Haley's Poker Blog

No bad beats, but still a poker blog... hence the anguish.

Sunday, January 24, 2010

Inside the UB Hand Histories, Part 1

"Sent down for the expert again, and this time he struck a most dazzling idea--he fixed the thing so that opening the kitchen door would take off the alarm. It was a noble idea, and he charged accordingly. But you already foresee the result. I switched on the alarm every night at bed-time, no longer trusting on Thomas's frail memory; and as soon as the lights were out the burglars walked in at the kitchen door, thus taking the alarm off without waiting for the cook to do it in the morning. You see how aggravatingly we were situated. For months we couldn't have any company. Not a spare bed in the house; all occupied by burglars." -- Mark Twain, "The McWilliamses and the Burglar Alarm"

I'm not sure exactly why the above excerpt strikes me as insightful when regarding the UltimateBet (now renamed UB.com) online poker scandal, and yet it does. After all, one must find a way to laugh at some of this stuff, else go insane. I've been throwing out the occasional post on the topic here, and even dropped in with a handful of posts in a related thread on 2+2. As a rule I don't post much on 2+2, for a couple of reasons. First, 2+2 tends to degenerate to a pack mentality, and I prefer to think independently; second, I just don't plan on wasting the best of my thoughts in a way that could put money in Mason's and David's pockets. Be that as it may, there has been some good thought expressed over there, so I've been obliged to chip in a comment or two.

One of the things I mentioned but in passing over there was my belief that the UB hand histories -- as being sent out -- were likely incomplete, and were chopped up to pieces and made incomprehensible as a bit of further obfuscation by UB. To this end I'd like to cite the hand histories sent out to Barry Greenstein, who had limited play against the cheating accounts, and did not receive a refund. Barry's in perhaps the most difficult position of all non-UB folks regarding the scandal, in the wake of his stepson Joe Sebok accepting a largish endorsement contract to shill on behalf of UB's interests. Barry's one of the most prominent and outspoken of all well-known pros regarding the UB scandal, and indeed has stated publicly his own feelings against Joe signing the deal with UB. As I mentioned in a post last fall, Joe's deal effectively neutralized Barry and Joe's popular PokerRoad site as a place where meaningful debate on the topic could occur. That pooch has been screwed; the PokerRoad playing field, viewed from afar, is tilted and has piles of dog poo littering its blades of grass.

It's forced Barry into protesting his own innocence as well, though anyone who knows anything at all of Barry knows the man is one of the game's straightest shooters, and is unlikely to have been involved with the scandal in any culpable way. Barry also defended Russ Hamilton early on, though only on the basis of Barry having talked to Russ about it and getting the feeling that Russ was telling the truth. Barry's since recanted that defense of Russ in the face of increasing evidence.

Another thing that Barry did is to post the hand histories that UB sent him, upon request. I've buried the lede deep here, but with reason: if I start out with a pile of tables and numbers no one's going to read on. Barry did not receive a refund from UB, yet did receive a smallish number of hand histories, which he posted on PokerRoad here.

As you can see, it's a jumble, and the vast majority of these files are two- and three-hand snippets with everything rendered in a way that makes deciphering any hand's contents a test of patience and tenacity.

Yes, The Hands and Formats Can Be Deciphered

Still, there's information to be had. Here's how the files were posted by Barry's minions, which is likely to be the same way UB sent them out:

