After all the recents scandals (WPN etc), i come to theses conclusions:

As result of collusion between bots, legits players will always run below EV,

For example; On monochrome board , going allin with your flushdraw, russian colluder calls with toppair because his bot(s) folded two cards from the board's suit, And his algo told him to call,

So lets say your tracker show you should have win the hand 50% of the time for exemple, It is not the case if you play against colluders, you might have only 40% equity instead of 50% if you take in count the dead cards from bots/colluders,

Cliffs:

• Legit players will always run below EV in an colluding environnment

• Regs running above EV are suspicious

• Don't focus on the bots, focus on the people playing with them,

• Bots can be even, of loosing a bit, human with very diferent style/stats from them is winning a lot, and will nerver raise attention/never get caught (because none bots stats).

this is getting ridiculous, the only thing joey has brought awareness to is losing players deluding themselves thinking that they are not losing players.

You cliffs are almost the same length as your post

Great post. People need to face the reality sooner or longer that collusion between humans, and between humans and bots is now intrinsic to the game

and with Mafia owned sites like acr, probably in some cases sanctioned by the site

of course we'll have the same naysayers afraid to go out on a limb and look like a paranoid rec. you know, the high post count idiots who probably don't even play poker anymore , but profit off the industry in someway or are just cheaters themselves
(Becoming more and more, the same thing, but really, we knew that already didn t we?)

Trust no one in poker. haven't we learned this already, 100 times over? 5years out from black Friday and even "solid" players are desperate for a winrate so and money always wins over morals. Especially in the faceless online poker world.

Here's my post about collusion war nets i put up just yesterday

Ideally, shouldn't all-in EV already take into account the dead cards and only count the cards that are actually left in the deck? People make plays based on some impression of the distribution of folded cards all the time.

Interesting. I'm surprised I didnt think about this before.

of the six posts ITT, this is my favorite so far cuz it made me laugh.

I like to laugh.

what OP is talking about is above my pay grade but sounds legit.

There haven't been any requests to specify dead cards in piosolver so the colluding bots don't work on pio solutions.

just this

I always thought that people who complained about internet poker being rigged were just losing players.
I've had people in the past ask me if I wanted to collaborate with them, in essence, play together on different devices but in the same room.
In other words, knowing each others hold cards.
I always refused and just shrugged it off as " oh well, they can't win legit that's why they resort to this " but it never really bothered me.
Now, I remember watching old YouTube videos of what was then aspiring pros, who many now have become well known pros, sharing an apartment together, all sitting in front of different computers at the same time, and all the while, bragging about how they were crushing the game.
I now realize, they maybe didn't think of it that way, but they were actually cheating.
Well, I've long stopped playing on the internet any way.
Who wants to spend all day locked in their house in their underwear!

It is not about being loosing player but apparently being unlucky

You got me



Ev would be more accurate if the HandHistory show folded cards, yes.
But you would know perfectly how opponent build ranges (pre and post) , the more information available, the more the game become 'machine friendly'.

Subbed bc I’m a bot

OP's points are legit with certain assumptions.

If this is true then Stars must be riddled with colluders. Only site I've always run way below AIEV on.

NickMPK is correct. AIEV is not even well defined in multi-handed games without full holecard knowledge.

Even a very simple example preflop: 20bb effective 3 way, BTN opens AQo, SB 3b shoves 33, BB folds and BTN calls. This is not the same as HU all in with the same hands and 1bb dead money thrown in DUCY (I would expect BTN to have a bit more equity 3 handed)?

There are biases depending on the way you define all in EV regardless of the presence of collusion.

Sure, and in 6handed, 9h games card removal have a big impact on the AIEV as well, but i think it is not relevant to my point no?

It's not very relevant.

But a site with full holecards should be able to find better ways to search for colluders than to come up with a definition of AIEV discarding some of that info, and look at a summed statistic for each player.

From the players POV it's not that bad- but you need to find enough players who are willing to share their EV with each other, including fish who may run above/below ev.

The AIEV-based method of bot detection is irrelevant at Stars because the company has more advanced algorithms for this. However, it's a useful heuristic for the regs of those networks that are too slow to catch bots, in order to avoid sitting at the same tables with the suspicious players until the latter are banned, which may take long or even never happen.

In 2014, a colluding bot ring was discovered by a reg in iPoker PLO, and one of the main giveaways was that those accounts were running enormously above EV, which was easier for them to do in PLO, where they had twice more hole cards to share than in NLHE.

As poker all-ins are zero-sum once the rake is taken, the fact that colluding bots run above EV implies that the humans competing with them will run below it.

as a bots network owner, i can confirm as i'm reading through my 10+ million hands collected.

groundbraking, collusion is a thing.

Neat idea, and once proposed it seems like it must be true. Nice one OP.

I’d guess the sample size would have to be absolutely massive to prove anything. You could combine databases to see if a group of players collectively run under EV vs someone, but I think you’d need much more. Could a site do it with all their data?

Well its much different scenario like these:

1. They are dead cards and NOBODY KNOW WHAT THEY ARE [Except the folder] - There is zero advantage to any other player in this scenario.

2. One player knows the folded cards via collusion - that is whole different story. i.e. the player with top pair knows his bot or buddy folded two of the likely flush cards [DD] on say a DDC flop. Now the flush draw is only 8 outs twice - not 10 outs twice or about ~34% Equity verses ~42%. That is a huge advantage.

Flush draw has less outs against him, but it is also less likely that he is against a flush draw and more likely he is against a set or two pair. Now what?

AIEV line in trackers doesnt work because it doesnt take cardremoval effects into account. They just assume that previous folders hand ranodm cards , even AA...If you get all-in SB vs BB with AK versus 77 in a fullring game, its more likely that the deck contains more A,K, Q and less low cards so the AK is actually a favorite. But the tracker doesnt account for that and the AIEV is incorrectly calculated.

That you are not doing as well as the software might predict does not show collusion. It could be that your play is not as good as you are the software thinks it is.

I agree with the comments and stay away from ACR. Collusion, bots, superusers etc. are possible on any site.

Just gonna state the obvious and point out that victims of collusion are not always running below EV. Victims of collusion would run more below EV than if there was no collusion, but this does not automatically make them guaranteed to run below EV. See the difference?