RGP post on online collusion

[MS Sunshine]

There is a post on RGP about two players using collusion, for about 5000 hands, at an internet poker site. The site is unnamed, but it looks like PokerStat was used for the info they gave, making Paradise the most likely site.

They used some fixed rules to avoid much active colluding and getting caught. They would fold a pocket pair, nines or below, if the other player had one. Big card conflicts were settled by one going all-in preflop, by disconnect. Other than that they played best hand. They also reduced the number of hours they played together by 2/3rds, 2 hours of single play for each hour of team play.

They played mostly $3-6 and $5-10 Hold'em and both were 1 BB/hr winning players before trying this for 5000 hands.

Their results:

Without playing in collusion;

Player1 played 11716hands with a winrate of 0.9 BB for every 50 hands
played

Player2 played 14913hands with 1.4 BB for every 50 hands played

With teamplay:

5097 hands were played

Player 1 won 292 BB=2.9 BB/hr
Player 2 won 201 BB=2BB/hr

It would seem that two competent players can each increase their win rate by about a bet and a half per hour by colluding in the mid-limit online hold'em games without getting caught by the software. With the single play counted in, it seems to add .8 BB/hr to each of their win rates.

This assumes that this was not a troll post.

MS Sunshine

[GrannyMae]

This assumes that this was not a troll post.

Sunshine, i have not seen the post yet.. i will look later. however, what came to my mind while reading your recap was:
"why in the world would successful colluders post their results?"

i fully understand the human nature of bragging, but for the life of me i do not comprehend what they could gain by this..

perhaps i need to see it in context..

thx for the heads up.

[MS Sunshine]

"Once ( and we realy thought security-software would catch us, but it didn't) we both held 88 in late position with 4 limpers already. Normaly a clear raise. We both folded. Being disconnected (mostly preflop) was one of our favorite. We were surprised by the fact how few situations realy come up where we could start a raising-war to trap somebody."

This was part of the post. From a few weeks ago this is a 6010-1 in a ten-handed game. You would think the poker sites should be able to check these hands to see if they can track these two down. Sorry, I enjoyed your post, but I hope you get caught.

As for the all-ins, why didn't this ring any bells?

It's just a matter of time before 60 minutes or someone like them shows on TV how easy it is for this stuff to happen and the software be clueless.

This kind of stuff costs us money in two ways. Directly by cheating us and by running players off when if it is shown that the software does little to caught cheats while seeing everyone's holecards.

MS Sunshine

""why in the world would successful colluders post their results?"

i fully understand the human nature of bragging, but for the life of me i do not comprehend what they could gain by this.."

Granny is clearly not properly equiped below the waist to understand this very basic principle.

Quantifying the collusion tax, a real breakthrough. It's right about what I expected except I assume that less careful cheats will win at greater rates for shorter periods of time. 4+ BB/50 hands going to the cheats then another couple going to the rake will make any such game unbeatable.

Gamble up!

[GrannyMae]

Granny is clearly not properly equiped below the waist to understand this very basic principle.

enlighten me honey

i read the post on RGP, and all i can figure is that perhaps this poster is looking for critiques on his alleged process, so that he may be successful at what he claims to have already done. my instinct is that he is full of [censored], and has not done it <font color="red"> YET</font color>

it is obvious, due to the language of this post, and the reference to 8-16 games that the poster is referring to paradise. if this assumption is correct, i would hope that, as MS Sunshine suggested, paradise has looked at all the recent 88 vs. 88 battles. if these two have indeed done this, and the site is indeed paradise, then it will be easy for paradise to identify them.

that being said, i would still like "anonymous" to share for those testicularly-challenged people what the advantage of this post would have been.

[mdlm]

There is a trivial method to detect this sort of cheating: Simply compute how often two players are at the same table, compare it to the average, and flag for further investigation if time-at-same-table is unusually high.

For example, if there are ten identical tables then the probability that two players will be at the same table is 10%. If two players are at the same table just 30% of the time for even 100 hours, this should be flagged and investigated.

There is no need to look at how hands are being played to detect this sort of collusion.

If Internet sites are not doing this that would be truly amazing.

[kamrann]

Friends?
People who go to the same table through habit?
I really don't think it is that simple.

What about someone who thinks someone else is a fish. They may purposely look for their victim and even follow from table to table. That would skew the statistics so they may look like colusion. [img]/forums/images/icons/wink.gif[/img]

[mdlm]

First, if the filter that I suggested is triggered then hand histories should be examined (as I suggested in the original post).

Second, any test for cheating will have false positives. Inadvertently preventing friends from playing at the same table is a small price to pay for significantly reducing cheating opportunities.