Pretty sure that's right for juk's program.

I'll ship $20 on Stars to Juk or anyone else who will walk me through this on IM and get it working for me to where I can see a graph. I play on Stars and my hands are imported through HEM. PM me if interested.

sorry 47 pages is too much to go through but id really like to try this program out on stars... could somebody please just let me know if i can run it with stars.... or maybe the original post could be editted to say whether or not it supports stars

Specifically: Any hand where either 2 or 3 people get all-in preflop (ie: all but one of the players puts all their chips in and the hand rolls from flop to river without action).

By definition this will have to go to showdown. I then look at the cards that each player had when the all-in occurred and calculate all the possible outcomes that could have occurred after the all-in and the chance of each outcome occurring (this is what makes SNGs "luck" harder to calculate).

The outomes are then passed through the ICM function, weighted by the chance that the outcome will occur and summed to work out the expected $EV for the hand from the hero's perspective. Then the expected $EV is subtracted from the $EV of the outcome that actually occurred to find your $EV_Luck.

There are no "tricks" used for hands that don't go to showdown like Phil uses in PEV (ie: purely Sklansky bucks). The only possible problem is due to the "bunching effect" (see other massive thread in the Zoo about this). There might also be a small effect from the fact that I use 169x169 and 169x169x169 sized lookup table to speed things up (which is already ~20MB zipped) as opposed to a lookup that accounts for suit match-ups (upon testing this effect seems almost insignificant in cases such as AhKd vs QhQs, etc because the flush domination happens so rarely).

The code I currently have will work on any all-in; however many players were involved and on the flop and turn also. I just can't release it yet without releasing the full source or using a different hand evaluator. The problem of the AhKd vs QhQs type matchups can also be negated in the new code, at the expense of extra CPU time required, by not using the pre-computed lookups if asked to.

Juk

I am trying to convert Stars HUSNG's but none of the blinds seem to convert . Every Blind entry in the converted file is "Level: Blinds(,/)"

I'm using the files that have been moved by PT into a processed_files directory.

Any ideas ???

Cheers

Hi I kinda spent half an hour now looking for a link....got one version but couldn't get anything working at all. Perhaps thats because of vista?

Juk, (or anyone that has party hand histories)
Any chance to get the line message expected when the SB is dead in the Party hand history files ?

Thanks
Jax

"There is no Small Blind in this hand as the Big Blind of the previous hand left the table."

needs to be inserted just before the "** Dealing down cards **" line.

Juk

Which post #?

My settings were ok. I don't have any weird lines. I'm not sure what you are suggesting I do.

Look at post #622 there are various solutions for the parsing of the 1st line that yield errors with Blinds and Date.

I am working on some changes to Juk's conversion program and that issue you have is fixed but I am trying to fix some others issues I have before I post it, unfortunately one of them may require a big change and the time is scarse.....

Cheers

Jax

It's dave_w11's conversion program, not mine I think his was used as the template for the FTP one too so it's him mainly that you have to thank for both Stars and FTP compatibility.

This afternoon I've finally replaced the GNU pokersrc hand evaluator with a modified version of Steve Brecher's hand evaluator and after some brief testing it appears to be working fine. This means I can finally release the new code which will work with any type of all-in (the old code only worked with 2-way and 3-way preflop all-ins). A few notes:

• I still haven't added the ability to do full enumeration and since it's still using Monte-Carlo simulation when the lookups can't be used, it'll give very slightly different answers each time it's run. This barely makes any difference anyway (even if the lookups are not used at all).
• I still haven't added the ability to disable the use of the 2-way and 3-way lookups (so as to allow for "suit domination"). Doing so slows the application down significantly (by a factor of 20 or more!) and testing shows using the lookups makes hardly any difference anyway (and it averages out - in the long run, you'll "suit dominate" just as often as you get "suit dominated").
• I still haven't added any filtering capability, so the (printed and ".csv") output it basically the same as it was in the old version.

I'm going to wait for Jax's improved FTP conversion script first (seems like 90%+ of all the errors/questions posted here have been FTP related recently).

Also, if possible, can dave_w11 post the most up-to-date Stars conversion script (I messed up last time and included an old script by accident...).

Juk

Outstanding. You are the f-ing man!

Also, does this program calculate a SD? I've never actually seen an output.

I am having lot of probs modifying the FTP converter and I need to do some pretty big changes if I want to fix them, before I spend more time on it can someone confirm I am using the latest version ? What I have is date 5/5/2008 1:46 AM for the source and 5/5/2008 1:47AM for the .exe

Cheers
Jax

Ok, this is getting silly, too many inconsistencies.
I need party original hands, not files imported/exported from ptracker but genuine party HH files.

