As a professional-grade poker player and long-time Economist reader, the following made me laugh:
Economist April 20th print edition
A POKER face. It is the expressionless gaze that gives nothing away. To win at poker, the face must be mastered, and master it is what the best players try their best to do. But a study just published in Psychological Science by Michael Slepian of Stanford University and his colleagues suggests that even people with the best poker faces give the game away. They do so, however, not with their heads but with their hands.
Mr Slepian made his discovery when he showed 78 undergraduate volunteers video clips of players placing bets at the 2009 World Series of Poker. (Bets in poker are placed by pushing chips into the middle of the table.) The clips were 1.6 seconds long, on average, and featured different parts of the players’ anatomies. Some showed everything visible from the table up: chest, arms and head. Some showed just the face. And some showed only the arms and hands. Each volunteer watched only one of the three types of video, but was shown several examples.
After each viewing, volunteers were asked to rate the quality of the player’s hand on a seven-point scale. Then, when they had finished watching all the clips, they were asked to rate their own experience with poker on a similar scale.
Mr Slepian found that students were poor at judging the quality of a player’s hand when shown just that player’s face. Indeed, he noticed a negative correlation of 0.07. This is not huge (a perfect correlation is 1.0). But it meant there was a statistically significant tendency that the better a volunteer believed the hand to be, the worse it actually was. When a player’s whole posture was considered, this misapprehension went away: if a volunteer could see everything about a player from the table up there was no correlation between his judgments of a hand’s value and its actual value. When a volunteer could see only arms and hands, however, Mr Slepian found a positive correlation, of 0.07, between his guesses and reality.
To confirm his discovery, Mr Slepian re-ran the experiment with a different set of clips. The results were the same. Students, even those who were poker novices, could judge the quality of a professional poker player’s cards from the behaviour of his hands. The next question was, how?
Mr Slepian knew from previous studies by other people that anxiety has a tendency to disrupt smooth body movements, and he suspected this might be the explanation. To find out, he showed 40 new volunteers the clips he had used in the previous experiment. Rather than asking them to judge the quality of a player’s cards, however, he asked them to rate either that player’s confidence or how smoothly the player pushed his chips into the middle of the table.
He found that when students rated players as being confident or having hands that moved smoothly, the cards they held were likely to be good. There was a positive correlation of 0.15 when the students considered confidence and of 0.29 when they looked for smooth movement, so they were actually more capable of determining hand quality from these variables than when asked to estimate it directly. The moral of the story for players, then, is don’t look your opponent squarely in the eye if you want to know how good his cards are. The secret of his hand is in his hands.
God bless academics… entertainment value aside, that study is completely worthless (albeit in an interesting way).
One can question whether a 0.29 correlation has any merit in the first place. But putting the dubious math and questionable control variables aside, findings like this one completely overlook the subtleties and complexities of the game.
For example, re, competitive poker edge, we can demolish the “watch their hands” argument with a simple application of experience and common sense:
If your opponent is weak enough to let hand movements broadcast information, you do not need subtlety to beat him. Bad players are legion at the poker tables. In any given tournament or cash game — even at higher stakes buy-ins — the amount of staggeringly bad play is eye-opening. Given this empirically observable reality, how do you handle terrible opponents? The answer: Very simply and straightforwardly. Subtleties are wasted on them. Not only that, applying subtlety in attempting to “read” a bad opponent can actually be counterproductive, because if he is bad enough, your opponent will misread his own hand values (and generate a fool’s confidence in doing so). Based on board texture and betting sequence, top pair / top kicker could be all but worthless, yet your awful opponent may treat it as the stone cold nuts in his mind — telegraphing the strength of a set, a straight or a flush (via confident hand movements) because he is too green (or dumb) to situationally distinguish between a strong hand and a weak one.
If your opponent has any kind of experience or skill, on the other hand, buttery smooth bluff movements will be second nature (thus invalidating the entire premise). Subtlety does have value when facing skilled opponents. But skilled opponents present a different kind of problem: They are well aware of telegraphed signals and have lots of practice concealing them. Worse still, a truly skilled player can run a large bluff with absolute calmness and confidence (not a tremor to be found) because, if he understands the game with sufficient depth, he will literally not care about the outcome of that single confrontation. Bluff spots are assessed probabilistically: If you know that X bluff in X situation has a certain favorable expectation, regardless of this result, you can make that bluff in the same way you would accept a 60/40 coin flip with the 60 in your favor… and do so with nonchalance bordering on boredom.
