I'm going to limit the scope of the non-slot games to what could be called "the standards" - Blackjack, Video Poker, Table Poker(s), Craps, Roulette, and like that. I won't be including in my mental considerations any of the vast variety of obscure game variants, custom games, and so forth.
I reread Part 1 a few times over the past couple of months. The content for this Part 2 post has been percolating in my brain for awhile now as I've gone about the work of collecting links to some good examples for the points I'll be raising. In doing this, I have observed in myself what some might label a bias regarding online Casinos. That bias is:
If the Casino thinks that they can cheat you and get away with it, then they will cheat you. They will cheat you to the extent of their comfort level in getting away with it.
They (typically) won't implement a cheat whose result can be quickly and easily discovered. They (typically) won't implement a cheat whose magnitude will invite further investigation - that is, a cheat that results in, say, a 10-fold increase in the house edge.
However, given human nature in general, the nature of the humans that appear to be involved in the operation of online gambling, the magnitude and persistence of the global economic downturn, the impact of the UIGEA, the ongoing (never ending?) actions of the US DOJ, etc. - they will cheat you.
Some Casinos cheat all of the time, but all Casinos cheat some of the time.
Onward to Part 2 - Do Online Casino non-Slot Games Cheat?
With slots, the only known performance details about any specific game are the ones disclosed by the software supplier and/or the Casino. As mentioned in Part 1, unless there is some sort of Monthly RTP Report provided by a respectable Certification company, any/all of this data can simply be lies. And even then there are problems because the Certification company lumps all of the slot data into a single number. One number for 150 slots leaves the Casino with a whole lot of wiggle room to cheat you and get away with it.
However, for the group of games covered in this Part 2, a great deal of performance details is known without any disclosure by the software provider or the Casino. A 52-card deck is a 52-card deck, and each card has the same 1-in-52 chance of appearing as any other card. It's just statistics, and there is no possibility of lying. This, however, does not preclude the possibility of cheating.
I'll wade into some details and statistics for one example, and then try to move more quickly from there on.
*** Some Gory Details ***
In Blackjack (BJ) it should be true that you have exactly the same chance to be dealt any starting card (first card a 7), any second card (second card a King), and any starting total (2-card total = 17) as the Dealer.
In BJ, the "problem" starting totals are from 12 to 16 (any Hit could result in a bust). The probability of getting dealt a 2-card total of 12 to 16 (for the typical BJ 6-deck shoe) is about 38%. 2,000 hands is a good sample size, so you'd expect to see an average of 760 deal hands in this range.
The Standard Deviation (SD) for a 12-to-16 deal in 2000 hands is about 22. +/- 3 SDs is the usual cut-off for the "Tin Foil Hat" crowd. ("Tin Foil Hat" crowd = "This game is totally rigged. I just played 10 hands and lost all 10. Everyone knows that's impossible. I want my money back.")
So, with an average of 760, +/- 3 SDs = 694 to 826. If you play 2000 hands of BJ and do nothing more than keep count of the 12-to-16 deal total, a count of 694 to 826 for you and/or the Dealer really can't justify crying "Wolf". (Well, you could try, but most of the villagers will ignore you.)
(For the sake of reference, 4,000 hands, average 12-to-16 count = 1520, SD = 31, +/- 3 SD range = 1427 to 1613.)
Bottom Line - The deal total is easy to keep track of, and easy to analyze.
Can the Dealer cheat on the deal? Of course they can.
One way would be to "work the up card" only on specific hand combinations. Example: The server deals your hand a 9 and a 7, total 16. The server deals the Dealer an Ace and a 6 (soft 17). If the Dealer shows the 6, Basic Strategy says Stand (and lose). So, although the Dealer's first card "dealt" was an Ace, they show the 6. The Dealer’s deal total is the same so doesn't show a problem, but you got screwed.
Precise calculations can be made as to when to "work the up card", such that the addition to the House Edge would fly below the radar - say, half a percent or so. Even something as small as a half a percent on a highly popular, high-bet-per-round game like BJ can mean a whole lot more money in the Casino's coffers at the end of the month.
(As the Wizard of Odds said in one post - "If you're going to cheat, you want to fleece the sheep, not slaughter them.")
