Thanks!

Are the columns expressed in $, bb, etc?

Everything is in dollars.
If you want to see results in bb, then I'd advise just using 0.5/1 blinds and (if applicable) set the rake and cap accordingly.

Or, just divide the dollar amounts by the blind sizes for yourself.

hi i tried to put the blinds to 0 so no money in the pot.
two players allin preflop with 100% of hands.
it says that the EV for both players is -0.97 but shouldnt it be 0 or is there something i misunderstand.

Did you adjust the rake?

It sounds like you're using 5% rake with 35% rakeback and a $3 cap.
If both players lose 97 cents, then 2*0.97=$1.94 goes missing.
And 65% of the $3 cap is indeed $1.95.

How can I get access to your youtube-videos?

CardRunnersEV
These videos are currently being produced and are not fully available yet. What's available so far can be reached by:
- installing the CardRunnersEV beta (http://www.cardrunnersev.com/download.html)
- run it
- in the startup screen, click on the "Videos" button

The videos that are available so far are linked to there. Please observe though that the ones that you cán get to right now are on the more advanced subjects of CardRunnersEV, such as writing scripts and the unexploitable shoving algorithm. The videos that are yet to come are the introductory ones.

Flopzilla
Should you be looking for videos on Flopzilla, please go here: http://www.flopzilla.com/video.html

If i bet for value with frequency x. I want to bluff with highcards so that the frequency of bluffs is x/2. Is there any way to do it? So when i change the hands i bet i dont need to recalculate the frequency of highcards to bet.

If i bet 40% hands for value, i want to bluff 20% with highcards. And if the hands i bet are 30% i want to bluff 15%.

Thanks.

You cóuld achieve that with a script.
Here's an example: www.cardrunnersev.com/rar/fractions.rar
This rar file contains both a savefile and a script.
To let the software detect the script, save it to CardRunnersEV's /scripts subdirectory.

Click the script icon in the main field's lower right to open the script window.
Load the script with "Load script" and run it with F4.

hey,
I just reinstalled crEV beta to new location. What files should I transfer there to get same settings that I used to work before?

Thanx.

Hi,

You will need:
- postflopconditions.txt
- prefloppredefs.txt
- preflopranges.txt
- structure.txt

Hi,

I used several times CREV in finding frequencies like Flopzilla does (last time I used Flopzilla was in free version before you had to pay for using it).
However I seem to get inconsitent results for 2pair hands in CREV.

Before the CREV beta release, I remembered that you had a pop-up saying something about that "pairs at the board are not taken into account when counting (calc frequencies) for 2 pairs".

To give an example (100 000 equity run)
SB MR 100%
BB calls 100%
flop JhJc5d
BB checks 100%
SB cbets 100%
BB calls with following sieve
1) better or equalt to 2 pairs (or better made hands) 9.87%
2) better or equalt to top pair 0%
3) better or equal to middle pair 13.6%
4) better or equal to 3rd pair 1.56%
5) better or equal to a pair 0%
6) better or equal to 2 overcards 4.18%
7) better or equal to 1 overcard 32.6%
8) all hands 38.2%

Now, as a poker player the 2 pair frequency makes sense (all pocket pairs, Jx 5x => counts to 2 pair)

Every other hand without a card of which the rank is at the table would result to a toppair hand because flushes, streets are not possible.
Nevertheless, we see toppair 0% middle pair 13.6% 3rd pair 1.56% a pair 0% 2 overcards 4.15% 1 overcard 32.6% all hands 38.2% which imo should be placed all added up after toppair.

