How online gamblers unmasked cheaters(news.cnet.com)
news.cnet.com
How online gamblers unmasked cheaters
http://news.cnet.com/8301-1023_3-10110040-93.html?part=rss&subj=news&tag=2547-1_3-0-20
8 comments
Me and 2 friends won a $13,000 (7000 GBP) car in 2000, sold it and split it 3 ways. It was an online competition. First off, you had to press keys on the keyboard, while moving the mouse in a perfect circle for as many hours as possible. We stayed up 40 hours, put a macro on the mouse, took our turn at the computer hitting keys. Then split the prize when we won. When we got to 40 hours, beating everyone else, that was a very jubilant 6am.
Awesome story :) I could so see this being spiced up into a Coen brothers movie. Instead of giving up after the admin posts an unbelievably good time, you call and blackmail him into sharing the prize money. After a lot of fuming he figures it is still better than getting fired and giving the money back. However, your friend has a crisis of conscience, so you plot together to "eliminate" him...
A Fistful of Euros...
I've followed this since it started; it has understandably caused a great amount of anxiety in the online poker community. Almost everyone was happy that the cheating was exposed, but I think most players would have preferred it if the 60 Minutes episode had not aired.
Prior to the 60 Minutes piece, knowledge of the cheating was limited mostly to online poker aficionados. Now that it has received national media coverage, casual poker players everywhere are likely to hear about it. People who were already hesitant to play because of online poker's grey legality and funding difficulties are not going to be swamped with "Oh, you know online poker is rigged. I saw it on 60 Minutes." Combine that with the waning interest in televised poker and the UIGEA legislation and you've got a bleak outlook for the future of online poker.
In the end, I think the 60 minutes episode will deal in a much greater blow to online poker than the money the cheaters stole.
Prior to the 60 Minutes piece, knowledge of the cheating was limited mostly to online poker aficionados. Now that it has received national media coverage, casual poker players everywhere are likely to hear about it. People who were already hesitant to play because of online poker's grey legality and funding difficulties are not going to be swamped with "Oh, you know online poker is rigged. I saw it on 60 Minutes." Combine that with the waning interest in televised poker and the UIGEA legislation and you've got a bleak outlook for the future of online poker.
In the end, I think the 60 minutes episode will deal in a much greater blow to online poker than the money the cheaters stole.
Here are some more links.
The 60 minutes segment:
http://www.cbsnews.com/video/watch/?id=4639016n
The Washington Post articles:
http://www.washingtonpost.com/wp-dyn/content/article/2008/11...
http://www.washingtonpost.com/wp-dyn/content/article/2008/11...
A summary of the original Twoplustwo investigation into the UB scandal. This is much more detailed and accurate than the reports linked above:
http://forumserver.twoplustwo.com/29/news-views-gossip/ub-sc...
The 60 minutes segment:
http://www.cbsnews.com/video/watch/?id=4639016n
The Washington Post articles:
http://www.washingtonpost.com/wp-dyn/content/article/2008/11...
http://www.washingtonpost.com/wp-dyn/content/article/2008/11...
A summary of the original Twoplustwo investigation into the UB scandal. This is much more detailed and accurate than the reports linked above:
http://forumserver.twoplustwo.com/29/news-views-gossip/ub-sc...
Interestingly these people would never have been caught (at least not by other players) had they been smarter about it. They could have made 100x what they did over a long period of time.
I've thought a bit about how I would do it if I had that opportunity and were trying to maximize profits without getting caught. It's an interesting thought experiment really. I'd probably do something like have preset starting hand standards that are on the loose side of what a winning player might have and stick to them religiously. Then I'd largely only use the cheating software on the river. I wouldn't call bluffs much, though I'd raise all-in against them frequently (since presumably the bluffer would fold and nobody would then see your hand). Etc.
I've thought a bit about how I would do it if I had that opportunity and were trying to maximize profits without getting caught. It's an interesting thought experiment really. I'd probably do something like have preset starting hand standards that are on the loose side of what a winning player might have and stick to them religiously. Then I'd largely only use the cheating software on the river. I wouldn't call bluffs much, though I'd raise all-in against them frequently (since presumably the bluffer would fold and nobody would then see your hand). Etc.
I think even if you tried, your PokerTracker stats would be abnormal, even if you only used your advantage sparingly. For example, the river strategy you proposed was what resulting in the monstrous river aggression factors that identified many of the accounts in question.
It'd be tough -- agreed that its an interesting thought experiment.
It'd be tough -- agreed that its an interesting thought experiment.
