PDA

View Full Version : $EV won in tourneys (PokerStars)



AndrewK
06-22-2010, 05:05 AM
Why is at that at PokerStars my multi-table $EV won column has all the exact same figures as my $won column, but for single-table results the two columns are completely different. Is this a bug?

Also, I can't seem to find an explanation as to how $EV won is actually defined in a tournament context.

Any help would be much appreciated. Thanks.

_Loki_
06-22-2010, 07:35 AM
Hi Andrew

I can't specifically answer your questions, but the support guys will be along in a while to do that...
If ever I can't sleep at night I count sheep or think about the logic of all-in EV :)

###
I can't imagine how EV can work across more than one table - it's about all-ins right?

###
Anyway - below is the FAQ on EV which you probably know about & a short quote from it:

Note: EV is calculated at the point in which everyone is all-in. Calculating it any other way would be biased since we don't know what hands people might have folded. Example: You have JJ VS two other players with larger stacks with AK and QQ. You get all in, the other players call. Flop is AK5 and QQ folds, is your pre-flop EV 55% (pair VS AK) or 20% (JJ VS QQ and AK)? We can't know because we don't know what people fold
FAQ - Hold'em Manager Poker Tracking Software :: EV Explained (http://faq.holdemmanager.com/questions/120/EV+Explained)

AndrewK
06-24-2010, 04:44 AM
###
I can't imagine how EV can work across more than one table - it's about all-ins right?



Hi Loki,

Yes, it's about all-ins. The thing is, on multi-tables you do get changes in $EV diff on the individual hands played whenever someone goes all-in (see DataView page) even if you invested absolutely nothing in the pot yourself. Clearly someone on the software design team planned that $EV does have meaning in the context of multi-table tournaments, but neither you nor I has any idea what exactly that is (and maybe there is a bug in it anyhow).

Please could you pursue this with the support guys, as it's frustrating that there are stats that appear on the standard screens in the software that are not clearly explained in your documentation.

Many thanks,
Andrew

The Minder
06-24-2010, 05:39 AM
Ok, here's my take on tourney EV, and once again I'll prolly upset a bunch of folks.

First. The HEM doco on EV is vague when it comes to tourney because (drum roll please)... it doesn't work! Tbh, I don't believe EV even has a place in tournies. Why, well consider this for a cash game. You've got AA and villain has KK. That means your EV is about 82%. You go all in and clobber villain... next hand AA vs KK again, repeat repeat repeat. After a bazillion hands your bazillion AA's would have held up about 82% of the time and the EV calc makes sense because when villain sucks out on you you just reload and fire away again.

But in a tourney, the first time your AA fails you're out of the tourney... you can't reload, you're SOL. And how can HEM account for AA getting crushed like this, after all they are tourney chips and have no value outside the tourney. Perhaps we can use the buyin value? Nope because that doesn't account for payouts, and besides it's dead money as soon as the tourney starts. Ok, use the payout structure? Nope, because then EV could/would total more than your BR... and that's just plain silly.

Here's a crazy situation. Pokerstars recently tried to set a world record tourney. They got about 45,000 entrants in a $1 MTT with (I think) T$3,000 starting stack. Can't remember but I think first place got about $5,000. But, can you imagine the stacks at the final table... average would be about 15 million! Any EV calc in this would be nonsensical.

Second. When would a cash game player call a 10BB all in when he only had 7,2 offsuit. Neva!! But tourney players do it all the time because of massive differences in chip stacks (I've seen tables with 2 million chips and villain only had 100k... I'll call him with ATC all day long and twice on Sundays just to knock him out of the tourney). No EV calc allows for this variation in apparent chip value.

On the other hand, an all-in statistic makes a hell of a lot of sense for a tourney player but not so much for a cash game player. There is a separate thread where Roy said he would create an all in report.