Cheating at Professional Bridge

Interesting article on detecting cheaters in professional bridge using big-data analysis.

Basically, a big part of the game is the communication of information between the partners. But only certain communications channels are permitted. Cheating involves partners sending secret signals to each other.

The results of this can be detected by analyzing lots of games the partners play. If they consistently make plays that should turn out badly based on the information they should know, but end up turning out well given the actual distribution of the cards, then we know that some sort of secret signaling is involved.

Posted on March 8, 2016 at 6:07 AM • 27 Comments

Comments

jggimiMarch 8, 2016 7:50 AM

I play competitively, but am still a "beginner" as I have only been playing for a little over a decade. I play in local clubs, or in regional events. At my (low) level of play, bidding or play irregularities are common and inadvertent, and a director is as likely to be called by the perpetrating side as by their opponents. Perhaps I'm lucky, having never come across it.

In bridge, pairs do need to practice their communication, and understand the meaning of their bids during the auction and of their card play signals when defending.

An opponent can ask me about my partner's bid or carding. And my least favorite answer to give is, "No prior agreement." That's bridge-speak for, "I have no effing clue what my partner's last bid means. We never discussed what he/she just did, so your guess is as good as mine."

jggimiMarch 8, 2016 7:59 AM

I wrote, "...having never come across it." For clarity, that should be "...having never come across cheating."

M. WelinderMarch 8, 2016 8:07 AM

That's one doomed game. It seems like the cheaters only need a side channel
for a few crucial bits.

What's described in the article is amateur cheating. They all seem to do
unsalted passwords: the cheats mean the same every time. For starters,
they should salt their codes: so maybe a horizontally-held card means "club"
today, but next time it'll mean "spades" and the time after that it'll mean
nothing whatsoever.

Done with just a minimum of skill, there is no way of detecting the side
channel's existence other that "you win too often".

Clive RobinsonMarch 8, 2016 8:53 AM

First things first, bridge is in effect a very dull set of rules laid over a more enjoyable card game...

Secondly the people who tend to play it "socialy" treat it more like warfare. And as the old saying has it "All is fair in love and war". So are they "cheating" or strategically "taking advantage of an opponent's weakness"?

@ Bruce,

This "catching cheating" by the use of Computers is becoming quite popular these days. As I pointed out on the squid page Crossword Making and Chess are likewise tainted.

It would appear that some people with a mentality for such games, have a habit of misusing their talents to make money or fame for themselves. But they totaly fail to realise that others with a similar mentality, now have an avenue via computers etc, to get their own slice of fame and fortune at the cheaters expense...

You would have thought if they were as smart as they think they are then being less obvious or covering their tracks would be part of their game plan...

But apparently not, which begs the question of what the real smart ones are getting upto ;-)

Ray DillingerMarch 8, 2016 10:44 AM


They're going to detect "cheating" that people aren't even aware that they're doing.

We are humans. Our brains integrate available information and make decisions. We don't get a real choice about what information we integrate into a choice; our best decisions are often made taking into account things we're not even consciously aware of knowing.

If people have been playing bridge together long enough to become an effective team, they have "tells" and are aware of each other's "tells" whether or not they know they are. Particular types of hands will make a partner nervous and she twists a lock of hair when she's nervous. The partner who's been playing across the table from her hasn't thought about it, but has memories of success with some types of play when partner twists that lock of hair and failure with others. Conditioning works. Big-data analysis will identify them as cheaters even though they had no such intent.

If you really want to lock out all other channels of communication you have to restrict play to a medium where there are no other channels of oh wait that's boring and no fun to play.

BrunoMarch 8, 2016 11:46 AM

Just read the article.
Actually, there isn't much on Big-Data Analysis - it's mostly about crowdsouring video-analysis trough youtube. Statistical analysis is only mentioned in one example.

Cyborg2237March 8, 2016 12:49 PM

Great example of how covert communications can be surveilled with "mathematical certainty" by means the communicators never suspect.

And how crushing that can be for their purposes, even if the exact nature of the communication is unknown.

This often happens in murder cases. Someone says they haven't talked to someone. Their phone records are pulled, and messaging metadata traffic is seen. What is said is not known, but that they were communicating, how much, and at the juncture of time -- with the denials -- all of that puts them in the spotlight.

keinerMarch 8, 2016 1:03 PM

Sounds like typical pre-crime nonsense. How many false positives? How many false negatives? You don't have the data? Forget this nonsense!

Statistics is a bitch! March the driest month of the year, I get a new roof on my house, 5 minutes after opening the old one it starts to rain. Needless to say there was NO rain at all in weather forecast....

Please, forget this fancy "we have a pile of shyte data and find I bucket of gold for you" Ponzi nonsense.

albertMarch 8, 2016 2:26 PM

@Bruce,

Are you big on big data analysis?

"...then we know that some sort of secret signaling is involved..."