They would be abnormal, but not egregiously so, especially if you were careful to change accounts periodically, or perhaps just paid off a better hand a few times for posterity. What tipped them off in the first place was people playing every hand preflop in a tournament and winning.
I can attest first-hand to the uselessness of the Kahnawake Gaming Commission. I was involved about a decade ago in proving that the blackjack games dealt by a couple casinos was rigged. Someone much smarter than me proved conclusively that they were dealing seconds, meaning that if the dealer's card would have busted him, instead it was digitally tossed out and replaced by the next card, totally unbeknownst to the user of course.
The KGC was given a record of hands that any stats professor would have been able to use to verify it. They did nothing.
The KGC was given a record of hands that any stats professor would have been able to use to verify it. They did nothing.
Here is the 60 minutes 10 minute clip
http://www.cbsnews.com/video/watch/?id=4639016n
Im an avid viewer of 60 minutes, 48hours & CBS evening news since they put them online in a simple Hulu like format!
http://www.cbsnews.com/video/watch/?id=4639016n
Im an avid viewer of 60 minutes, 48hours & CBS evening news since they put them online in a simple Hulu like format!
I wonder if a provably unbiased distributed gambling application cannot be made so as to guard against the potential for such attacks.
Generating random numbers in a distributed and uncheatable way is already possible, I wonder what is holding this back.
Generating random numbers in a distributed and uncheatable way is already possible, I wonder what is holding this back.
The problem isn't bias -- the problem is one actor having more information than the others, earlier than the others. There is no evidence that the RNG was anything other than fair.
A related attack on poker is just stuffing a game with colluding players. You don't have to know player A's cards to beat him in the long run if you know the cards of B, C, D, and E (and, if you want to get really tricky -- they can know yours, too!).
Plus, given that this is online poker, you can have players B, C, D, and E be bots, programmed to talk to your server to get the odds, throw enough hands to look reasonable, and bet big only once in a while when their collusion means The Syndicate rakes in the moolah.
Example: the flop is 5, 6, 7 with 3 different suits. The best hand from The Syndicate is pocket 7s, which is (poker amateur hour commences) pretty strong but not unbeatable. Everybody is out but one player. The next card is shown, and it is a 3. There are two worlds the game could be in now: the world where there are fours available, in which case the adversary might have one. Then there is the world with no fours available, in which case The Syndicate has the best possible hand and will only lose by a freak accident.
The adversary goes all in. The Syndicate checks with the other TEN cards from the deck they saw, notes that they've seen three of the 4s already, and calls, with a smile on his face. (Again, amateur poker player but I'm thinking this is a 95%+ win for the syndicate.)
A related attack on poker is just stuffing a game with colluding players. You don't have to know player A's cards to beat him in the long run if you know the cards of B, C, D, and E (and, if you want to get really tricky -- they can know yours, too!).
Plus, given that this is online poker, you can have players B, C, D, and E be bots, programmed to talk to your server to get the odds, throw enough hands to look reasonable, and bet big only once in a while when their collusion means The Syndicate rakes in the moolah.
Example: the flop is 5, 6, 7 with 3 different suits. The best hand from The Syndicate is pocket 7s, which is (poker amateur hour commences) pretty strong but not unbeatable. Everybody is out but one player. The next card is shown, and it is a 3. There are two worlds the game could be in now: the world where there are fours available, in which case the adversary might have one. Then there is the world with no fours available, in which case The Syndicate has the best possible hand and will only lose by a freak accident.
The adversary goes all in. The Syndicate checks with the other TEN cards from the deck they saw, notes that they've seen three of the 4s already, and calls, with a smile on his face. (Again, amateur poker player but I'm thinking this is a 95%+ win for the syndicate.)
Your point is valid, translating existing games which are built on a set of assumptions (physical presence -> unique non-communicating players, physical deck + shuffling -> unique cards and true randomness) may not be translatable to an online experience in a sustainable manner.
I was refering to the general problem of not being able to trust the single point of control, which applies to all types of online gambling. In this context, I was wondering why games that can be gambled on have not appeared that work in a fully distributed manner.
I was refering to the general problem of not being able to trust the single point of control, which applies to all types of online gambling. In this context, I was wondering why games that can be gambled on have not appeared that work in a fully distributed manner.
Because it doesn't solve the cheating described in the post you replied to. Distribute all you want but if 5 of 6 players are colluding player #6 is gonna be screwed.
there is a way to derrive an unbiased random number, even if everyone else is cheating. take this:
http://everything2.com/title/Coin%2520tossing%2520protocol
and extend for N participants and results from 1 to N.
Given this, a lot can be done.
Given this, a lot can be done.
