I swear to God I want to make a debunk video about this topic, because everyone who plays online poker seems to believe this, and for good reasons: It happens. There's so many different sub-topics to talk about when discussing this topic.
OUR ABILITY TO ONLY REMEMBER CERTAIN HANDS
We don't have perfect memory. For example, you could play 10,000 hands, and you can't tell me off the top of your head what that distribution of dealt hands looks like. Our perception is skewed a bit, so we only seem to remember memorable things... makes sense.
WHAT EXACTLY ARE YOU REFERRING TO
If a
poker site would set up hands, what part of the hand would they set up? For example, the easiest way to gather more rake, if I was a corrupt poker site, would be to deal AA and KK in the same hand more often. So setting up preflop monsters is a good way, regardless of flops, turns, and rivers. I have yet to see, in my millions of recorded hands, setups like this that are happening more than the actual probability.
A better, more stealthy way to generate a bigger pot would be to create flops that give players monsters. This is a lot more difficult to accurately test. But again, my hands hit the flop pretty close in probability to a flopzilla calculation (it's skewed slightly because opponents have ranges Ex. An UTG open would slightly decrease the chances of an A hitting the flop (VERY SMALL PERCENTAGE). Let's say they open with a tight range of 66+, ATs+, AJo+, KQs. 58 combos have at least 1 ace as a hole card, out of 110 combos, as opposed to ATC, which would have at least once Ace as a hole card 198/1326. Even when an UTG folds, for example, it skews the probability in the opposite direction. Testing flops for set-ups is quite the difficult task. Turns and Rivers are a bit easier, as a set-up could be your opponent hitting a gutshot or something where both of you'd stack off, which is not as easy as the example I gave, but a lot easier than the flop.
NOT ENOUGH EVIDENCE
Everyone who believes that
poker sites set up hands to generate action to build the pot to get more rake doesn't have enough evidence. And there's different ways to gather evidence about this hypothesis, some of which are more precise and accurate than others. But here's the problem: it's quite hard to get enough evidence to prove this. I guess the most inaccurate way to prove this hypothesis would to look at average pot sizes of all the hands of a particular stake on one poker site vs the average pot size of all the hands of a particular stake on another poker site. This has so many flaws, but it's an example of one way to gather evidence.
SOFTWARE AND ALGORITHMS
I'm a programmer, and quite a knowledgeable one at that (if I may brag a bit). First off, most poker sites that are decent in popularity explain their RNG systems.
pokerstars randomization system is so advanced I don't even understand it; it shoots like quantum particles in the air that are random in nature and use that to determine their deck (IDK much about this). A lot of sites actually don't know what the flop will be preflop, according to these algorithms. They also don't know which of the remaining cards in the deck will be chosen for the turn until it gets to the turn, etc. This is different than an online slot machine, that when you choose "spin" or whatever, the outcome is already known, but the animations make you feel as though it's "spinning" or "calculating". These sites "allegedly" use pretty impressive CryptoAPIs.
THIRD-PARTY TESTERS
Also, there's usually a third-party that validates these claims; noncompliance would lead to serious lawsuits and shut down the site. The risk is too high to rig flops.
Yes, conspiracy theorists can deny the last two points I've made; after all, I didn't write the software. I didn't sit in the board meetings at poker site headquarters. I can't prove much with my set of hands I've captured.
Conclusion: So, we've all seen a couple of hands that look like clear setups. Our brains are horrible note-taking machines, and also horrible at math. You haven't taken into account all the hands played that didn't do this. Not only that, most of us conspiracy theorists haven't even taken an Introduction to Probability course where you'd learn concepts like variance. I've seen some pretty crazy percentages happen, because of the sheer amount of hands I've played. I've seen 3/3 people flop a flush. I've seen AA vs TT where AA ends up winning on a TT4 flop. I trust Intertops, PokerStars, Winning Poker,
888,
partypoker, and other sites in these calibers for their randomness.
EDIT: I just realized the best way a poker site could rig hands would be to give both players the nuts; it'd end up all in, and the pot would be split, giving both players their money back, minus rake.