Issue with Amigotechs Jacks or Better 50-line software

[Eliot Jacobson]

Did they respond to your findings, Eliot?

I had a long conversation with their CEO yesterday. They acknowledge that their Jacks or Better 50-line software malfunctioned in the way I described between December 15 and December 28, 2011.

I was told that they have a number of fixes that were put in place for security reasons because people hacked their flash software. Such hacks are constant and are a challenge for all such companies. They are conducting an internal investigation to determine if it was a patch that went wrong, or if there was an employee who purposely did this, or if there was some other cause. I asked that they make a public statement once they have concluded their investigation.

As for their other games and huge rebates and bonuses, there is a simple and non-rogue reason the games may lose quickly. The software is specifically designed to have the ability to have poor pay tables if the client requests. This option is there so that their client casinos have a lot of room for rebates and bonuses. For example, the Jacks or Better pay table they use at Youwager has an RTP under 95% (most are above 98%). I cannot argue with this business model. Many players love the bonuses. In the end the result is the same as good pay tables and smaller bonuses. Players may have the sense that they lose quickly there after they cash their bonuses. Some may blame this on conspiracies about their programming. However, there is no evidence for this and I do not believe this is the case. The source comes from the poor pay tables.

[spiderlegz]

FTR - here is a list of affected casinos and poker rooms:

SBG Global Casino
Pamper Casino
Poker Host
AC Casino - Always Cool Casino
JustBet
BetGuardian
Tropical Cash Casino
Pro Poker Evolution
Hard Poker Club
Pepita Poker
Poker Face
Poker Texas Club
PokerRooM.eu



AST - well, that's a blast from the past! I thought they were defunct. AST had some serious issues. Search the site and you'll see what I mean.

Same crap but in different wrapping. All are basically the same software (AST/Amigotechs/ThrillX). Differences are maybe like MG Thumper and Viper.

[jstrike]

I just came across this on the Wizard's site, and my first thought was I wouldn't wanna be them, being audited by Eliot. That'd be pretty embarrassing. I was going to say -- the short period of time between when this problem surfaced in the logs and when it was discovered makes me think it probably wasn't intentional, though. It was so obvious, it doesn't make sense for it to be intentional, if you subscribe to the theory that people only cheat when they think they can get away with it.

More than likely it was a tiny mistake with huge repercussions, when someone uploaded the wrong file. When you're dealing with hundreds of source code files for a big software project, with hundreds of revisions and versions and bug fixes on each one, it's sometimes possible to screw up and put the wrong file to the live server.

Back when I was testing my software -- before I took deposits, when I was just handing out cash to beta testers -- I ran into a problem with the Craps game not showing the win animation sometimes. So while I was tweaking the flash animation, I set the back end (on the local version, on my own computer) to ignore the RNG and just roll 7's every time. That way I could test the win sequence over and over, and see what was going on. Unfortunately, I forgot to change it back to the RNG before the next time I uploaded a software patch, so the next guy who came along was the luckiest guy in the world, he sat down and rolled about a hundred consecutive 7's on a $5 table. For real money.

So as a programmer, you have a lot of pieces of code for testing that do all kinds of weird things you'd never upload to the live server. Having said all that, this means they had a piece of code floating around, for whatever reason, that double-deals cards. That's not a very complicated piece of code to write. On the surface it doesn't make a lot of sense for testing purposes, but consider this -- this code doesn't just double-deal. It keeps dealing until there's definitely no win. True, that's a very similar piece of code to one that double deals and lives with the result of the second card, but it's both more complicated to write (because it's iterative), and more detectable; I mean it's blatantly obvious.

Here's my guess...and this is my best guess without ascribing any malice to them, because like they said about every President since FDR, never assume malice when you can assume incompetence: When you write a 50-hand poker game, you're taking one bet and splitting it 50 ways. You could test it a lot of times, maybe thousands of times, without ever seeing a situation where not one of those hands paid off. Somewhere in your payout code, you're going to have to take the initial bet and divide it by the number of winning hands to get the score. If the number of winning hands is zero, then you're dividing something by zero... and when you do that in a computer program, the program throws an error and crashes. You can't divide by zero in most programming languages without the program crashing.

