Tuesday, February 28, 2006

Online Poker and Cheating

I've found a lot of the reactions out and about in the blogsphere to the ZeeJustin and JJProdigy situation pretty interesting. Especially in regards to the possible repercussions it might have as far as the timing of the cheating coming to light, with efforts in the US to ban Internet gambling once again rearing their ugly heads.

I've never cheated at poker. I've created second accounts for rakeback deals, but I've always abandoned the original account, and played only on the second account. I've never discussed my play with others while playing in a MTT. I've never shared an account with others. I've never intentionally soft-played anyone, for any reason, whether they be stranger or friend.

If it's not clear from the above, I don't condone cheating, never have cheated, and feel pretty strongly that we should all be allowed to sling chips on an equal playing field.

All that said, cheating is inevitable in online poker. It just is. It's been happening for years and it's happening right now, as I type this. And the sad, awful truth is that there's nothing that can be done to prevent it, unless sites pretty dramatically change certain basic structures of the current online game.

Poker's growing worldwide popularity is also, ironically, likely a prime factor that encourages cheating at online poker. Organizing a well-oiled team that works together to net $3.75/hr for each team member by colluding in Party $10+$1 SnGs isn't an appealing activity for most bored teenagers in the US, but it's a much more attractive option for enterprising Madagascaran teens, where the average daily wage is $1. Just do a Google search on "click fraud India" if you'd like to see a real world example of the same concept in action.

Yes, in its blatant, idiotic forms (see ZeeJustin and JJProdigy), it can be unearthed and dealt with. Those cases should be unearthed and dealt with, and the poker sites of the world should definitely take notice and step up their own efforts to do so. Proactively, not reactively. As other people have pointed out, don't fall over yourself thanking Party and PokerStars for their actions in these recent cases, as all of the initial "investigative" work was done for them by others. Your security efforts shouldn't depend heavily on the propensity for young kids to act like dumb, young kids, and to tell other people about it.

If you have the ability to isolate multiple accounts playing from the same IP/MAC address (and you obviously do), then do it. Publish the results every week, as far as the multiple accounts you unearthed, investigated, and closed down. Include the total sum of money seized, and how you re-destributed that, either by bumping up winners in MTTs or by offering it as freeroll prize money. Make your security and policing efforts transparent. Players would love you for it and be more inclined to play on your site.

Honestly, though, while that's a good start, it's only going to address the most blatant and dumb forms of cheating. The very nature of online poker is always going to attract cheaters, as the potential payoff is high enough to encourage very smart, organized groups to basically pillage higher stakes games at will. And there's absolutely nothing anyone can do to stop it, if you want to be truly and brutally honest, unless you change the current way in which players sit at tables online.

How many players, working together, do you think you'd need to ensure a healthy profit for the team playing the $200+15's or $500+30's SnGs on Party? It's definitely debateable, but I'd guess as few as 3 players colluding via IM could guarantee a substantial profit, with all 3 constantly playing at the same SnG table. You wouldn't collectively profit every SnG you enter, and would have some complete washouts where your cheating efforts were all for naught, but overall I have to think you'd make a very healthy collective profit over time.

And they could easily practice it to the point that even a hand by hand analysis couldn't prove they were colluding, as they'd be good enough to avoid leaving obvious signs. Find 15 or so people willing to engage in the above activity, randomize the playing order so that different team members are in the same SnG together, and you've basically got a license to print money.

The same can be done in cash games, although it's a little harder, due to rake and the lessened ability to squeeze non-colluding players out of hands. Still very possible, though, and virtually undetectable, if done skillfully.

And if you don't think people aren't currently doing all of the above, right now, think again. If the potential to profit is there, people will abuse the systems. And they'll do so constantly and creatively, with a level of organization and sophistication that'd likely amaze you.

One glimmer of hope, though, is that sites could actually address a lot of the above pitfalls, if they're willing to re-think a few things, and take the time to educate players as to why they should accept what on the surface appears to be unpleasant change.

Ignoring cheating done by an individual creating multiple accounts (which can currently be policed and unearthed and requires no real change in operating procedure), nearly every other form of cheating involves collusion between players working together at a table. As long as you allow players to choose where they sit (and a window of opportunity for colluding players to sit at the same table), some degree of cheating is always going to occur. If done skillfully, this cheating is also unfortunately almost impossible to detect and eliminate.