Can someone post or attach a tourny.

Thanks

Jax

Man I cant figure it out. I tried to read the readme.doc but get this message:

adobe reader could not open" readme.doc" becasue it is not a supported filetype or the file has been damaged

So I figured I would try to use ftpconverter.exe. since ftp is the site I am trying to use the program on. A box came up saying select a folder with only full tilt HH in it. So I created a new folder and moved the 1 month or so HH into the new folder on my desktop. I selected that folder and it started going thru all the tourneys, then it went to over the number of tourneys. I think It was converting my HH into new files so that it would maybe work.

Not sure what to do,

I went to C:/programfiles/sngluckanalzer to run the program and The program pops up and a error message saying note: The version has timed out pops up

I also went to excel and tried a few things but nothing worked.

Sry, I am such a noob. any help is greatly appreciated. Thanks

Open the file with either windows "wordpad" or rename it to ".txt" (I only named it ".doc" so as to not end up with it getting read as a HH file by accident).

Juk

Use two different folders: one containing the unconverted hand histories and a second one to store the converted ones. If it's "going over" the total number, you are probably creating new files in the old folder, which the converter then finds and attempts to convert again. Why it stops at two passes I don't know.

Well I am happy to modify the FTP converter so we get more consistent results but to do so I need Party hands.
Can someone post a complete SNG or send me a PM I'll send you my email so you can send it.

Thanks
Jax

If nobody has sent you any by tommorrow them PM/email me and I will.

Juk

Ok, This is where I am at. I just took all my FTP HH's starting at August 1st and converted them into a new folder.

It appears that each HH also came with a tourney summary. So actually the number of files in each of the folders is double the amount of sngs I have played.

1)Do I need to remove all summaries and start the process over so the number of files is the same as actual sngs played?

2)Should I be putting in HH into the analyzer prior to the beginning on my downswing or just start when my downswing happened?

All of my sngs have the same entry fee%. These sngs are 6 man Turbo FTP sngs with buyins of $55+5, 110+9,and $220+16

3) Do I need to create different Batch files for the different buyins?

4)How do I create a batch file.(sry, such a n00b)

Sorry for all the questions, I just want/need to see something. Any help is greatly appreciated.

"All of my sngs have the same entry fee%."

No.

Is it a bug or do I miss something?

allin_luck_results_by_hand.csv
--------------
HandNumber,ActualWinnings,AdjustedWinnings,$EV_Luck,cEV_Luck
...
7352683776,180,-83.2204,263.22,-3469.81
7352794560,-30,-392.226,362.226,-2420.37
...
--------------
#Game No : 7352794560
***** Hand History for Game 7352794560 *****
NL Texas Hold'em $210 USD Buy-in Trny: 41859607 Level: 2 Blinds(30/60) - Friday, September 05, 19:28:29 ET 2008
Table Speed #1384179 (Real Money)
Seat 2 is the button
Total number of players : 10
Seat 1: leonenivanca ( 2,270 )
Seat 2: Mathilda1988 ( 1,760 )
Seat 4: palau777 ( 1,800 )
Seat 5: Bubbelpojken ( 2,000 )
Seat 8: TURAN4OKS ( 2,550 )
Seat 3: MsC0inflip ( 1,830 )
Seat 7: ElBarone ( 3,450 )
Seat 10: paakallo ( 1,940 )
Seat 6: sst027 ( 1,860 )
Seat 9: Tupoy ( 540 )
Trny: 41859607 Level: 2
Blinds(30/60)
** Dealing down cards **
Dealt to MsC0inflip [ Kd Kh ]
Bubbelpojken folds
sst027 raises [240]
ElBarone folds
TURAN4OKS folds
Tupoy folds
paakallo folds
leonenivanca folds
Mathilda1988 folds
MsC0inflip is all-In [1,800]
palau777 folds
sst027 calls [1,590]
** Dealing Flop ** [ Jh, 4d, Ks ]
** Dealing Turn ** [ 7h ]
** Dealing River ** [ 6s ]
sst027 shows [ Qd, Ah ]high card Ace.
MsC0inflip shows [ Kd, Kh ]three of a kind, Kings.
MsC0inflip wins 3,720 chips from the main pot with three of a kind, Kings.

Are you using the correct "Entry Fee Percentage" for the $210 level (ie: it's not going to be the same as the 11s, 22s, 33s, 55s, etc and you need to make your own batch file for it)?

Juk

Any update for make your program working with the last update of PS (HH modified) ?

Thanks a lot

Yes, I made such file. For the most part the program works quite well. BTW Thank you. But sometimes I got some hands like the quoted one that seem out of line to me.
If you need, I can send you the whole HH file.