When games of chance become games of skill, and the house gets an attitude

I know that they’re not moral because they’re kicking out people whose only sin is to play well. Basically, you seem to be claiming that disqualifying one from playing Rock-Paper-Scissors because she is good at randomizing is excusable simply because the guy organizing the game is a crook. This is not an obvious conclusion, and you are going to have to construct an argument for it if you wish me to understand it.

Re: Blackjack:

Epstein, pg. 225

A player walking in with $100 and playing the optimal zero-memory strategy can expect to walk out with $100.10. Darrell Huff, in his 1996 The Complete How to Figure It gives a diluted version of Thorp’s optimal strategy that is “almost as good” as the more complicated version.

As a benchmark, consider the following:

Epsetin, pg. 248-9

So, even with a computer at hand to calculate all possible permutations of the remaining deck and spit out the optimal strategy at each instant, and a “conservative” betting strategy, a player who walks in with $100 can expect to leave with $113. And even with that advantage, knowing what the dealer’s down card is will let a player expect to walk out with $140. Now, that would be one hell of a profit for sure; but, not one any punter is going to get.

Page 244 of Epstein gives a table comparison of counting strategies (20 cards remaining, bet variation of up to four units). The expectations offered are all in the range of 0.019 to 0.025; i.e. the best strategy for a human player will allow a punter to enter the casino with $100 and expect to leave with $102.50. The “Griffen-1” strategy can be tweaked to increase the punter’s expectation on leaving the casino to be $103.70.

I don’t know if, or to what extent, the casino rules have changed since Epstein has published his book, nor do I know what new strategies have come to light since then. I nevertheless remain skeptical that individual counters are a real profit threat. The argument that others can mirror the successful counter and thereby gain unearned profits at the casino’s expense seems fair. Worth noting is that the win-lose correlation between players doing so is about 0.5 and the results are “modest.” (Epstein, pg. 249)

As far as I am able to discern, the math does not encourage the idea that counters are going to be very profitable from their efforts. One must realize that the personal experience of a person who made a living gambling is not compelling evidence until that person can distinguish herself from the Royal Head-Flipper. The survivorship bias in such accounts makes them inherently questionable as usable data.

That is not to say that one who learns to count and has a long career in the black isn’t honest, trustworthy, or intelligent. It’s just that when the starting pool is large, some people will do exceptionally well by chance alone and their accounts do not change the results we can expect for a randomly chosen counter.

js_africanus:

First, the game Epstein is describing is a single-deck game with all the good rules. As I said before, you can sometimes find this game at very low limits - it’s a loss leader for the casino. You will NOT find that game in the high limit pit where serious money can be made. In fact, you’ll almost never find it anymore at all.

In Reno, you can find a reasonable number of single-deck games that have a small house advantage of around .25. More common are the four and six deck shoes. A 4-deck game that allows double after split generally has a house edge of around .5. The six-deck shoes have a house average generally a little higher, at maybe .6-.7.

BTW, the reason the advantage moves towards the house as you add more decks is because there are fewer blackjacks dealt out.

As for Epstein’s playing strategy-only technique, that is mostly true in single deck games. Playing strategy variations are critical to beating single deck games, which is why players who play them learn complex multi-level counts, side counts of sevens and aces, and memorize many strategy variations. In shoe games, most of your profit goes from bet variation, and strategy changes alone are not enough to overcome the house edge. That’s why I keep pointing out that shoes are actually easier to count than single deck games.

As for taking the word of someone who claims to have beaten the house, you don’t have to just take my word for it. If you need more empirical evidence, you can start by explaining why the casinos spend millions of dollars on anti-counter countermeasures. Or why the casinos in Atlantic City were willing to fight tooth-and-nail in court to protect their right to bar counters (they lost (to Ken Uston, btw), and you can’t be barred in Atlantic City - which is why the games there suck). And I’m not the only one who’s made a living playing blackjack. I know literally dozens of people who have done it or are doing it. The ranks of professional blackjack players is thinning, however, for two reasons - one is the rise of poker, which has attracted away a lot of blackjack players. And the other is that the casinos have gotten very good at spotting counters, and making a living at it for big stakes is increasingly hard to do.

As for a player “Walking in with $100 and walking out with $113”, what you are neglecting to consider is that the players play MANY trials. If you have a 1% edge per hand over the house, and you play 100 hands, you expect to make 100% of your average wager. If you are betting $100 per hand, and you get 50 hands per hour, that’s $50 per hour profit. I don’t know about you, but most people would think that was worth playing for.