This is a great story, the only sad thing is that most of the affected high-stakes players still do not have their money back, and the companies involved still maintain their innocence blaming 'previous management' for the scandal. The latest news is available on the popular twoplustwo poker forums.
The hack will make you laugh : right click on the flash app, select "Stop", and enjoy unlimited time to answer the questions instead of the 10 secs countdown. Then, when I was ready to submit the answer, I'd change my Windows clock time back in time since you'd have to be the fatest player as well.
But the interesting part is how I started playing this quizz. One night, very very late, I bumped into the site and I decided to try the practice mode, wich was free. I started the quizz, and I had a surprise when I discovered that a small white asterix showed next to the right answer every time. I thought that it was dumb even if it was the practice mode and I suspected something wierd was going on, so I immediatly started a real paying game to verify my suspicions. Again, a white asterisk would sit next to the right answer.
You'd have to answer 5 grids of 10 questions successfuly to make the best score, then the winner would be the overall best time. I managed to finish the 1st grid, then in the middle of the second one, the white asterisks suddenly disappeared. This is when I understood what happened : the admin was probably cheating the game, and he made a mistake wich made the asterisks visible by everyone instead of himself only.
Still, the game was under French juridiction and supposely enforced by an "Huisser", like every other game in France. My bet is that the huissier didn't understood jack about computers and didn't care at all about this internet quizz.
I talked about this quizz to my best friend and we quickly discovered the exploits. We figured out that if we set the Windows clock back in time before I even started to answer the question, the final time per question would be a couple of hundreds milliseconds, varying from 50-300.
So I decided I would take my chance against the admin and try to win the 1000 euros, wich was plenty of money for a teenager. It was still a lot of work because most of the questions were insanely difficult and even with Google it was often hard to find the answer. As a matter of fact, the quizz was plain impossible to answer for a human.
Nevermind, with a lot of patience and a couple of days, I managed to answer the 5 grids with an amazing final time, I figured out that if the admin was cheating, then he couldn't say anything. The game would automatically ends at midnight every Sunday and declares the winner. I did the last grid on Sunday evening so that I wouldn't appear in the high scores too early and the admin wouldn't suspect anything.
I watched anxiously the table of scores and as midnight approached I started to relax, nothing unusual happened. Then suddenly, at 23:45, 15 mins before the game would end, a player appeared out of nowhere in the scores and was quickly climbing the ladder. I knew it was him.
When he finaly ended his fifth grid, he reached the top scores, but only took the second position because I had a best overall time, so my plan worked perfectly as I knew he wasn't prepared for this. A couple of seconds later, the game ended and I was declared winner. I was given a phone number I had to call to claim my reward.
On the next day, I dialed up the number. When he picked up, you could almost hear the spider webs being removed from the phone :
- "Hello ?"
- "Hi, I won the game yesterday, and I was given this number to call."
- "Ohhhh, it's you", his tone was priceless. You could guess I was the first person ever claiming the reward. At this moment, I knew he knew and he knew I knew. He asked for my details and told me I would receive the check soon.
I tuned up to the radio show in the afternoon, and waited for my minute of glory. "And congratulations to our internet winner of the week, , living in , for winning the 1000 euros". Good.
A couple of days later, I received my 1000 euros check and I went to see the updated list of winners on the site. Unsurprisingly, I was the only one listed with a real name instead of a nickname and an actual town showing in the details.
The following week, I decided I would beat the admin again, for my best friend this time. I tweaked my tactic a bit so I would valid the last grid in the last minutes. It wasn't as easy as it sounded since the pages had an expiration time, but I figured it out.
This time, the admin figured out he should make a better score and his final score was much better than the last week. Still, I had predicted it and when I valided my last grid with a nearly perfect overall time (50 questions in about 20 secs), it was too late for him and I won again.
My friend called and he received his check. The rules wouldn't allow us to participate again before 6 months, but I knew there was a time to stop, and we decided to never participate again. Moreover, I had beaten the admin twice and that was my main satisfaction, and I knew I had played all my cards and beating him again would be very difficult if not impossible.
The third week, I went to check the winner and his overall time was indeed ridiculous. It was clear the admin was unbeatable from now. Well, in one sense only, because a couple of weeks later, the winner prize turned from the 1000 euros to a ticket to participate in the actual radio show. I laughed hard, because I knew I forced the admin to make unbelievable times and they figured out he was cheating. I don't know if he was fired, but he certainly didn't have a 4000 euros monthly bonus anymore.
Since this time, I have a different look on this kind of games, especially when there is an admin behind. Moral of the story : Poker is no exception, and if you cannot fully trust the admin, even with the best legislation, you cannot trust the game either.