68720,147524_9173956_67065_67202.txt
68720,638132_9173956_66826_66826.txt
68720,966687_5415005_1559_1561.txt
68720,445519_9973341_3770_3772.txt
68720,638132_9173956_44933_44940.txt
68720,445519_9973341_3674_3675.txt
68720,638132_9173956_66821_66824.txt
68720,638132_9173956_45005_45005.txt
68720,638132_9173956_44930_44931.txt
68720,966687_9173942_39001_39003.txt
68720,445519_9973341_3768_3768.txt
68720,966687_9173942_39024_39025.txt
68720,638132_9173942_24376_24378.txt
68720,445519_9973341_3678_3679.txt
68720,445519_9973341_3725_3728.txt
68720,147524_5026496_3190_3190.txt
68720,147524_5026496_3193_3230.txt
68720,638132_9173942_24541_24542.txt
68720,638132_9173956_45024_45026.txt
68720,638132_9173956_66799_66804.txt
68720,40018_9173956_51543_51543.txt
68720,40018_1612888_17144_17180.txt
68720,638132_9173956_44943_44943.txt
68720,40018_1747573_6712_6748.txt
68720,638132_9173956_44981_44982.txt
68720,966687_5415005_1555_1555.txt
68720,445519_9973341_3681_3682.txt
68720,40018_9100457_387_476.txt
68720,638132_9173942_24383_24383.txt
68720,638132_8513886_4518_4518.txt
68720,966687_9173942_39005_39008.txt
68720,638132_9173942_24373_24374.txt
68720,445519_9973341_3701_3706.txt
68720,638132_9173956_66834_66856.txt
68720,638132_9173956_66791_66795.txt
68720,966687_9173942_38994_38996.txt
68720,445519_9973341_3657_3661.txt
68720,638132_9173956_44946_44956.txt
68720,638132_9173956_44994_44994.txt
68720,966687_9173942_39021_39021.txt
68720,147524_10339613_31834_31868.txt
68720,40018_10339613_28578_28620.txt
68720,445519_9973341_3732_3732.txt
68720,638132_9173942_24544_24547.txt
68720,638132_8513886_4521_4522.txt
68720,40018_9173942_43575_43597.txt
68720,638132_9173956_66782_66785.txt
68720,638132_9173956_45028_45029.txt
68720,445519_9973341_3763_3766.txt
68720,638132_9173942_24529_24533.txt
68720,445519_9973341_3719_3721.txt
68720,966687_9173942_39018_39019.txt
68720,638132_9173942_24380_24381.txt
68720,445519_9973341_3668_3668.txt
68720,445519_9973341_3698_3699.txt
68720,445519_9973341_3691_3691.txt
68720,445519_9973341_3693_3696.txt
68720,638132_9173956_44975_44978.txt
68720,147524_1612888_17406_17480.txt
68720,445519_9973341_3745_3748.txt
68720,638132_9173956_44958_44963.txt
68720,638132_8513886_4511_4511.txt
68720,445519_9973341_3670_3671.txt
68720,445519_9973341_3753_3760.txt
68720,445519_9973341_3735_3736.txt
68720,638132_8513886_4505_4509.txt
68720,966687_5415005_1564_1565.txt
68720,638132_9173956_66787_66788.txt
68720,638132_8513886_4502_4502.txt
68720,638132_9173956_45009_45011.txt
68720,638132_9173956_44966_44970.txt
68720,445519_9973341_3663_3666.txt
68720,638132_9173956_66807_66815.txt
68720,638132_9173956_45020_45021.txt
68720,445519_9973341_3739_3740.txt
68720,638132_9173942_24536_24539.txt
68720,40018_8513464_43158_43260.txt
68720,638132_8513886_4513_4515.txt
68720,147524_10341187_9585_9656.txt
68720,638132_9173956_66819_66819.txt
68720,966687_9173942_39010_39013.txt
68720,445519_9973341_3723_3723.txt
68720,966687_5415005_1552_1553.txt
68720,40018_1612888_17187_17189.txt
68720,638132_9173956_44984_44989.txt
68720,638132_9173956_45017_45017.txt
68720,40018_10339613_37627_37627.txt
68720,445519_9973341_3688_3689.txt

Looks like a true M&M's chocolate mess to me. But there are patterns afoot, within these 88 files. Let's take all these filenames, dump them into an Excel database, and sort them. The results show clear groupings:

(Group 1)
68720,147524_10339613_31834_31868.txt
68720,147524_10341187_9585_9656.txt
68720,147524_1612888_17406_17480.txt
68720,147524_5026496_3190_3190.txt
68720,147524_5026496_3193_3230.txt
68720,147524_9173956_67065_67202.txt

(Group 2)
68720,40018_10339613_28578_28620.txt
68720,40018_10339613_37627_37627.txt
68720,40018_1612888_17144_17180.txt
68720,40018_1612888_17187_17189.txt
68720,40018_1747573_6712_6748.txt
68720,40018_8513464_43158_43260.txt
68720,40018_9100457_387_476.txt
68720,40018_9173942_43575_43597.txt
68720,40018_9173956_51543_51543.txt

