ICM question - what +EV% should you use on SNGEGT's?

[campfirewest]

I've been building a spreadsheet of various scenarios varying the number of players, blinds, chip stack sizes. I've been primarily using this website that someone here pointed out to me: http://www.holdemresources.net/hr/sn...alculator.html This will calculate push-call-fold ranges for the +0EV breakpoint. I've also been using SNGEGT (I purchased the $100 version) Using SNGEGT you can put in the +%EV gain you want and it will calcuate the ranges accordingly. It defaults to +.5% EV. This makes sense because if you make decisions at the 0 EV point you're basically coinflipping. And against bad players why shove when you're +.1%EV and risk going broke when you know they'll be making a lot of mistakes and you'll get better EV situations later?

So after that long winded intro, my actual question is: What is a good min %EV to use when calculating push-call-fold ranges for low limit ($<20) sng's?

There is often a huge difference in the percentage of hands to play depending on whether you use 0%EV and +.5%EV, as much as 45%!!! So what I want to do is add columns to my spreadsheet to include push-call-fold %'s depending on various EV %'s. I'm thinking against strong players you want to play hand ranges that correspond to a lower min +EV%. Then the weaker the players the higher the +EV% because better opportunities will present themselves.

[Cadaver]

- buy-in (ok lowlimit.. so you have to pick your spots.. obv not abusing every +ev situation)
- players left
- presence of uberdonks\very solid players

and..
sngegt and all other "realtime" tools - BS. review your HHs.

[campfirewest]

I'm not using the realtime feature, I bought SGNEGT for the game theory feature that calculates the push-call equilibrium points. I've been comparing them to what the website in my last post predicts. They're not quite the same but they're close. I think some of the difference is because the 2 programs weight the hand values a little bit differently. What I'm trying to do is get to where I have a very accurate 'cheat sheet'. I review hand histories all the time, the problem is after studying a particular hand and learning something, I won't necessarily remember it a year later. Hence writing everything down and building a giant spreadsheet that I can review and study whenever I want. I think sng's are simple enough that this is feasible.

[HajiShirazu]

I don't like having any % at all unless you are playing at low buyins. IMO if you start passing these up you aren't going to have as big an edge as you think. Most people who have at least some clue what to do with shorter stacks aren't going to be missing many pushes with a wide edge but won't consider many of the more marginal shoves. This is one of the things you have to do better to beat tough games.
More importantly your overall "skill edge" over the rest of the tournament is usually going to go up, possibly quite a bit, if you double up or even just add to your stack in many situations. The equity "edge" calculated from the ICM stacksizes before/after the hand is like an absolute minimum edge that you will gain from making the shove. In reality your equity will usually go up just a tiny bit more than what ICM calculates in most situations because you will almost always do better on future hands simply due to the fact that you will have more opportunities with a bigger stack (although there are rare cases such as with microstacks about to take the blinds where this is not true.) IMO it is (usually) better to make a shove that is just barely ICM wrong than to pass up a shove that is just barely right.

[plexiq]

[ QUOTE ]
I'm not using the realtime feature, I bought SGNEGT for the game theory feature that calculates the push-call equilibrium points. I've been comparing them to what the website in my last post predicts. They're not quite the same but they're close. I think some of the difference is because the 2 programs weight the hand values a little bit differently.

[/ QUOTE ]

Actually, the difference should be pretty drastic in some cases - as SNGEGT does not consider overcalls in their GT approach.