The downside is risk. Blackjack has a lot of variance. Much more so than poker. So while your expectation may be to make $50 per hour, the variance around that may be $1000 per hour. The variance is so large that in a given hour it is impossible to see the trend through the noise. That’s why bad players keep playing, after all.

Given a variance that large, after 400 hours of play, a player would expect to be up $20,000. But his standard deviation will also be $20,000. So being even one standard deviation out on the low side, which is quite common, can have a good card counter losing money after 400 hours of play, which represents probably close to half a year of play for a counter because they get so few hours at the table before they have to leave if playing high limits.

This is the real reason why there aren’t that many card counters. Because to make serious money, you have to have a HUGE bankroll. Especially if you have to pay your living expenses out of it. As the numbers above show, if you want to make $50/hr as a card counter, and you want a 5% risk of ruin, you had better start with a bankroll of maybe $40,000, plus enough to pay the bills each month if you aren’t winning. The thing is, most people who have those kinds of financial resources already have good jobs and wouldn’t dream of giving up their careers so they can sit in a casino all day.

So the average card counter probably plays with a bankroll that’s too small. This is the real killer of professional gamblers everywhere. They start with a too-small bankroll, they pay their bills out of it, and what’s worse, when they go on winning streaks they treat it like a windfall and spend even more. If you’ve read Epstein, you understand the Kelly Criterion, which states that if you continually bet more than twice the percentage of your bankroll compared to your advantage, at some point your bankroll will crash to zero no matter how big your edge is over the house. This, combined with those ‘counters’ who think they are winning when they have giant holes in their play, and casino countermeasures, is what really limits the number of serious card counters around.

This is also why the real threat to the casino are the teams. And there are some BIG teams arount. The MIT team was mentioned. There was also a team from Jet Propulsion Labs (they offered me a chance to play with them one year, but I couldn’t make it). There are other, even bigger teams. There was a team so large at one point that it had its own full-time accountant and hired people to fly around the world scouting for good games. When they’d find one that was especially profitable, the team would fly in, bet insane amounts of money, and win big. The key to a team’s success is that they can pool their bankrolls and each player can play as if the bankroll was theirs alone in terms of risk (as long as they all play on separate tables). This, coupled with the fact that teams have their own methods of camouflage such as the ‘big player’ technique which is very hard to spot, and the casinos sweat them big time.

Also, counters in the internet age tend to pool together a bit. I used to subscribe to an expensive notification service run by a mathematician and expert blackjack player named Stanford Wong (a pseudonym, of course). He had stringers scouting casinos looking for good rules, promotions, or mathematical errors, and notify the list when one cropped up. For example, one year a casino in Vegas introduced a new game called “Free Ride Blackjack”. The innovation was that if you got a blackjack, you got a little chit that allowed you to surrender a hand in the future and get your bet back. Well, it turns out that this game had a 2% advantage for the player. Stanford spotted it, worked out the optimal play, and sent a notification out, and when that casino opened the next day there were about 50 professional card counters waiting outside. That game didn’t even last 24 hours - towards the end, the casino was having to employ their security guards as chip runners to buy back chips from the players because the cage had run out.

By the way, “Theory of Gambling and Statistical Logic” was the first serious gambling book I ever bought. Great book. If you want something a little lighter and fun to read, I recommend “Theory of Blackjack” by Peter Griffin. He does fun stuff like work out the mathematical expectation of those little coupon books the casinos hand out, and optimal play for using them to make a profit of a few cents. Fun stuff, but also very educational in showing how to look for advantages.

js_africanus,

I don’t have the reference that you mention handy, but if I read the quoted section properly, it is claiming with a point count system, positive expectation can achieve .05! That’s is significant. The model I was using reported returns of around .025. Consider that is for every bet. Every bet you make, you expect an average of 2.5% return. How many bets per hour? A quick dealer can yield maybe, 50 hands/hour. With an average bet of $75, that’s about $93/hour. Ummm, that’s good.

Well, Sam states he could teach someone how to count in a weekend. I think that’s fair, as I am very familiar with the system he employed. I think it may take a few weekends to develop some polish, but that’s negligible. Considering that that system is pretty well known, and you could see droves of people with the bankroll taking $93/hour from the casino. When I played, I usually played over a weekend when crowds were thick. In a weekend, I would average 20 hours of gambling. That’s just shy of 2k with your estimates, average. A single person is not a profit threat, but several hundred or several thousand would be. It’s not data, and I’m not in the casino business, but I believe the most compelling evidence is the casino’s consistent behavior in expelling those who successfully count. I experienced it as a player, and heard about it when I was a dealer (not high enough on the food chain to be privy to real discussions).