*edit: The right way to do the actual payout would be to take (x/50)*bet, and I can't really imagine why they'd have something divided by zero, but maybe it was a second operation for a bonus feature or something else.

So this piece of code was probably written specifically to find out if their game was crashing in those rare cases where no hands paid off.

There was probably an easier and better way to do this, like short-circuiting the payout routine instead of screwing with the cards, but my bet is this came out of a crash they observed, and then they uploaded the wrong file after testing it. Mass incompetence, yeah...but too stupid to be intentional.

[Eliot Jacobson]

This issue has been resolved. The issue raised in the original complaint was caused by malfunctioning and poorly written software. Specifically, software designed to detect and deal with certain forms of player cheating began operating in normal play mode. Amigotechs had neither malice nor intent.

The quick version is that their software was being abused by certain people playing JOB50. Some players discovered that they could win multiple times on a single hand. They did this by closing a session in pause mode and then re-opening it in multiple browsers at the same point, getting to play out the same hand from the draw. Amigotechs put in a fix for this attack, and it was the fix that went wrong.

The full version is here:

http://wizardofvegas.com/forum/gambl...tions-warning/

[TheLastCylon]

The software is specifically designed to have the ability to have poor pay tables if the client requests. This option is there so that their client casinos have a lot of room for rebates and bonuses. For example, the Jacks or Better pay table they use at Youwager has an RTP under 95% (most are above 98%). I cannot argue with this business model. Many players love the bonuses. In the end the result is the same as good pay tables and smaller bonuses. Players may have the sense that they lose quickly there after they cash their bonuses. Some may blame this on conspiracies about their programming. However, there is no evidence for this and I do not believe this is the case. The source comes from the poor pay tables.

This part of your post really blew my mind. I kind of knew that this was happening but this confirms it. Is this understood? Why would people willingly accept stupid terms and conditions and a crappy pay table? The end numbers are the same. the probability of losign money is the same. It's like buying a car for ONLY ONE DOLLAR, but the tires cost $19,999.

You can't argue with this business model, but I can. Am I missing something?

[jstrike]

This part of your post really blew my mind. I kind of knew that this was happening but this confirms it. Is this understood? Why would people willingly accept stupid terms and conditions and a crappy pay table? The end numbers are the same. the probability of losign money is the same. It's like buying a car for ONLY ONE DOLLAR, but the tires cost $19,999.

You can't argue with this business model, but I can. Am I missing something?

Well... I think the point about adjustable paytables is as long as they're putting the payouts where the player can see them, and it's not changing in the middle of a gaming session or paying less than it says it will, then it's the player's responsibility to shop for a good deal. So in that way they're not ripping anybody off.

Personally I was a little confused by part of this. Is double- or triple-dealing until all 50 hands lose supposed to happen when someone opens more than one window? Like, that's their defense mechanism? I don't get how that can be considered a good countermeasure against hackers. What it sounds like is that Node A on the server can't check to see if the re-deal has already been done by Node B. If that's true though, how do they decide when to deal 50 losing hands? It doesn't make any sense, unless they're doing it on a timer or something. But if it's true that their socket processes have no way to communicate this with each other, then the pause hack described will probably still work for any game where a player has to make a decision after a bet's in play.

While it's kind of understandable that a game developed for single-window use would be lazy about bothering to check for rogue windows, why go to the trouble of creating this crazy 50-hands-lose system to mitigate it...and then why deploy that system if it's obviously broken? The right way to fix it would be to store an InnoDB table of current hands with their current status, and check that table when a player makes a request. Sure, it's more overhead, might take an extra 100ms for each re-deal and lower a server's concurrency by a few players, but it's the only way to be safe. You can't just assume that every thread is talking to a unique client. Set up a tiny database, lock the row for update while the re-deal is happening, and if the player asks for a redeal on a hand that's already finished, just don't deal it. Close both windows, kick and ban the player. Problem solved. This can work fast with row-level locking, and assuming you're regularly flushing finished hands from the table. You can shard it to each server. It might even be faster as a MEMORY table, depending on the load and how often they're dumping it. Takes less time to code and works for all games, not just multi-hand VP.

