Sklansky Bucks Calculator: PokerEV

[Phil153]

[ QUOTE ]
Main play is at Crypto, only play Stars when there are not enough tables there so would love Crypto support.

[/ QUOTE ]
I'm trying to get more sites supported for the next release (as well as tournaments, finally), but I need a lot of hands for testing and debugging. The more the better, especially for smaller sites.

If you want to help, you can email hands to phil@pokerevsoftware.com

You can also upload hands very easily here which get automatically emailed to me:

http://www.pokerevsoftware.com/form/use/HHUploader/

Everything's kept confidential of course. In terms of sites:

Priority
few hand histories, not currently supported

Everest
Cryptologic
Prima
Ongame
Bodog

Not priority
but helpful for tournaments and further debugging

Party Poker
Full Tilt
PokerStars
iPoker

Any versions are fine, including PokerTracker output, raw site histories, site xml, .dat database files, etc.


[ QUOTE ]
phil:

When Are you Adding SD/100 for Graphs? Would be nice to know so We could run some simulations or something.

[/ QUOTE ]
Next release, mid-month. It won't be SD/100 but something similar.

[schwza]

[ QUOTE ]
Thanks, that is a bug. It doesn't affect the calcs but will give the wrong value in the "total pot size" column where hero overbets and gets called by a shorter stack.

You can look at/verify the all-in values if you scroll to the far right of the analysis tab.



[/ QUOTE ]

i'm looking at those two columns and i have a question about how you're calculating AllinNetWon and AllinNetExpected. it looks like those values only look at the size of the all-in bet (or maybe all of the bets on the all-in street). is that correct?

it's more intuitive to me to look at those variables for the entire pot size.

for example, say two players each put in $100 preflop and $200 on the flop and they are all-in on the flop. A is a 9:1 favorite over B on the flop, but B outdraws A. how much bad luck did A just suffer? by your calculation, it is

.9 * $400 = $360

in my view, he suffered more than that. when he put his last chip in the pot, there was $600 and he had 90% equity, so he had .9 * $600 = $540 of bad luck. regardless of how the money went in, once the luck factor showed up, there was $600 at stake, not just the $400 that went in on the flop.

also, i noticed your doing 3way equity calcs. how are computing the side pot and main pot equity? are you assuming that winning the side pot and winning the main pot are independent events? i.e., running the sim once for the main pot and then if the shortstack wins then run another sim for the side pot?

finally, are you actually enumerating preflop all-in? i would think it'd be easier and faster to just have a master chart stored inside the program so if Td8d gets ai against AdAh you just look up the equity for T8s vs AA. you'd lose the fact that having the Ad helps a little but since we're talking about all-in preflop it's just luck how the suits match up anyway.

[Phil153]

[ QUOTE ]

it's more intuitive to me to look at those variables for the entire pot size.

for example, say two players each put in $100 preflop and $200 on the flop and they are all-in on the flop. A is a 9:1 favorite over B on the flop, but B outdraws A. how much bad luck did A just suffer? by your calculation, it is

.9 * $400 = $360

in my view, he suffered more than that. when he put his last chip in the pot, there was $600 and he had 90% equity, so he had .9 * $600 = $540 of bad luck.

[/ QUOTE ]
That's the way the all-in-calc program does it, but I don't like it. Say for example you put in $380 preflop as a 5:1 dog, and then your final $20 on the flop (this happens sometimes) as a 9:1 favorite. If you lose, did you really get unlucky for $720-odd dollars? That's the problem with basing it on the entire pot size. I went for the option of going for a smaller calculation that's a far more accurate indicator of luck. I've tried fiddling with a calc where if you put say 80% of the your stack in, then it calcs both streets, but the potential for skew is there. I may add it as a user option.

[ QUOTE ]
also, i noticed your doing 3way equity calcs. how are computing the side pot and main pot equity?

[/ QUOTE ]
Two separate sims for each pot.

[ QUOTE ]
finally, are you actually enumerating preflop all-in?

[/ QUOTE ]
I'm doing a monte carlo for preflop,which is slightly inaccurate (1-2%) but doesn't affect cumulative results as the inaccuracy is randomly spread on both sides. It'll throw a 2000 hand, $400 graph out by ~$1 for example, it really is inconsequential and worth it for the speed.

However, I've written a PokerStove-speed preflop enumeration algorithm and I hope to get it into the next version.

[schwza]

[ QUOTE ]

That's the way the all-in-calc program does it, but I don't like it. Say for example you put in $380 preflop as a 5:1 dog, and then your final $20 on the flop (this happens sometimes) as a 9:1 favorite. If you lose, did you really get unlucky for $720-odd dollars? That's the problem with basing it on the entire pot size.

[/ QUOTE ]

it's true that sometimes in your example the guy who was a 9:1 favorite was actually a 5:1 dog preflop. but it's much more likely that the 9:1 flop favorite was also the 5:1 preflop favorite.

your way of calculating luck magnitude basically assumes that the money that went in preflop went in as 50-50. the other way basically assumes that if the flop money went in as 80-20 then the preflop money went in 80-20 also. that seems more right to me but i think it's just a matter of preference.

the biggest reason i want to see your program done in the same way as all-in-calc is so that i can run them both on the same database and see how close they are to agreeing. (you would also need an option for ignoring 3-way pots for this to really work).

[thac]

Why am I always running ahead of S-Bucks? Is it possible to run hot for 45k hands?

[xd40]

[ QUOTE ]
[ QUOTE ]
Main play is at Crypto, only play Stars when there are not enough tables there so would love Crypto support.

[/ QUOTE ]
I'm trying to get more sites supported for the next release (as well as tournaments, finally), but I need a lot of hands for testing and debugging. The more the better, especially for smaller sites.

If you want to help, you can email hands to phil@pokerevsoftware.com

You can also upload hands very easily here which get automatically emailed to me:

http://www.pokerevsoftware.com/form/use/HHUploader/

Everything's kept confidential of course. In terms of sites:

Priority
few hand histories, not currently supported

Everest
Cryptologic
Prima
Ongame
Bodog

Not priority
but helpful for tournaments and further debugging

Party Poker
Full Tilt
PokerStars
iPoker

Any versions are fine, including PokerTracker output, raw site histories, site xml, .dat database files, etc.


[ QUOTE ]
phil:

When Are you Adding SD/100 for Graphs? Would be nice to know so We could run some simulations or something.

[/ QUOTE ]
Next release, mid-month. It won't be SD/100 but something similar.


[/ QUOTE ]

ye, recently i changed to prima and i noticed it wasnt supported. continue with the good work

[HesseJam]

What a wonderful program Phil!

If my red SB luck line has a negative slope over 100 all-in hands, chances are that I was tilting, right? => I was constantly getting all-in not having the best of it.

[awjpoker]

[img=http://img2.freeimagehosting.net/uploads/th.b6115b5bee.jpg]

Looking for an overview of the graph in the link, does this mean i am running bad or just playing bad?

[delta k]

Does Absolute Poker work yet?

[1968]

[ QUOTE ]
[img=http://img2.freeimagehosting.net/uploads/th.b6115b5bee.jpg]

Looking for an overview of the graph in the link, does this mean i am running bad or just playing bad?

[/ QUOTE ]
Means you are running as expected but losing too much money in non-showdown hands (eg- you are putting money in, then folding)..