"But wait." you say. "This would get caught in a first-card/second-card distribution analysis of the play logs". Simple solution - the Casino showed the 6 as the first card in the hand, but they recorded the Ace as the first card in the database. Analysis of the play logs is clean.
"But wait." you say. "This would get caught if I kept a record of the first 2 cards dealt, for both me and the Dealer. Then my record would show what they’re doing." True. However, I don't know of any Casino that doesn't say somewhere in their complaint processing documentation that in the event of a disagreement between your data and the Casino's data, the Casino's data will win.
(A very recent example, on target but sadly about Roulette rather than BJ. But you get the point.)
If you want to see what it took for a Player to deal with cheating software, have a look at this thread on The Wizard of Vegas forum:
In this case, the Player played thousands of games over an extended period of time, video recorded several hours of game sessions, uploaded these videos to YouTube, and then struggled through the skepticism of the forum members because he wasn't "speaking their language" before finally causing the whole issue to explode. (I plan to talk about this specific example in greater detail in a later post.)
BJ third-card analysis is a little trickier, because Basic Strategy is working in your favor. The Dealer can't follow Basic Strategy - they always have to Hit if their hand total is less than 17 (typically, this includes "soft 17"). Example: Your deal total of 13 only takes a third card if the Dealer is showing anything higher than 6. The Dealer's 13 always has to take a Hit.
So, the waters for a third-card analysis are going to be pretty muddy anyway. If the dealer is going to cheat, this is an even better place to do it. One quick example, pulled in from Part 1 - the "Win Percent" factor on your Hit request for the third card.
The software code for the game has as one of its input parameters a "Win Percent" factor. The game runs, and let’s say that this "Win Percent" factor is 90%. For reasons, you take a third card. If it is possible for a third card to bust your hand (a hand total greater than 11), the first thing the game code does is generate a random number. If that number is less than 90, then the game plays normally. If the number is greater than 90, then the game plays your hand to a bust.
Note that this 90% number can be adjusted to accommodate the comfort level of the Casino. A Casino that is more concerned with getting caught might be more comfortable with a 98% adjustment value.
*** End of Gory Details ***
To enable cheating, the Casino software application would have to provide the capability, and the Casino would have to use that capability.
Obviously, all of the Casino software companies have for years been swearing under oath that they do not provide this capability, or indeed any capability for the Casino to directly manipulate any aspect of the games. These statements have repeatedly proven to be lies.
Here is a link exposing the BJ game from Rival, one of the larger software suppliers in the industry.
And another link exposing the Video Poker games from SkillOnNet, a smaller software supplier to the industry.
(In my opinion, this SkillOnNet issue exposes the use of that "Win Percent" factor I've mentioned several times. The factor setting that led to the discovery of the problem might have been 80%, or 70%, or whatever. Once the problem was exposed, the factor was immediately set to 100%.)
And another link exposing the Craps game from BLR Technology, another smaller supplier to the industry. This one has a second link in which it is statistically demonstrated that the probability that this Craps game is NOT crooked is 1 chance in 4,500,000,000,000,000,000,000. (I don't even know how to say that number in words.)
And finally, I have the following link. This is an odd post in an odd thread. (I would tend to distrust its contents if it weren't so odd, and so long.) However, it appears to indicate very clearly that software provided by Real Time Gaming (RTG), another one of the larger software suppliers in the industry, allows a Casino operator to modify game returns on the fly. (If I had to make a guess, I'd conclude that the "Win Percent" factor method was at work here.)
It also states that 2 separate attempts were made to make this capability public on the "watchdog" web site Casinomeister, and that both attempts were promptly shot down by the Casinomeister administrators. (Say "Thank you Casinomeister".)
So, there you have it. I'm sure there are other web links (a handful? hundreds?) that are NOT published by the "Tin Foil Hat" crowd and which disclose the details of online Casinos cheating at non-slot games. This is what I was able to pull together in my limited time.
I'll end this the way I started it - with my "Statement of Bias".
If the Casino thinks that they can cheat you and get away with it, then they will cheat you. They will cheat you to the extent of their comfort level in getting away with it. Some Casinos cheat all of the time, but all Casinos cheat some of the time.
What can you do about it? How can you protect yourself? See Part 1.