Mental floss's intrepid reporter immerses himself in calculus to master the game.
(Image credit: Nazario Graziano)
The bus to Atlantic City is oversold, over-air-conditioned, and struggling to get out of Manhattan. Normally, I’d appreciate the irony that Greyhound dubs this shuttle the Lucky Streak, but right now I’m too busy sorting through my notes about implied odds, effective value, and something called “M-ratio.”
Two weeks ago, this pile of equations would have meant nothing to me. Today, however, it means next to nothing. A marginal improvement, sure, but isn’t massaging the margins what gambling is all about?
Poker Theory and Analytics is a graduate-level MIT course taught by Kevin Desmond, a former pro player and Morgan Stanley analyst. The school offers the course online, meaning video lectures, assignments, and class notes are available to anyone for free. Inspired by Bringing Down the House, the 2003 book about the MIT Blackjack Team who used their card-counting smarts to outwit Vegas, I formulated a simple plan: Take the class, hit the poker tables of Atlantic City, and profit.
The Jersey Turnpike, however, has a way of shaking one’s confidence.
I’m what seasoned poker players would call a “donkey.” I’ve played only small games with friends, and every hand I’ve ever won has been the result of pure luck (try as I might to convince myself otherwise). I lack every quality required of good poker players: risk assessment, pattern identification, stoicism, basic math proficiency, and attention span. If poker can be taught, as MIT’s course materials suggest, it’ll be put to the test here not by genius-level MIT students, but by a bumpkin who barely knows his multiplication tables.
But why would MIT offer a course on poker in the first place? According to its official overview, the class “takes a broad-based look at poker theory and applications of poker analytics to investment management and trading.” The bulk of the course consists of eight video lectures. One is guest-led by poker player, author, and financial risk manager Aaron Brown and covers the history of poker and how it relates to economics.
Poker is an American game (invented on the frontier in the early 1800s) with American sensibilities (the decidedly anti-monarchical bent that ranks the ace above the king). But what made it truly special was its use of chips—a novel idea at the time. These markers freely flowed between individuals, creating upstart economies complete with risk, debt, and credit, all in a time and place where actual currency was sparse and stagnant.
It makes sense, Brown asserts, that the first futures markets sprouted up in poker-crazy parts of the country, some two decades after the game first became popular. “Futures exchanges are populated by tough, brawling innovators who often make fortunes or lose fortunes,” Brown tells the class. Poker games are named after places that were populated by these types of people—Texas, Omaha, Chicago, etc. That’s why, he argues, “there is no poker game named after any place except places where, if you lose all your money in a game … you float down to New Orleans.”
This history is why the game once conjured images of Stetson-wearing toughs bluffing through cigarillo smoke. The rise of online poker means that today’s stereotype is less Maverick, more Mark Zuckerberg. Now, players can rapidly play through multiple tables and tournaments simultaneously, amassing years’ worth of experience in just a few days.
Students who took MIT’s course for credit (and not Internet observers watching later, like me) were asked to rack up hours in a private league created for the class by PokerStars, a major online gambling site (the students used fake money). They were granted free access to a poker tracker that enabled them to archive and tabulate their statistics. It was odd to see such product placement in a college class—both the online league and poker tracker were heavily branded—but I’d rather not clutch pearls when I’m learning how to better separate people from their money.
The course focuses on Texas Hold ’Em, a popular game you may have seen on ESPN’s annual World Series of Poker broadcast. While the goal is ostensibly to have the best combination of cards, it’s just as important to wear your poker face—either to convince everyone you have the best cards (and scare them out of betting against you), or the worst cards (and sucker them into betting against you).
Everybody playing Texas Hold ’Em starts with two cards. Then players take turns placing bets. You can “call,” or match the current bet, “raise,” or up the current bet, or “fold,” and throw away your hand, leaving any chips you’ve bet on the table.
A dealer then lays down shared community cards on the table faceup. This is called “the flop.” After a round of betting, a fourth card, “the turn,” is laid out. Players bet again, followed by a fifth card, “the river,” and then one last round of betting. Whoever has the best five-card combination wins.
It’s a simple game made more complicated (and fun) by the infinite number of factors in play—namely, the qualities of the other humans you’re up against. It’s a nonstop mind game in which players must figure out why, or why not, competitors are betting. As the old poker saying goes, you play the players, not the cards.
There is math involved, of course; MIT isn’t known for its mind-reading classes. While Kevin Desmond does offer some broad insider tips early on in the course, like the best times to play (“a lot of the newer guys only play poker on the weekend”), the workload is heavily analytical.