We 'know' nothing of the sort. What we 'know' is that computer 'analysis' returns some range of -probabilities- that -some- cheating -may- have occurred.

All of these finer points will, of course, be left out or ignored by the gov't and the MSM; much more dangerous adversaries than bridge players. :)

. .. . .. ---


WaelMarch 8, 2016 4:09 PM

Big data wasn't big back in the day. Lucky Omar Esharif got away with his ancient secret bridge signals!

Speaking of analytics, cheating, and cards (Black Jack specifically,) casinos employ software to monitor players and how they vary their bet spreads (which the table can determine because they have RFID sensors that read the RFID tags embedded in the chips) in correlation with the count. If a player is observed to increase bets as the count goes high, and decreases bets when the count is negative, they are marked as an Advantage Player, then they get barred (86'd.) -- this is an analytic / statistical approach. Apparently this information is shared with other casinos so they're on the guard. Not sure this qualifies as "big data" because I have no insider knowledge of how the database is shared. In a way, casinos are allowed to cheat! In addition, the so called "random" shuffling machines don't really produce random cards. The player is playing against a controlled deck (I hinted to @ianf about it here and he refused to bite.) The odds are stacked so high against players, they stand little chance. Although only people with trivial minds approve of gambling, it's still just sad to see casinos ruining people's lives...

I wonder what kind of "testing" these machines have to pass to be certified as "fair". A simple statistical analysis of the RNG output isn't sufficient, because the use of the output of said RNG can be "deceiving"...

albertMarch 8, 2016 5:30 PM

@Wael,

The first link you posted (patent filing) contains a good description of Black Jack, for non-gamblers (me) and gamblers alike.

Cheating at games of chance is an art form that's been around since gambling was invented, but the addition of technology has changed the game. Like computerized voting machines, computerized slot machines, poker machines, even roulette machines effectively hide any sort of cheating by the casinos. Admittedly, they reduce cheating by players, but if both sides cheat, then the game goes to the best cheater, and that just seems wrong.

Casinos make billions, so I shouldn't be surprised if they 'stack the deck'. Greed is only natural. The sad thing is that folks keep gambling. There -are- worthy causes to spend money on, instead of giving it to blood-sucking parasites.

. .. . .. --- ....

KenMarch 8, 2016 6:04 PM

@ Wael

"I wonder what kind of "testing" these machines have to pass to be certified as "fair". A simple statistical analysis of the RNG output isn't sufficient, because the use of the output of said RNG can be "deceiving"..."

It was many moons ago that I enjoyed frequenting the many casinos in the world. As a courtesy, the house by policy or observed coincidence usually let the newcomer win a few hands before it take his chips away. Thus, I roamed around and ended up with positive hands. As I roamed, it wasn't before long that I accidentally overheard the inside jocket whisper to the table dealer he's a swindler don't let him win.

and then I realized the fix was in...

WaelMarch 8, 2016 6:07 PM

@albert,

Cheating at games of chance is an art form that's been around since gambling was invented

Players were ahead of casinos in technology, and maybe some still are! This guy is the most interesting one I've read about. The book is fascinating. On a relevant note, the funniest cheat I heard about back in the early eighties took place in Atlantic City. I cannot find a reference to it in search engines, but it went something like this:

A drunk guy stood in front of a roulette machine. He wasn't placing any bets. Whenever the the wheel (or the ball) stopped, he'd throw a $5 chip to the dealer and say I lost the bet. The dealer told him: but you didn't place any bets! The player said (in drunken voice) I am an honest player! I made a mental bet and lost it. The dealer looked at the pit boss who signaled him to take the money. This kept going on for a few hours! The guy kept loosing five dollars every turn of the wheel. Then suddenly the player started jumping up and down screaming: I won, I won! The pit boss told him you didn't bet anything! The player said: I made a mental bet and won. The pit boss told him: here, take your $5 bet and leave! The player said: oh, no! I mentally placed the maximum bet on the table! You owe me $35000! The pit boss kicked him out of the casino without paying him a penny...

So the guy went to court and sued the casino. In court, the judge told the casino: I understand he didn't place any bets when he won! But if you take his money when he lost, you must pay him money when he wins. He's an honest guy, after all ;) Reminds me of a joke. Perhaps I'll leave it for another day as I exceeded my quota for today.

Not sure if it's true, but I heard it from several sources.

WaelMarch 8, 2016 6:39 PM

@Ken,

and then I realized the fix was in...

Sounds like it wasn't an expensive lesson, and hopefully you stuck it to the man!

Jonathan WilsonMarch 9, 2016 12:51 AM

If the card shuffle machine has any way to know what cards its dealing or what cards are in the deck, its not possible to be truly fair.

A truly fair machine would shuffle the cards randomly and then deal the cards in the order that results without ever having a way to know the cards comming out.


rMarch 9, 2016 1:29 AM

@clive,

If someone has a knack for cards, sure they shouldn't try to live off from it... But honestly? I'm not going to get in their way: and neither should any of us.