.025 is consistent with what I used. Changing bet strategies would have a large effect, but at the strategy that I employed, average hourly expected return was around $45/hour. That with a standard deviation of about $250/hour. That’s a big swing, big enough that I believe many who would be interested casually get dissuaded.

.05? I may still be playing for .05! Like Sam, I got tired of being surrounded by the scum of the earth, and could make more money by toiling and paying taxes.

Another recommended book, one which I think is a great primer, is coincidentally, by Standford Wong. Professional Blackjack.

How fairly does the house have to play?

My first reaction is “They don’t need to play any way except that which will make them rich. Fairness is a liability. They will squeeze the marks for everything they have.”

But then, these are games with rules. Hoyle and them publish unbiased books on how to play blackjack, poker (any style), euchre, baccarat, etc. If the house is marking cards or otherwise giving itself inside information, is that not false advertising and, in essence, fraud?

I could perhaps think up a few weasely methods to get around the most naive readings of those terms. (Like a casino playing Blakjak or ‘Hold-em: Texas-style’, the same way cereal makers sell non-fruited Froot Loops.) But there’s still a vague reek of deception around the process. (Well, a different kind of deception than usually exists in Vegas.)

I know (thanks to the SDMB) that gambling machines have to give you a real chance of winning. (That is, the slot machine can’t have all lemons on one reel.) Do human-moderated games follow the same premise?

I have constructed an argument, you are just ignoring it. Casinos are businesses whose profit is made by offering games of chance that you can win, but the odds are you will lose. Casinos do not hide this fact. Nor do they hide the fact that if you employ a strategy that will move the odds in your favour they will stop you playing. Nor do they say anything different in their rules.

You continue to labour under a misapprehension that casinos are or say that they are neutral venues offering fair games. They are not, nor do they say they are.

You use words like “sin”. It is not a question of sin, it is a question of it not being profitable for the business in question to offer a service out of which it does not make a profit.

Comparing what a casino is and does to an amateur game of rock paper scissors just shows your fundamental lack of understanding of what a casino is. Your position is like saying a clothes shop is immoral because it won’t give you clothes at cost when your mother would.

If I were a professional rock-paper-sissors player, and someone appeared to have a system that was causing me to lose and pay out money, I’d stop playing with them.

But we can also turn this argument around. Why do some people choose to play only blackjack, and not roulette? Isn’t it unfair to the casino that more people don’t take them up on their efforts to provide a routlette wheel for anyone who wants to play that game?

If it’s fair for people not to choose to play roulette for fear of losing money, its fair of casinos not to play against people who can take the business’s money in blackjack.

But you might be laboring under a misapprehension of what constitutes a “fair game.” I don’t think js_africanus is actually suggesting that casinos are not allowed to offer games which are biased against the player. But a game that’s biased against the player can still be called a “fair game.”

Here’s an example from Scarne. Many funfairs or carnivals have midway games where a player must throw a hoop over a wooden block to win a prize. The game is difficult to win, and the prizes are not worth anywhere near what they look like, but the game can be considered “fair”–what you see is what you get. Biased against the player, but still “fair.” However, many carnies “gaff” the blocks so that a hoop will not fit over them under any circumstance. This is an “unfair game”–rigged so that the player thinks he’s playing a game of skill, but instead is being ripped off.

How can blackjack be an “unfair game”? In the old days, dealers didn’t deal from a shoe, but rather from the hand. That led to all sorts of shenanigans from crooked casinos in which dishonest dealers would introduce “cold” (stacked) decks, or deal the second or last card from the deck, to guarantee that the player would lose at inconvenient moments. That was an “unfair game,” because the player was manifestly being cheated. When blackjack players started getting wise to such actions, the dealing shoe was introduced, which ended “card mechanic” dealers from enriching the casino in that way. Later, state regulatory bodies further protected players. Nevada, for example, has very strict regulations on such minutinae as how video poker machines may be coded, or how roulette wheels must be constructed.

The question now becomes not “Is it fair for casinos to make a profit?”, as obviously it is, and nobody doubts that. Rather, it’s “What methods of profit-making are ‘fair’ or ‘unfair’?” Not all methods of profit-making by a casino can be considered “fair”–e.g., undercounting chips at the change booth, or having a pit boss take your chips off the table when your head is turned. Personally I think that ejecting card-counters is a “fair” procedure–after all, they are not obligated to serve anybody they don’t want to serve. But I could figure that some constituencies might disagree. That doesn’t mean they don’t think the casino has a right to a profit, merely that they disagree on the allowable methods of doing so.

Sorry, the quote in my post above was from Princhester.