As MIT students (even those of us watching in our underwear at home), we would be learning to rely on numbers, not hunches. Betting or folding—the life-or-death decisions made at a poker table—are matters of calculated probability. “Expected value is the same in poker as it is in math,” Desmond says, not helping this lifetime C math student one bit. “It’s win percentage times win amount minus lose percentage times lose amount.” I pause the video, which is titled “Basic Strategy,” to write this down. It doesn’t help. I’m lost.
My ears perk up when Desmond brings up bluffing. Finally, I think, some instruction on how to steel my guile with some sexy poker deception. “We’re going to have to use calculus for this,” he says, bringing up a slide with a curved line graph. My heart sinks—I find myself back in summer school math class. A key difference is that now I actually have an answer to that classic slacker refrain: “When will I have to use this in the real world?” I was going to Atlantic City in two weeks to play a poker tournament.
Luckily, I’ve got a genuine ace up my sleeve: my friend Will. Will has been playing since the online poker boom in the early 2000s, starting as a precocious high schooler. I’d watched him play dozens of tables at once, Bobby Fischer–like, spread across two massive computer monitors. He could tell me the hand history and style of any given player, like a hummingbird returning to a crowded field, knowing precisely which flowers had already been pollinated.
When I hit him up, he’d just returned from a summer of playing tournaments in Las Vegas, South Korea, and Monte Carlo. But he only got into live games once the government cracked down on online poker. The adjustment wasn’t easy—he had to teach himself how to play in person. The toughest change, he says, was learning to cope with the boredom of playing only one hand at a time. I asked him to watch some of the MIT videos. “Some of this stuff,” he says, laughing, “is beyond me.” He had watched a lecture on game theory led by computer scientist and professional poker player Bill Chen. One key element Chen covers is “regret minimization,” which I gather is a way to determine how adversaries are playing, and what their next move will likely be. It was explained like this: R*T/k = T/∑/t=1*ut *(σk) – ut (σt)
I ask Will if he knows what all this alludes to, and he does. “I just don’t think of it like that,” he says with a shrug. “You just have to kind of internalize vague types of these ideas.”
Poker, I realize, is a skill in the way language is a skill. It’s a set of rules under a structure of infinite nuance and variance. Professionals separate themselves from the pack with an ingrained understanding of these nuances—smart decisions, made instinctively. I couldn’t expect to learn a language in two weeks, and poker would be no different. All I could hope to do is pick up enough of the basics to survive.
Early in the course, Desmond explained the four types of poker players:
1. Tight-aggressive: You bet only when you have a good hand, but when you do, you don’t back down.
2. Loose-aggressive: You bet often, but you don’t let people push you into folding.
3. Tight-passive: You rarely bet, and when the action gets hot, you’re content to fold away.
4. Loose-passive: You call all bets without dictating the game.
The only players who win, Desmond says, are the aggressive types. As for passive players, “There’s virtually no way that these guys are making money in poker.”
From there, we covered more complex concepts. Your “effective stack” is “the most chips you can lose in the hand.” My “M-ratio,” an equation popularized by poker pro (and tight-aggressive archetype) Dan Harrington, is that effective stack divided by the sum of the “blinds,” default bets players have to make to play the game, and “antes,” raises to stay in the game. The closer that number gets to zero, the more vital your need to win, and this helps dictate how aggressively you should play. “In tournaments,” Desmond says, “most of your value is going to come from what you do preflop,” meaning before a single community card is shown. If you’re going to play well—aggressively and smart—you’re going to have to do so as early in the game as possible.
I’m still studying my cheat sheet of the best hands as the towering casino-hotel complexes of Atlantic City come into view. I remind myself what kind of player I want to be, and it becomes my mantra as we speed past marshland down the long access road: tight-aggressive, tight-aggressive, tight-aggressive. Windbreakers crinkle as excited passengers shift in their seats. Optimism fills the Lucky Streak, and it’s contagious. MIT’s Poker Theory and Analytics treated luck as an irrational variable, but the subtext was always there: It helps if you have it.
For $45, Will and I sign up for an afternoon tournament at Bally’s poker room. The first thing I notice is how quiet it is—the cacophony of the main casino floor seems far, far away.
It’s probably not a good endorsement of my character, but casinos put me at ease. Entering one, you become a citizen of a domineering surveillance state, and there’s some perverse comfort in that simplicity. Like windows and clocks, ambiguity has no place here. There are clear rules and, as long as you play by them, you are A-OK in the casino’s book. Heck, you might even make a few bucks! It may seem like an Orwellian nightmare, but Orwell never had a hot night at the craps table.