[TheLastCylon]

Well... I think the point about adjustable paytables is as long as they're putting the payouts where the player can see them, and it's not changing in the middle of a gaming session or paying less than it says it will, then it's the player's responsibility to shop for a good deal. So in that way they're not ripping anybody off.

I'm not accusing them of ripping anyone off. They are playing by the rules. Their software is apparently a disaster. But they ARE playing by the rules. My shock was more in the rules themselves.

The fustercluck in their software happened because of patches with their paytable, but the variable paytable exists because of what amounts to a game of numbers. Yeah, that's what all gambling is, but instead of simply putting the numbers up front, they hide them under bonuses. And then, if the casino uses bonuses, that means their paytable is based on bonuses, and that means that NOT playing with the bonus is mathematically stupid.

I'm attacking this, but is this completely normal and I'm just being an ass?

[jstrike]

I'm not accusing them of ripping anyone off. They are playing by the rules. Their software is apparently a disaster. But they ARE playing by the rules. My shock was more in the rules themselves.

The fustercluck in their software happened because of patches with their paytable, but the variable paytable exists because of what amounts to a game of numbers. Yeah, that's what all gambling is, but instead of simply putting the numbers up front, they hide them under bonuses. And then, if the casino uses bonuses, that means their paytable is based on bonuses, and that means that NOT playing with the bonus is mathematically stupid.

I'm attacking this, but is this completely normal and I'm just being an ass?

It's become normal, which is kinda sad. Used to be only porn actors cared about who had the biggest bonus. The arms race now to see who can throw out a higher percentage bonus has pretty much ruined bonuses, and it's screwed up a lot of non-bonus games too. What's sad is that it reflects sound marketing principles, which in turn reflect the way players think. People get a coupon and for every one who reads the fine print, there are two people who just see "2000%" and think "oh wow, I gotta try that". So the real numbers game is in how many new players they can get. Which gets more and more important the more their loyalty programs suck, and the more players leave them after the first deposit. It's turned into a vicious cycle.

Variable-paytable games are pretty standard, in and of themselves. It's easy to reconfigure a real Video Poker game down from 9-6 to 8-5 if a brick and mortar casino wants to, and they do that all the time. But the drain on casinos that focus on wild bonuses can really impact the games for straight-up players, because basically you're paying to make up for the bonus grinders.

Just ballpark figures, so don't hold me to this: The only two ways a casino can not lose money against a perfect blackjack player with a 100% bonus is to either make the WR 65-70x including deposits, or to set it up so only around 10% of blackjack bets count towards WR. That's just to break even. For a good-to-average blackjack player, say with a 98.25% RTP who knows a little bit about bonus advantage play, the casino will always lose money with a 30x WR including deposits. How much they lose depends on the size of the bonus. It's true that blackjack's a really high-RTP game, but perfectly played 9-6 Jacks or Better isn't that much lower. It's still barely break-even for the house with a 100% bonus, and you can imagine how bad it gets with 1000% bonuses flying around.

Legit casinos only have two basic levers they can pull to increase their earnings. One is bonuses, comps and freerolls, and the other is pay tables. As a software maker you have to think some of your operators are going to take the bonus marketing approach, and others won't as much, and allow for the pay tables to be flexible according to their needs. I personally think there are still a lot of players out there who appreciate a small, flat $10 bonus with no WR if they know they can get a solid RTP on the games...and I think a lot of people here would agree...but CM attracts more serious players, so people's preferences here don't necessarily represent 90% of players who don't know what RTP stands for.

[TheLastCylon]

Thanks a lot for the response Jstrike. I'd thank it if I could.

Well I am a big table player, and the few people I know who play online games are all table players. We love blackjack. I know that they don't like bonuses because they complain about them. Actually, they are the ones who introduced me to the concept of bonuses.

I know that we are not perfectly representative of the gambling community at large, but we definitely represent <i>something</i>. I hope that some casino decides to take a chance and completely eschew bonuses in favor of just straight-up GAMES. I don't want to have to read fine print.