(Group 3)
68720,445519_9973341_3657_3661.txt
68720,445519_9973341_3663_3666.txt
68720,445519_9973341_3668_3668.txt
68720,445519_9973341_3670_3671.txt
68720,445519_9973341_3674_3675.txt
68720,445519_9973341_3678_3679.txt
68720,445519_9973341_3681_3682.txt
68720,445519_9973341_3688_3689.txt
68720,445519_9973341_3691_3691.txt
68720,445519_9973341_3693_3696.txt
68720,445519_9973341_3698_3699.txt
68720,445519_9973341_3701_3706.txt
68720,445519_9973341_3719_3721.txt
68720,445519_9973341_3723_3723.txt
68720,445519_9973341_3725_3728.txt
68720,445519_9973341_3732_3732.txt
68720,445519_9973341_3735_3736.txt
68720,445519_9973341_3739_3740.txt
68720,445519_9973341_3745_3748.txt
68720,445519_9973341_3753_3760.txt
68720,445519_9973341_3763_3766.txt
68720,445519_9973341_3768_3768.txt
68720,445519_9973341_3770_3772.txt

(Group 4a)
68720,638132_8513886_4502_4502.txt
68720,638132_8513886_4505_4509.txt
68720,638132_8513886_4511_4511.txt
68720,638132_8513886_4513_4515.txt
68720,638132_8513886_4518_4518.txt
68720,638132_8513886_4521_4522.txt

(Group 4b)
68720,638132_9173942_24373_24374.txt
68720,638132_9173942_24376_24378.txt
68720,638132_9173942_24380_24381.txt
68720,638132_9173942_24383_24383.txt
68720,638132_9173942_24529_24533.txt
68720,638132_9173942_24536_24539.txt
68720,638132_9173942_24541_24542.txt
68720,638132_9173942_24544_24547.txt
68720,638132_9173956_44930_44931.txt
68720,638132_9173956_44933_44940.txt
68720,638132_9173956_44943_44943.txt
68720,638132_9173956_44946_44956.txt
68720,638132_9173956_44958_44963.txt
68720,638132_9173956_44966_44970.txt
68720,638132_9173956_44975_44978.txt
68720,638132_9173956_44981_44982.txt
68720,638132_9173956_44984_44989.txt
68720,638132_9173956_44994_44994.txt
68720,638132_9173956_45005_45005.txt
68720,638132_9173956_45009_45011.txt
68720,638132_9173956_45017_45017.txt
68720,638132_9173956_45020_45021.txt
68720,638132_9173956_45024_45026.txt
68720,638132_9173956_45028_45029.txt
68720,638132_9173956_66782_66785.txt
68720,638132_9173956_66787_66788.txt
68720,638132_9173956_66791_66795.txt
68720,638132_9173956_66799_66804.txt
68720,638132_9173956_66807_66815.txt
68720,638132_9173956_66819_66819.txt
68720,638132_9173956_66821_66824.txt
68720,638132_9173956_66826_66826.txt
68720,638132_9173956_66834_66856.txt

(Group 5a)
68720,966687_5415005_1552_1553.txt
68720,966687_5415005_1555_1555.txt
68720,966687_5415005_1559_1561.txt
68720,966687_5415005_1564_1565.txt

(Group 5b)
68720,966687_9173942_38994_38996.txt
68720,966687_9173942_39001_39003.txt
68720,966687_9173942_39005_39008.txt
68720,966687_9173942_39010_39013.txt
68720,966687_9173942_39018_39019.txt
68720,966687_9173942_39021_39021.txt
68720,966687_9173942_39024_39025.txt

There's at least a partial decoding for these hands, too. "68720" is likely the numeric ID for Barry's UB account "crazyplayer". The last two numbers are clearly the starting and ending hand numbers for each text file, and by clicking the link above and examining any of the files, one can see that the hand numbers match up. Notice that this also gives us an exact count as to the number of hands each text file contains, ranging from one (in several instances) to 138 in txt file "68720,147524_9173956_67065_67202.txt", the very first file in the original list. I'll even note these counts as (1) and (138) and others when we add more information in a bit.

Left to be determined is the middle section, such as "966687_9173842", which I deem to be a cryptic combination of table number, game type, and some translated date/time stamp. Since the dates are also included in the files it doesn't matter much; we have the information we need anyway. The 4a/4b and 5a/5b groups share like data in the first half of that field, and we'll look at that in a bit. The latter portion, such as "9173956", is tied to the table, so a given hand at a given table is expressed as "9173956_67065" or similar.