On a personal note, I remember the first time (circa eight years ago) I made a big bluff at a cash game pot. It was actually a little nothing amount — something like 75 dollars — but it felt absolutely massive at the time. I remember an adrenaline rush so strong it felt like my heart might pound right out of my chest.
Now, these many years later, I have made — and called down — countless bluffs in the +$1,500 range, or even larger, without so much as a half-degree rise in temperature. Firing, say, $600 into a cash game pot on a casual bluff or semi-bluff with a net positive expectation produces no more internal drama than, say, tossing in a $5 chip.
That’s a long way from sweating bullets over 75 bucks…
And it is the same for other experienced opponents, thus making the smoothness of hand movements completely worthless as a tell. If the opponent is a tourist, subtleties are a waste or even counterproductive. If the opponent has been around the block enough times, he will possess a smooth bluff confidence that originates from a place wholly other than the value of his cards.
(The situation gets even more complex, by the way, when an opponent is smart enough to understand that a gigantic value bet is more likely to be called if it is made to appear as a bluff. On the river, your opponent fires $1,200 into a $700 pot — a bet way too large to make logical sense, unless he is trying to scare you off with sheer force. Meanwhile, he looks away when you make eye contact with him. When you say “time” and take another thirty seconds, he spreads out the twelve benjamins he has laid out in an intimidating peacock fan. Very briefly, he tugs at his collar in just the way Joe Navarro says indicates discomfort. These are nervous tells that practically scream “I am bluffing!” Which is exactly the point. He knows the tells just as you do, and is attempting to induce a hero call.)
Ah yes, complexity… the stuff that only seasoning and experience can overcome.
Are there ways to win at poker, and win consistently, by walking a higher skill path than one’s opponents? Oh indeed. We could give a sixteen-hour lecture series, sans notes, on the finer details of poker theory (and even most “good” players are not nearly as good as they think they are). But silly stuff like “watch their hands” would not be a part of any theory curriculum. (Except maybe a section titled “dumb stuff not to rely on.”)
Articles like the one in The Economist cited above, and the research studies that generate supposed “insight” into poker and trading, highlight a key point:
In respect to highly competitive, skill-based human endeavors with zero sum or minus sum outcomes (like poker and trading), tentative hypotheses generated by non-practitioner research are generally worthless.
This is because far too many nuances, subtleties, and empirically critical factors (hidden edges discovered and verified through ample experience — things one wouldn’t even know to consider without experience) are completely overlooked by those outside the game.
This train of thought also explains why academics who condemn trading as impossible, or say trading outperformance is impossible based on various “studies” or “models” or Wall Street test cases blah blah blah, are almost invariably nitwits with an agenda.
(My favorite class of folly in academic research, as it relates to trading, involves the studies that purport to investigate whether chart patterns “work” by, say, looking at 7,000 separate instances of data-mined head and shoulders setups in U.S. equities of a certain market cap over the years 1981 to 1995, assuming equal fixed fractions of risk capital — no situational adjustments, position sizing, trend management, nothing — uniformly applied in all 7,000 cases. As far as useful conclusions go, the implied shortfalls here are so glaring, they might as well be doing flight tests on jumbo jets made of cream cheese, or something of comparable idiocy.)
Bottom line: There is no substitute for getting in the thick of things, accumulating insights and a-ha moments born of real world experience, and stress-testing your research with hard-nosed verification of what does and doesn’t work. Rick Bookstaber has a useful take:
[In respect to how markets work,] a better analogy than physics or biology is a military one. The point is that there is a strategy of intelligent reaction to any action, an arms race to leapfrog one another in information gathering and technology, to know what others are doing, and to react in a way that they will not anticipate. This is the point where I could pull out quotes from The Art War about seeing into the mind of the enemy, attacking when your opponent believes you will retreat, and the like. That is not physics.
Nor is it hubristic academic findings with gaping holes big enough to drive a mack-truck through… so remember that next time a non-practitioner tries to lecture you on what can or can’t be done.