I bought my woman a membership to a lodge so she could play eucher with real people twice a week and you know what? She enjoys it, as do I... and 9 out of 10 times she's in the top three when she plays.

It's not a big pot, it's maybe a pack of smokes and some gas... But it's buckets of fun and an outlet for both of us.

In the game of life and cards: the side+channel is your face.

I have pretty good luck on hunches in bridge canasta and euchre too, I'm quite sure I would be a potential suspect of this program via known deviations of luck and outcome too.

I'll check out the article, but i can't see this being anywhere near the accuracy of day the scandal brought to tennis throught data mining.

:) I wonder if the other r will post here too.

KevinHMarch 9, 2016 3:29 PM

@Raydillinger, re:

"If people have been playing bridge together long enough to become an effective team, they have "tells" and are aware of each other's "tells" whether or not they know they are. Particular types of hands will make a partner nervous and she twists a lock of hair when she's nervous. The partner who's been playing across the table from her hasn't thought about it, but has memories of success with some types of play when partner twists that lock of hair and failure with others. Conditioning works. Big-data analysis will identify them as cheaters even though they had no such intent."

@r, re:

"In the game of life and cards: the side+channel is your face."

This is why, in bridge tournaments, a vertical divider is placed diagonally across the table so that partners cannot see each other during the bidding. The bids themselves are placed by putting pre-printed cards into a tray. Players are expected to maintain the 'tempo' of the bidding, so that their partners get no clues from very slow (or rapid) bids.

For the games discussed in the article, where cheating was detected by the analysis of play, the players would not have been able to use their face, voice, or other tics and tells to leak information to each other--inadvertently or deliberately. The "friendly" game with the neighbours, of course, is a whole different story.

albertMarch 9, 2016 4:27 PM

@Wael,
I like a good story, even if it's not true. :)
. .. . .. --- ....

KeithBMarch 10, 2016 9:07 AM

I refuse to believe that counting cards in blackjack is "cheating". Casinos are short sighted to kick those players out - most people do it badly anyway.

albertMarch 10, 2016 10:22 AM

@KeithB,
Agreed. It's integral to the game. Besides, can't counting be mitigated by shuffling the decks? If they don't like the game, they need to change the game. Imagine playing poker, where you throw away all the cards in the last hand.
. .. . .. --- ....

KenMarch 10, 2016 6:54 PM

@ Wael

"Sounds like it wasn't an expensive lesson, and hopefully you stuck it to the man!"

An expensive lesson would be a lesson unlearned, but it was indeed inexpensive as I had put on the stop loss, which was no play. It's a racket back in the days and I believe it still is. Those inside it either enjoy the benefits in tacit or be buried/dumped in the locale, and thats the subject of the other thread.

jdgaltMarch 10, 2016 11:26 PM

I play tournament bridge and have been hearing about this for a couple of years now. The main reason it's become news is that the organizers of top-level tournaments have started posting videos of at least the final rounds on the Web, where if anyone is using visible "codes" to communicate, viewers can see it. I'm rather surprised at how many cheaters there are. Maybe I shouldn't have been.

But I want to correct a wrong impression which I think some people may draw from the article: "Professional" bridge in the sense of either high-stakes betting or prize money doesn't exist, and never has. Bridge "professionals" are like golf pros: they either write books, teach, or play as the partner of a paying "sponsor" who doesn't play as well as the pro, thus enabling the sponsor to win tournaments and masterpoints.

The reason I'm surprised at the number of cheaters is precisely that: there aren't any big bets or prizes for cheaters to collect. Maybe some of them sell more books because of their perceived skill, but I think that effect is minor. Some of the biggest selling authors (Eddie Kantar, for one) haven't won major tournaments in decades, but sell well because they write well.

JulianMarch 11, 2016 2:41 PM

@ jdgalt

The big money isn’t in writing or teaching, but in being paid to play. The top pros make a pretty good living at it. The more high-level events they do well in, the more they can charge the client. (And it’s not just as the client’s partner; most of the prestigious events are team games, where you usually have one client and five pros.)

OscarPhilipsMarch 15, 2016 9:26 AM

I am not a bridge player, but was a heavy euchre player in college, spending 4 to 6 hours a day playing with the same partner the bulk of that time, and while we did not cheat, we played with each other so often that we had expectations of what the other player was doing, just based on the calling of trump process, I am sure there were small clues in body language, timing of the bid, and tone of voice that we did not understand ourselves.

One particular tournament comes to mind when our opponents were so bent on catching us cheating, we successfully passed the deal back and forth between us for a good 80% of the hands.

Leave a comment

Allowed HTML: <a href="URL"> • <em> <cite> <i> • <strong> <b> • <sub> <sup> • <ul> <ol> <li> • <blockquote> <pre>

Photo of Bruce Schneier by Per Ervland.

Schneier on Security is a personal website. Opinions expressed are not necessarily those of IBM Resilient.