Barry's Hands are Curious... To Be Continued

Let's start with Group 1. This session for Barry started on November 26, 2005, at about 10:24pm system time, when he took a seat at table 9173956 during Hand #67065. Barry, again playing as crazyplayer, didn't receive cards until the next hand, #67066. The cheating account in play against him here, according to the file, was "snowball". (according to the KGC report, "snowball" was one of 19 different screennames used by one of the most active cheating accounts, #147524.) "snowball" played continuously against Barry and other accounts through Hand #67176, sitting out at that time, a little over an hour later. Barry himself sat out at this table in Hands #67079 through #67092. Therefore, Barry payed against this cheating account for only 97 of the 138 hands in this file. At the time Barry ("crazyplayer") sat down, snowball had $2,521, but all those hands later, snowball departed with $34,071, a neat $31,500 profit. Barry sat down with $5,000, rebought once for another $2,500, and 111 hands later, was left with only $1,935.

Here's the drama bomb: Barry didn't lose any money to snowball, though snowball was clearly up big in this brief session, and can be assumed to have cheated the other accounts. Barry mixed it up in several hands against snowball and lost sizeable pots only twice. The first time snowball won a big pot after sucking out a river flush against the odds while getting the chips in way bad in percentage terms. In the second hand, Barry raised pre-flop with K-Q, snowball flat-called with 4-3, looking for more action, and none came. The flop came 6-4-3, Barry bluffed by betting $850, and snowball immediately flashed a warning signal by min-raising with bottom two pair, all while presumably knowing Barry had air. It's hard to imagine a circumstance where a cheater, if trying to cheat Barry, wouldn't just call Barry's flop bet and let Barry try another bluff on the turn.

In the other hands in this file, assuming snowball was cheating during this stretch and had access to opponents' hole cards (which the huge profit for snowball suggests), the cheating account put in money bad against Barry while either drawing completely dead or against irrecoverably long odds.

Now, think about this: The best way to make a cover of this nature look real is to lose occasionally, on purpose. And how better to do it than against one of the most respected, intelligent players in the poker community, one who cold be presumed to be more likely than most to smell a rat under his very nose? After all, one pays the most attention to the hands one is personally involved in, and a bit less in other hands where one has folded early.

In this session, Barry lost his money against other players, and was well up here against snowball. It looks to me as though snowball softplayed against or cover-dumped to Barry during this session, and in some of the other files I've examined, I've seen similar curiosities.

I'll post some of these hands in the future, but the specifics of such wonders can wait for another post or two, as this should be enough of a drama bomb for today.

Or is it? There's still another dangler to wrap up:

Why All the Missing Hands?

When reordered as I did so above, it's clear that a lot of hands are missing from whatever UB sent out to Barry. Let's look at Group 5b from above, though I could have selected most of the groups above for this point. Here 'tis:

(Group 5b)
68720,966687_9173942_38994_38996.txt
68720,966687_9173942_39001_39003.txt
68720,966687_9173942_39005_39008.txt
68720,966687_9173942_39010_39013.txt
68720,966687_9173942_39018_39019.txt
68720,966687_9173942_39021_39021.txt
68720,966687_9173942_39024_39025.txt

We know the last two sets of digits represent the range of hands at a given table that were played, by hand number. But why all the gaps? Look at it this way:

(Group 5b)
68720,966687_9173942_38994_38996.txt [3 hands]
-- missing hands, 9173942_38997_39001 [4 hands]
68720,966687_9173942_39001_39003.txt [3 hands]
-- missing hands, 9173842_39004 [1 hand]
68720,966687_9173942_39005_39008.txt [4 hands]
-- missing hands, 9173842_39009 [1 hand]
68720,966687_9173942_39010_39013.txt [4 hands]
-- missing hands, 9173842_39014_39017 [4 hands]
68720,966687_9173942_39018_39019.txt [2 hands]
-- missing hands, 9173842_39020 [1 hands]
68720,966687_9173942_39021_39021.txt [1 hand]
-- missing hands, 9173842_39022_39023 [4 hands]
68720,966687_9173942_39024_39025.txt [2 hands]

Overall, Group 5B runs across a stretch of 32 hands, #38994 through #39025. However, six different clusters of hands,representing 14 of the 32 hands (in italics above) are missing. These hands took place on November 22, 2005 and all the mini-groupings featured Barry (crazyplayer) and others against "Crackcorn55" (ID #966687), another of the cheating accounts. There is no explanation for the gaps in the hands provided, yet as one can clearly see, these gaps extend throughout virtually the entire collection of hand histories.

Could UB please explain this?

