatever, they probably aren't losing any discernable huge amount of $EV, if it's just a quick and easy answer that they want. Sure you might miss the odd one where, like you say, an any two call is 0.1%, but for 0.4% that any two would no doubt be 80%+ anyway. So it's not horribly wrong to want a default % if it makes someone feel more "comfortable" with a move.. maybe lose the few odd cents here and there.

[/ QUOTE ]

Meh.. I think we should suggest the play that earns a player the most money in the long run on these forums unless someone asks very specifically for something different.

I can tell you very easily how much money you're losing if you pass up a +.03% of the prize pool play if you just tell me the prize pool [img]/images/graemlins/smile.gif[/img].