If I remember correctly the pre-beta popup for 2 pairs could be seen as:
(ie JhJc5d --- Th4c4d --- topcard = J or T --- bottom card = 5 or 4)
counted to 2 pairs -- all sets, trips or better AND all pocket pairs with rank > topcard
counted to top pair -- impossible if 2 topcards at board AND if 1 topcard at board all hands forming a pair with topcard
counted to midpair -- all pocket pairs with rank not equal but between topcard and bottom card AND if 1 bottomcard at the boards all hands forming a pair with bottom card
counted to 3rd pair -- all pocket pairs with rank < bottom cards
a pair -- nothing as no pairs are left with above sieve
2 overcards -- 2 cards of different rank in hand AND none forms a pair with topcard or bottomcard AND both > topcard
1 overcard -- 2 cards of different rank in hand AND none forms a pair with topcard or bottomcard AND 1 card > topcard
all hands -- 2 cards of different rank in hand AND none forms a pair with topcard or bottomcard AND both cards < topcard

Can you confirm the above?
Now, if I do a paired board analysis (inserting a paired board); this is not a big issue once you know and realize this as it is little effort to go back to normal 2pair, pair definitions ...

Now, what happens if I keep the board random?
Will the 2pair, toppair, ... frequencies still contain 'miscounted 2 pair hands' in the case a flop is dealt that contains a pair?
Or is in that case the toppair frequency correctly used as it should be?

trips and boats?

My sieve as shown in CREV:
>=2p
>=tp
>=mp
>=3rd p
>=1p
>=overcards
>=overcard
all hands

You get >= by selecting at least.

As I should understand >=2p: trips, sets, boats, all kind of 2 pair hands (Jx 5x)
As I understand >=2p after pop up and my awareness of inconsistency @ paired boards: trips, sets, boats, all 2 pairs when you have a pocket pair with rank > topcard
>=2p at random boards: ???

Could you send a savefile to support please?
And when you mail a brief description of what decision you're looking at and what you thought the % should be?

Sorry, not really able to follow... but there is something wrong, imo... maybe we're talking about the same thing. Beta 2.8.6.

Board JJ5r, 100% ranges.

Evaluating in following order.
Trips+:8.33% ( any J, 55 )(correct)
2pair: 0%
2pair+:1.53% (AA,KK,JJ) (doesn't seem like correct to me)

Is it possible to do an max exploit on only one "action"?

Lets say,

I open BT with 60% of my hands
Villian 3bet
I want to 4bet some hand of the 60% range. <- In this action I want to see the maximum EV hands to 4bet of the 60% I opened on the button
Villian 5bet push (unexplo push)

No that's intentional and it becomes a bit more obvious once you do the same thing for a paired turn:
The board is JT55r.
What is better?
The overpair QQ or 2 pair (JT)?
The software considers QQ as being TRUE for >=2p, since it beats JT.

You could just 4bet with "all hands" and then use Alt+D to remove the -EV hands.

To make math manageable, I replaced sb with 5c5s and kept ATC for BB

Below what I would expect after should be

SB 5c5s
BB ATC
flop JhJc5d -
100 000 cash mode relative run

>=2p
CREV 10.2%
SHOULD BE =6.11+12.49=18.6%
AA KK QQ TT 99 88 77 66 44 33 22 = 11 * (4*3)/(47*46)=6.11%
JJ = 2*1/(47*46)=0.09%
all hands without a 5 J = (52-8)*(51-8)/(47*46))=87.51%
all hands containing a 5 J = 100-87.51=12.49% <= contains JJ


>=tp
CREV 0%
SHOULD BE 0%

>=mp
CREV 6.88%
SHOULD BE 0%

>=3rd p
CREV 1.69%
SHOULD BE 0%

>=1p
CREV 0%
SHOULD BE 0%

>=overcards
CREV 4.56%
SHOULD BE = 6.11-1.67=4.44%
AA KK QQ = 3 * (4*3)/(47*46)=1.67%
2 cards out {A K Q} = (12*11)/(47*46)=6.11% <= but contains AA KK QQ

>=overcard
CREV 35.3%
SHOULD BE 35.52%
1 card out {A K Q} + 1 card out {T 9 8 7 6 4 3 2}=2*(12*32)/(47*46)=35.52%

all hands
CREV 41.3%
SHOULD BE =45.88-4.44%=41.44%
2 cards out {T 9 8 7 6 4 3 2} = (32*31)/(47*46)=45,88%
<= but contains TT 99 88 77 66 44 33 22 = 8 * (4*3)/(47*46)=4.44%

File is send by mail

What I have been doing is that I've tried to come up with an unexploitable 4bet range (if his 3bet range are static, for example always 12%).