There's a pretty simple answer to that problem, though. Implement a queue system that prevents players from choosing their seat, either in SnGs or at ring games. If you want to play a $200+15 SnG, you click a button to say so and are entered into a queue. When 30 players (or 50, or 100, or whatever number) are in the queue, players are then randomly assigned to one of 3 tables (or 5 tables, or 10, or whatever). If you want to play $10/20 limit, you get in the queue and are assigned to the first spot open.

On the surface, that sounds pretty drastic, especially for ring games. Proper table selection is obviously +EV, so robbing a skilled player of the chance to select his or her tablemates online will definitely impact their bottom line. It'd also likely lead to more churn, as some players would keep rolling the dice, getting in the queue, getting a seat, leaving when the table doesn't look juicy, re-entering the queue, getting a new seat hoping for better, and on and on and on.

Bottom line, though, it wouldn't be all that painful. And, ironically, it'd basically mirror what happens in brick and mortar rooms when you want to play poker. You get on a list and take the first available seat. If you decide you don't like it, you request a table change. You don't get to analyze 50 different tables minutely and choose the absolute most perfect one. You take what's available and, if you don't like it, take a shot on another table being better. Basically you'd just force everyone to take the approach that the lazy currently do, as far as clicking on "First available seat" at certain levels, and do away with the ability to get on waiting lists for particular tables.

Could you end up at the same table with a team member eventually, by persisently re-entering the queue? Yeah, you could. But it'd definitely be more difficult, and even when accomplished you wouldn't necessarily have the correct positioning to truly maximize your cheating efforts.

It also wouldn't address collusion at the highest stakes, where there's often rarely a waiting list for games and always open seats. Not sure what you do about that, other than tell yourself that someone playing 100/200 or above is a big boy or girl and can take care of themselves, if they're playing that level.

It seems more of a no-brainer for SnGs, though, as far as randomly assigning seats to entrants, instead of opening a table for anyone to sit down, and starting when the table is full. While it would be hard for low traffic sites to implement (you can't just assign players in the queue to a new table when 10 sign up, as colluding players would just sign up simultaneously, go into the queue at the same time, and be assigned to the same table in the end anyway), the bigger sites have enough interest that they could wait until 30 players had signed up before randomly filling tables, with no one waiting too long to play.

It's far from perfect, though, as you encounter problems at higher limits, where it's naturally slower to get new tables filled and started. And those higher limits are exactly where, by most accounts, you're most likely to encounter organized cheating. In the end, though, players might accept a 20 minute wait to play a SnG, if they knew it came with most assurance possible that they were playing in a fair game. I know I would, if I was playing that high.

Which, in the end, is just a whole lot of unsolicited babbling from me. I think the real lurking issue is that if online sites want to get serious about addressing cheating, they need to do more than simply comb through their data, looking for multiple accounts. There are some pretty fundamental, exploitable flaws in the very nature of how most sites offer games today that also need to be addressed, even if changes to them might initially cause some players to squawk loudly. Given the current structure of online poker, cheating is only going to get worse and more prevalent before it gets better.

4 comments:

SirFWALGMan said...

Excellent post and very good idea's.

CC said...

I agree with much of this, scurvy. To me, the real a-ha of all this rhetoric is the seeminly embryonic security and fraud analysis by the online sites. From an ROI/profit standpoint, I think the major reason that this lacks a sense of urgency by the sites is that the money being taken is from other players and not the house. I think that's probably why even in cash games similarly the overt collusion detection efforts are fairly minimal. For the online sites, the only real risk is dampening growth of current players or increasing the number of dormant players. There should be next to zero impact on the cost of new player acquisition unless this hits the mainstream media. I can't imagine there are many poker newbies lurking around 2+2, just us addicts.

ScurvyDog said...

CC,

That's a really good point, as far as a general lack of urgency from online sites thus far, since from their point of view cheating is almost a victimless crime, as long as it's not exposed.

Which is pretty apparent when you look at the other side of the coin, as far as some of the extraordinary efforts online casinos go to in order to try to crack down on people creating multiple accounts and abusing promotional bonuses, since all of that "cheating" takes money straight out of the house's pocket.

I guess it's a good thing that we have healthy competition in the current world of online poker, as some site will likely (hopefully) step up their security efforts in an attempt to differentiate themselves from the lackadaisical pack, which will likely have a ripple effect of causing all major sites to take security and fraud analysis efforts a little more seriously.

kurokitty said...

That's a real insightful post and all, but, um, where are the ScurvyRat pics???