Monday, January 18, 2010

Just Conjecturin’, Part 6: More on Excapsa: The Missing 2.6 Million Forfeited Shares

It's been a couple of months since I delved into the fallout from the long-running UB online poker scandal, and I left off at that time with a bit of a cliffhanger. Let's resume the discussion with two numbers, which are the total extant shares of UB-software licensor Excapsa, before and after the limited resolution of the scandal as published by the KGC:

Excapsa, 2006: 201,804,155 shares
6356095 Canada (formerly Excapsa), 2009: 194,880,883

The difference between the two is 6,923,272 shares, which seems to represent the total number of shares cancelled due to the scandal. Of those, 4,304,720 comprised a single share bloc solely owned by Russ Hamilton, the UB co-founder publicly outed in the KGC's findings as the primary culprit in the scandal. But what about those other 2,618,552 shares? In the court documents, these other shares were described as being transferred back to Excapsa, but it seems as if they were cancelled outright, perhaps as part of an agreement by other shareholders to not be publicly outed.

An additional complication is that not all of the original UB ownership is known. 56.4% of the original ownership was made public when Excapsa found it necessary to switch liquidators, and of late an additional 25% -- some 49.3 million shares -- has become public knowledge as that portion of Excapsa which was acquired by Joe Norton's Blast-Off Ltd. interests in late 2006. This leaves roughly 19% of the shares still unknown, though it is my personal belief that most of those are Phil Hellmuth's for reasons detailed in an early post.

Still, one can't help looking at the list of known share blocs and trying to piece together that 2.618 million. Click here to examine that list of the known 56.4%. Right... there's no direct match.

But before exploring other possibilities, let's note this: In terms of actively participating in the online cheating, any other owner with shares totaling more than 2.618 million can be presumed innocent. This includes Annie Duke, Phil Hellmuth (other idiocies aside), secret owner Ted Forrest, even Mansour Matloubi, whose reputed failure to heed a prop player's warning back in 2006 (as detailed in one of the lengthy 2+2 UB scandal threads), may have cost affected players several millions more. But being dense, perhaps even intentionally so, isn't the same as actively participating in the cheating.

From here on we have to enter the land of conjecture, meaning that these are opinions and theories only, and anyone with evidence or reasonable alternatives are invited to share. (And save the idle threats for someone who's likely to be influenced by them... and that's not me.) But in sifting through the tens of thousands of posts on the matter, I was intrigued by the references to the "Fred" account reported by posters to be intrinsically involved in the money transfers to and from Russ and other players. At least one other poster reported that this "Fred" was Fred David, which is likely the same "Fred J. David" listed as originally owning 1,680,000 Excapsa shares.

1,680,000, however, is not 2,618,552. Let's run with this hypothesis, though:

Unknown forfeited shares: 2,618,552
(less David bloc:) 1,680,000

Possible unknown shares TBD: 938,552

Ruh-roh... 938,552 doesn't match anything either. But wait! There's an entry for the TipTop Irrevocable Trust for 9,385,822 shares, which is almost exactly 10 times the 938,552 we seek under this scenario. It's a matter of 30 shares, which could possibly be explained by a transposition of a single number. Note that the share totals were obtained from proxy ballots where the share amounts were handwritten, meaning that such a single-digit transposition is not unreasonable to assume, or at least to speculate about.

Now, who says that these trusts have to have single owners? Perhaps 10% of the Tip-Top Irrevocable Trust belonged to Hamilton or another presumed guilty party as part of a group package, and it was forfeited along with the 4.3 million bloc solely owned by him at the time. It's at least a possibility.

What's more than just a possibility is that Russ was not the only cheater, and the KGC has refused to release the other 31 names involved. Is Fred David's name among those 31? I believe that on the basis of the math alone, that's a reasonable speculation. Then again, perhaps there is another bloc among that missing 19% for a similar 1,680,000 shares. Come to think of it, 1,679,970 would work even better.

I should also point out that even if all this turned out to true, it doesn't mean that David necessarily was an active cheater. Instead, his presumed actions (as reported in those lengthy 2+2 threads) may have facilitated the cheating, which might well be grounds for the shares' forfeiture.

Still, someone in addition to Hamilton forfeited his Excapsa shares. Maybe more than one someone. This is at least one possibility.

Mr. David? Would you care to refute this hypothesis?

Thursday, January 14, 2010

Back on the Horse

I'm finally working again, after way, way too long of a dry spell, and I will be catching up on some of those long overdue post topics. Stay tuned.