So,
1. I open 60% hands on BT
2. get 3betted by 12% (a range I believe he has)
3. I 4bet
4. Villian unexplo push
5. Hero perfect call

I want to see the maximum EV for me in the 4bet spot AND exclude hands I usually calls, and then see the EV. Is it possible to exclude the call hands (AJ for example), which obvious have higher EV than for example A2o.

I could do this manually, but is there a faster way?

Ok, though it's a bit weird that QQ will be either 1p or >2p, depending on which one is evaluated first. The flop JJ5 is a bit confusing still, but I'll live with it... again, QQ is either 1p or >2p, but it's never 2 pair ( as nothing is on this flop). Reqular math and logic doesn't apply... division by zero -error in my brain.

I like the fact that you chose to do the evaluations on paired board like this (i.e. not counting PP as 2pair), but I guess this is the price we have to pay...

True.
However, its entirely your fault
You see, you're asking a question to which there is no correct answer.
After all, you're asking for >=2p.
But 2 pair does not exist on a J55 board.
So there is no correct way to interpret this query.
It's just impossible.
However, given that this is a computer program, it's forced to either assign TRUE or FALSE to QQ being >=2p.
Given that:
a) If the board were JT55 QQ wóuld clearly be >2p
b) QQ on J55 cóuld be interpreted as 2 pair, depending on your definitions
the software goes with TRUE.
To clarify point b further, if someone asks you if QQ on J55 is >=2p and you are forced to answer YES or NO, then, even though the question is faulty, it's simply a clear YES.

QQ is not considered as 2 pair.
It's considered as beating 2 pair.
After all, you're asking for >=2pair.
If you enter a condition "2pair", then you'll find that QQ is not included in it on this board.

However, once again, the reason for the confusion here is that 2 pair does not exist on a J55 board. And overpairs fall into a range of borderline hands that can either be considered TRUE or FALSE. They're either better or worse than 2 pair, depending on how you want to define things. There's just no clear answer. And yet the software is forced to choose one or the other.

Ok, I've received your savefile.
It sure makes things a lot clearer.
See pic below for a screenshot of that tree



Now, the issue becomes a lot more clear when we're able to perform an EV run in this tree (F7), since this will allow for popups. However, currently this is not possible, because the tree is incomplete. For that reason, I've just made BB openfold the turn. It ain't pretty, but the tree is complete and we'll be able to perform an EV run (F7). See pic below:



And when we now, for example mouse over >=2pair we get a popup that shows which hands specifically are in that condition. Press Alt+C to toggle to combo mode and you'll see how many combos there are of each hand.



So, there's a total of 109 combos in total that are considered as >=2pair. Given that there's a total of 47*46/2=1081 ways of dealing BB a hand out of "all hands", this means that the frequency for >=2pair should be 109/1081*100%=10.1%.

If you want to see how the software arrives at the other frequencies, then just do the same thing and mouse over the appropriate condition to see the number of combos in it.

If you want to know why QQ+ is considered as beating 2 pair on this board, then please read my previous post on that (#2623).




PS:
Card removal with combos only work if SB is dealt one specific hand. If SB holds a range, then card removal is ignored.

Is this what you mean?

SB's "raise to 3" range is fixed at 60%.
BB's 3bet range is fixed at 12%.
At point 2, SB just calls with some fixed range and checks down the flop.
Otherwise, at point 3 and beyond, you want the equilibrium where SB 4bets and BB makes an unexploitable shove?

If so, then yes, this can be done with a script.
Otherwise, please send a savefile to support with a bit more details about what you're looking for.