Sklansky bucks calculator (php script)

[NoMeansYes_]

Changing my starting hands, I have found that the hands I lose the most with are off suited. I'm guessing this is mostly because I fold 90% of them, playing a 20/15 style. As a 5.3BB/100 winner over 72k hands, my graph shows that if I would break even playing off suited hands like 97o, 62o, AJo etc better then I would be running at twice my BB/100. Does this mean my biggest leak is folding off suited cards?




This is all my hands, which shows how much I am losing out from not playing off suited cards/playing them badly.




btw great tool!

[Phil153]

Yeah, that's pretty amazing, but it's hard to tell how much is just blind folds. Try filtering for hands where you saw a flop and looking at those numbers. The best way to do this is to select made hand types.

I'll be adding stats and filters for proper positions and blinds which should make it easier to separate this stuff out.

[mustmuck]

Phil,

I've had loads of thoughts while using the app over the past few days, but you've probably got enough on your plate at the moment.

One thing though might be to include a equity while all-in before river vs actual earnings. At the moment you have one that uses the NetSB. While this is good, if you added one that was just plain equity at the time of the all-in then the lines would definitely converge.

[NoMeansYes_]

What is system.outofmemoryexception mean? I have 3gb ram but I doubt ram is the cause.

[holdme]

fyi prima looks very inaccurate
not that it matters
prima sucks

[Phil153]

Out of Memory - I used managed memory and the GC chokes on multithreaded memory intensive tasks (it doesn't release it fast enough after disposal). I can fix it but not immediately, for now stay below 100K hands.

holdme - Prima is not supported, so you'll see some crazy graphs from there. PokerTracker stores data differently for each site. I left in the option to run calcs on untested sites but will remove it given the confusion it's causing. [img]/images/graemlins/smile.gif[/img] For now, stick with Party, Stars, UB, iPoker and Full Tilt.

[ImsaKidd]

NoMeansYes:

You are running as expected with offsuit hands. Your winnings are negative because you are (correctly) folding these hands to a raise or reraise. Overall, you are losing money out of showdown. This means you should probably be bluffing more, folding less, or something like that. However, your winrate is still very good, so an adjustment might not be necessary, just something to consider.

[ImsaKidd]

Pete:

I think adjusting for whiffed sets or SCs is going to add more inaccuracy. Phil could have separate stats for "hit set" or "hit combo draw", but not factor them into equity adjusted winnings. This would tell you that your equity may be higher or lower than true expectation because you hit many/few sets.

I dont understand how a large sample is bad for accuracy. Was this meant only for the PP/SC argument?

[DWarrior]

[ QUOTE ]
Yeah, that's pretty amazing, but it's hard to tell how much is just blind folds. Try filtering for hands where you saw a flop and looking at those numbers. The best way to do this is to select made hand types.

I'll be adding stats and filters for proper positions and blinds which should make it easier to separate this stuff out.

[/ QUOTE ]

I did the same "All Unsuited" filter and my graph looked pretty similar, straight down losing -6.8bb/100

Then I added filters "Position OTB at most 3" "No. of Players Exactly 6" and they were winners. Same with "Position OTB at most 2" "No. of Players Exactly 5". So it looks like the blinds are causing this.

Oh yeah, then out of curiosity I cleared the filters and checked All Suited. I'm also a loser with that, though a bit better.

Seems like the only profitable hands are Pairs, 117 BB/100!!!

[ImsaKidd]

Phil:

Any way to figure out the variance or standard deviation of the equity lines? Especially interested in the SD/100 of the teal line. We can figure out the equivalent # of "regular" hands froma given sample.

Like, Maybe a pokerEV analyzed sample of 10k hands = same variance in winrate as 40k regular hands. Something like that.