SNG All in EV

[CougarBait]

Hello,

I have spent about 2 hours googling this and found a program by adon.albert called tourneyluck which doesn't seem to work anymore.

I understand PT3 has all in EV for sngs if you select hands played then either chips won or BB won. How accurate is this feature?

Looking at my graph I am 200bbs under EV over 10k hands, i'm not sure if this is a big deal or not but basically should I be placing more weight upon my results or the all in EV of BB won?

I understand if I am playing cash games then I should choose to look at AIEV as my true winrate rather than my actual winrate. Can the same be said for sngs or is the EV of BB won measurement not as good as results?

[kraada]

I personally believe all in EV is highly overrated, especially for tournaments - I just don't personally believe as a measurement it tells us anything that's particularly useful. That said, I understand that it's a very important number for many people and our calculation method is completely accurate.

For a more detailed treatment of how we perform the EV calculations, please see this tutorial.

[CougarBait]

Well I understand it's not a perfect measurement, I just wanted to know if it can be compared to AIEV at cash or whether it is somehow distorted in sngs/tournaments.

It is clearly more useful than results but obviously no where near 100% fullproof or conclusive.

[kraada]

Well, for tournaments you're always going to have skew based on tournament payout issues - you'd need to use something like the ICM model to account for such a thing (which PT3 does not do at this point in time).

Think about it like this.

Imagine you have two situations where you are exactly 50% to win (I know it's rare to be exactly 50% but work with me on the math :)). You win one, and you lose one. In both cases the pot was the same size - say 20BB (and you and your opponent each contributed 10BB).

In this case, you are running exactly at EV - you would expect to win one of the two coin flips. If it were cash that would be the end of the story, but it isn't for tournaments. Say one of the coin flips was for first place in a huge MTT - if you win that flip you move up $150,000 in prize money. The other coin flip was in a single table SnG - winning it moves you up $20. You won one and lost one - but which one? Clearly you're a lot luckier (and richer!) if you win the $150,000 coin flip.

This is just one of the reasons why all-in EV is overrated in my opinion - I'd rather look at each hand individually and try and ascertain if I made the best decision I could. As long as I do that and can do it consistently, the results will take care of themselves.

[CougarBait]

Yes I should have made clear I was talking about 1 table sngs all of the same stakes.

As you mentioned though this does not take ICM into account.

[kraada]

Even on SnGs all of the same size, some flips could be early, some could be on the bubble, etc. It's a very rough guide at best.

[lolfish]

sorry to bump this thread, but I'm a bit confused about this whole allin ev thing. Isn't it so that it's just for chips won instead of $ won like in cash games? so the EV line for sngs would not factor in the $ payout in sngs/mtts?

I might be dead wrong here, but since I'm running way under ev (like >600K in 90K hands) I'm looking into this just to check if this is normal or that is just way off.

[kraada]

Honestly I wouldn't worry about it that much - since whenever you bust you lose more chips than when you successfully survive (until you win the tournament) in the long run you will end up negative in chip EV. Think about it like this:

Flip for 100 chips: WIN, flip for 200 chips: WIN, flip for 400 chips: LOSE.

Assuming those are complete flips (50% equity) you'd be "due" 50 + 100 + 200 = 350 chips. You won 300 - so you're below EV. But you actually won 2 flips and lost 1 so you're doing better than expected that way.

It's just not a very good measurement for tournament poker. PT4 will include ICM calculations which are much more useful for players like yourself.

[lolfish]

I thought about that as well, I know my ev line is going to be off a little (or a lot) because I could win 2 flips early on for say 3K chips and then loose a flip in my endgame for say 10K chips, so that calculation is going to be off obviously, I just didn't expect it to be this much.

I'm not staring myself blind on my ev line as doing so is -EV already lol. Was just wondering if it's that accurate for sngs.

I'd love to see how that ICM stuff works in PT4, beta version should be released soon right?

[WhiteRider]

The PT4 beta should be within the next few months but we don't have an exact date yet - it could be anything from a few weeks to a few months: we won't release it till it is ready.