Someone else may have already explained this but I didn’t see it - but I think this is how the MIT team worked. One player would signal the others when a table was “good”. He would always play, and wouldn’t stick to any pattern. When the table was good, the others would come over and bet heavily, and then leave. It was a lot harder to spot.

Sorry for this unrelated post - but I wanted to get Sam Stone’s attention and this seemed like a fresh thread he was on.

Anyway, Sam, can you do me a favor and e-mail me at senorbeef@wSorlPdnAetMoh.com - remove “SPAM” from that. I’d like to discuss some things privately with a poker pro, if you wouldn’t mind. Thanks.

This is GQ and I don’t think it appropriate to argue about what is or is not fair in any subjective sense. I use the word here to mean simply “offering each participant an equal chance of winning”.

Are you sure? I think that is pretty much the upshot of what js_africanus is suggesting in relation to card counting, albeit that I’m not sure js_africanus has thought the issue through.

All the “unfair” examples you give are examples of casinos doing things that are illegal and/or not allowed even by their own published rules. Kicking out card counters does not fail either of those tests. Indeed I challenge you to lay down any objective test by which it is unfair, except in the sense that it stacks the odds against the player, and no casino could survive but by being “unfair” in that sense.

But that’s the crux of the argument. Every person who comes to a casino blackjack table theoretically has the same chance of winning. That the house holds the odds in their favor is irrelevant. The question is what method the house may use to maintain their mathematical superiority.

On re-reading js’s posts, I’m not sure he thought through the issue completely either, but, as I’ll mention below, I don’t think there’s an iron-clad answer to what a “fair game” entails.

Nitpick, but an important one: At least in the States, the casinos didn’t write their own regulations. They only did so when the individual state gaming boards stepped in. I would venture to say that, without state regulation, casinos would be just as lawless as they were in the '40’s and '50’s–merely because they could be, “their own published rules” notwithstanding. It’s the state, and the state alone, which is the arbiter of legality.

Well, for starters, I never said that kicking out card-counters was unfair–in fact, quite the opposite:

But, secondly, I don’t think that you can make an objective argument over whether this can be fair or not. As has already been mentioned in this thread, casinos cannot discriminate against race or creed. Apart from that, a casino is well within its rights to serve or not serve anybody they want, much like a bar that might prevent a known troublemaker from entering. Who the casino ejects, and their reasons for doing so, will perforce be subjective, as will any arguments as to what casinos should and should not do.

Almost every counter I know is actually in favor of allowing casinos to kick them out if they catch them. Because if you take that power away from the casino, the only tool they have left to thwart counters is to change the game so that it can’t be beaten for much money. That hurts everyone - counters, the casino, other players.

The current ‘balance of power’ seems to be a reasonable compromise that everyone can live with.

I actually asked this question here before. I wondered how Atlantic City casinos, which are “Public Accommodations” in the “State of New Jersey,” can deny service to people who play in a way that makes them more likely to win. I still don’t see a difference between counting and remembering if the Queen, KIng and Ace of Spades have been thrown in a Hearts game.

saoirse, look at it the other way around. What exactly would prevent the casino from denying anyone they choose, other than the Civil Rights Act? And that doesn’t say anything about protecting card counters.

If I choose not to play Hearts with someone who can keep track of the Ace, King, and Queen of spades, isn’t that my right?

saoirse: You can’t be barred in Atlantic City. A counter named Ken Uston was barred in the 1970’s, and took the casinos to court. He won, and since then casinos in New Jersey have been forbidden to bar card counters.

As a result, the games there suck. 8 deck shoes, bad penetration.

Casinos in Atlantic City can ban anyone they suspect of counting cards.

Gambling is regulated out the wazoo, so it comes as no surprise to find that kicking out card counters appears to fall under this regulation. IANAL, but cursory research appears to back up what Sam Stone said, and suggests that regulations governing the ejection of card counters differ quite dramatically from jurisdiction to jurisdiction.

Blackjack Insider addresses the question in their January 2004 newsletter: “Is Card Counting Legal?”

Note that the line of reasoning given for Nevada is pretty much what many have been saying in this thread. Also note that the New Jersey courts have apparently not agreed with this line of thinking.

I have not a clue whether this guy’s analysis of the case is accurate, but this link purports to give a brief summary of (what I guess is) the case Sam Stone was referring to, Uston v. Resorts International Hotel, Inc.:

This line of reasoning is similar to what many anti-counter-ejection folks have been arguing in this thread.

Disclaimer: This was the result of a very cursory search of the internet, and I have no particular expertise. I could easily have missed or misinterpreted something, and if I have I’d appreciate correction/clarification.