I leave the river blank and get incorrect results. I then input a river and did the max exploit feature (still get incorrect results), I then left the river blank (again) and did the max exploit feature again, and now I get the correct results.

Is it possible to form a graph of villain's equity against our hand, with villain's calling range % as a variable? If so, how?

major issues with the beta. just updated to newest beta as well (was having issues in older versions, too).

sometimes when the program starts i can't even select players (clicking on "SB" for example has no effect) and I have to close and reopen for it to decide to work sometimes but still have the same problem other times.

right now i'm having an issue where I had to close/reopen a few times to get it to let me pick players and set stacks, and then once I was at the start of the tree I set SB to raise (I don't see the options above raise anymore, used to be able to mouse over "raise" and see things like "all-in", now can only click raise and then manually enter the amount after) I enter "9999" as raise size, and it will show SB raising his full stack (in this example ~2k chips) and then everything freezes. Same thing has happened 3x now on times when I could get past the start screen.

edit: and it looks like I can't use the non-beta since my beta is taking up the register? that seems silly. I'm getting a "this key has been locked to a different machine"

How is it possible that hands which shove have -Ev under unexploitable shoving?

Are they very -ev or like -0.01 EV?

Might be some kind of rounding error if its only very slightly -ev

More than 1bb, less than 2.

http://forumserver.twoplustwo.com/sh...postcount=2170

read this.

But then villains calling range is exploitable, i.e. we can fold our -ev hands and gain.

Okay, i see.
all i can say is:

Latest beta is very unstable. Crashes if I modify raise size in a preflop only tree.

Is this bug, intentional or do i not understand the math here?

Structure .5/1, SB shoves 100% BB does x. With no rake, EV of SB strategy+EV bb strategy=1.5 (the blinds).

But if SB has a fold range, in CREV BB doesnt get awarded the EV when SB folds so SB strategyEV +bb strategy EV=less then 1.5. Is this a bug?

its been months ago that I emailed about bugged import for hands played on microgaming. its still the same mistake, it messes up the betsizes if someone raises a bet.
when I take the HH out of HM1, it only imports at all when i click prima, not when i click HM like it works wiht other sites.

Hi guys,

Ok, it appears something has gone horribly wrong.
I usually get an e-mail message that someone has posted here, upon which I go to the thread.
The last time I received such an e-mail was on 8/6/2012.
I missed all the other posts ...
I guess I have some catching up to do.

Sorry for any inconvenience.

Scylla

In order to make this graph I would simply create a hand of SB vs BB.
Make SB push all in.
BB is hero, who calls with the desired range (I've chosen "all hands").

Now edit SB's pushing range.
Double-click his "all hands" and in the preflop menu click on "use variable" and select, for example, variable 1.

Now set a checkpoint at SB's push decision and create a graph.
Vary variable 1 from 0 to 100 and as output measure Equity.

Your tree should look like this:


The graph will look like this:


Because this is the first decision for SB his raising range will be considered to be his entire range (a player's preflop range is decided in his first decision; it does not need to be "all hands"). Also, this is BB's first decision. His calling range will also be his entire range, meaning that his calling range is the one against which SB's equity is meaured (if you were to include fold "all hands" for BB, the equity would be measured against "all hands").

Please note that I haven't bothered to create a "correct" tree here. Since we're only looking at equities there's little point in creating a tree that's focussed on EVs.

This is the first time I've heard of this issue.
It could be a capturing problem, meaning that the program's window does not catch the mouse actions.
Please try if selecting some other window, and thén selecting the program's window suddenly makes it react again.
I'm rather curious if this is the first beta where you're having this specific problem?

Sounds like you're in LHE mode.
To fix that, click the LHE button to the right of the "raise" button.
You'll get back to NL mode and the raise menu will again appear.

Yep, that's a known bug.
I'll post an update tomorrow.

By including some -EV hands you're forcing villain to widen his calling range. As a result, your +EV hands will make even more profit. The additional profits for the +EV hands are greater than the losses suffered by adding the -EV hands.

If you were to remove the -EV hands and make villain make an optimal call your EV for the shove would be less then when you include them.

Edit:
There's a lot more to be said for the éxact mathematical solution.
The solution that the program arrives at it very very close to the exact mathematical solution, but it's not the optimal one.
I'm expecting to post some videos next month where I'll go over this.

It's about finding an equilibrium.
For unexploitable shoving we asume that villain knows our shoving range and make a perfect call.

The question is what is the optimal range for us to shove with, given that villain knows our strategy.
So if we were to fold our -EV hands villain would know we did this and adapt.
We'd make less money.

I know.
I'll post an update tomorrow.

No, that's not a bug.
The times that SB folds and BB wins the 1.5 pot also needs to be taken into account.
Please have a look at the tree below.



BB's EV for the 40.6% of the time that SB shoves is 0.42.
The other 59.4% BB will win the 1.5 pot.
That makes BB's EV for the entire situation 40.6%*0.42+59.4%*1.5=1.06.

And SB's EV is 0.44.

So that makes their combined EV 1.06+0.44=1.5.

Ok, I found your e-mail and it indeed hasn't been fixed.
I'll make sure to include it in the update for the beta tomorrow.

Sorry for the inconvenience.

PS: To get the hand history from HEM, select it by left-clicking it in the "Reports" section, then right-click it and select "view selected hand histories".

thanks! sorry if I came off to rude, was pretty tilted after loosing a bunch...and its pretty annoying when I analyze the EV of everything and then realize it was the wrong potsize and all my "discoveries" of thin EV dont make sense. other than that support was always excellent!

If you hadn't reminded me I would never have fixed it
This bug report definitely used to be on my to-do-list, but it appears I misplaced some items on that list.
Anyhow, I have fixed it and it will be included in the v2.8.1 beta which will be released tomorrow.

PS:
Sorry guys, I said the update would be today, but I found that after implementing quite a few changes today that it was already 22:00. I prefer to give the new features in the beta a good looking-over before releasing it. I'll be working on the beta first thing in the morning. I expect it to be up shortly.

just a quick one:

simple tree, heads up, SB open shove and BB calls with all hands. SB open shoves, let say 22-55, AQs (for argument sakes) but he open shoves AQs 10 times less often than 22-55.
How do I enter that? create a condition with 22-55 at 100% and AQs at 10%?

Yes, I would imagine so.
Asuming what you mean is that SB shoves 22-55 100% of the time he has that, and AQs only 10% of the time that he's dealt that hand.

In order to set such a weight, select one of the miniweight buttons in the upper right of the matrix in the preflop menu. Select AQs. And now drag the weight bar to the right of the matrix down to 10%.

Or did I misunderstand your question?

first of all you have to be the quickest suport ever! 3 minutes !lol

I must be blind can't see a miniweight button? just the one next to all hands