The poker room feels different from the rest of the casino. Gone are the crystal-clear roles of player versus house. In the poker room, it’s human versus human, and the benevolent dictatorship that is the casino can only watch. (Well, they also take entry fees or a small percentage of every bet, called “the rake.”) The people here have agency and control, and the air weighs heavy with consequence.
Despite the tension, this is about as low-stakes as poker tournaments get. Most poker pros won’t even get out of bed for $45, let alone waste a few hours playing in a tournament.
Another player in the sign-up line excitedly asks if Will and I have played before. Will points at me and says, “This guy’s been studying poker at MIT.” “Wow, that’s a great school,” the guy replies, and I shrink inward. Before I can elaborate, he explains that this is his first-ever poker tournament and that he’s been walking around for 15 minutes trying to find where he’s supposed to pick up his chips. If this is a hustle, he certainly is committed to it.
Players are allowed to rebuy in this tournament, meaning those who lose can still purchase more chips with which to continue to play. By the time I get settled, some players have already taken advantage of this, and their initial chips have gone to other players who now have a distinct advantage. I’m chasing the pack before I’ve placed a single bet. My first action is to call a bet—matching an opponent’s current bet instead of raising it. It’s a passive move that to the rest of the table might as well be a tattoo on my forehead reading chump. Already, I’ve ignored my tight-aggressive mantra.
With a few exceptions, calling is often a sign that you just want to live long enough to see more cards. When the dealer reveals the flop—the first three community cards—it reveals the straight I’ve been chasing is no longer a possibility. A middle-aged man across from me wearing a baseball cap and sunglasses (nice poker getup, albeit overboard) raises me more than half of my chip total. Even though I had good cards (ace-queen) going in, I’m forced to fold, forfeiting the chance to find out whether he was bluffing or truly had me beat.
I showed weakness and let an opponent muscle me out of playing a good hand. I couldn’t help but feel like I had let MIT down as the dealer shoved my share of chips to the other side of the table.
I overcorrect and start playing unhinged, or as Desmond would say, loose-aggressive. At first it works, and I take my turn forcing players into handing over blinds they clearly aren’t confident they can keep. The guy to the left of me keeps shooing away his friend who asks when he’ll be done so they can go eat. If he really wanted to go hang out with his buddy, I think to myself, then he’d have pushed all-in by now. But he’s hanging on to his chips for dear life, playing tight-passive, so his blinds are mine for the taking.
Unlike Will—absentmindedly watching football while playing on autopilot at an adjacent table—I soon find myself overwhelmed by the pace and start to lose track of everyone’s bets. Even though there’s only $45 on the line, the undulating stacks of chips in front of me make it seem like so much more. I lose a few hands, and those once-proud stacks dwindle to a single column.
Then it hits me: This is my effective stack. The players around me fade away and I’m back in MIT’s virtual classroom. I divide my stack by the sum of the blinds and antes on the table to get my M-ratio. It’s a hair over zero. The math is clear: I have to go all-in and bet everything. Desperate it may be, but my decision is all analysis, no guesswork involved.
One other player—a confident, quiet guy three seats to the left who has been playing tight-aggressive to a T all afternoon—calls. We show our cards.
My queen-seven off-suit isn’t as good as his hand—queen-10 of clubs—though it isn’t tragically far behind.
The flop comes: two fives and a jack, one of the fives bearing clubs.
Then, the turn—the ace of clubs. If the next card also shows clubs, I’m toast—he’d have five cards of the same suit, a flush.
The next card is flipped over: It’s the queen of diamonds, meaning we both have the same winning hand: a pair of queens and a pair of fives, with the ace serving as a mutual high card. It’s a tie, but it feels like a win.
Eventually, however, I lose. I won’t bore you with the details, but I can assure you: I was unlucky. That you can play well and still lose is a fact that haunts poker players at every level; it’s a simple truth that can make high-level MIT courses seem comically futile. Hidden beneath all the numbers was an unavoidable fact: Sometimes your luck just runs out.
But then, an announcement comes on over the PA: “Ten minutes left until rebuy closes.”
I wonder what the odds are of suffering a bad beat like that again. I then ask a better question: What are the odds I’ll play as weakly as I just did? MIT couldn’t prevent that from happening, but it did help me diagnose my poker ills. Fixing them could get expensive.
Bolstered by the confidence that can come only with a combination of empirical data and a little experience, I make my way to the teller window, $45 cash in hand.
The article above by Nick Greene appeared in the January 2016 issue of mental_floss magazine. It is reprinted here with permission.