The ethics of Candy Crush Saga

A coercive monetization model depends on the ability to “trick” a person into making a purchase with incomplete information, or by hiding that information such that while it is technically available, the brain of the consumer does not access that information. Hiding a purchase can be as simple as disguising the relationship between the action and the cost as I describe in my Systems of Control in F2P paper.

Research has shown that putting even one intermediate currency between the consumer and real money, such as a “game gem” (premium currency), makes the consumer much less adept at assessing the value of the transaction. Additional intermediary objects, what I call “layering”, makes it even harder for the brain to accurately assess the situation, especially if there is some additional stress applied.

This additional stress is often in the form of what Roger Dickey from Zynga calls “fun pain”. I describe this in my Two Contrasting Views of Monetization paper from 2011. This involves putting the consumer in a very uncomfortable or undesirable position in the game and then offering to remove this “pain” in return for spending money. This money is always layered in coercive monetization models, because if confronted with a “real” purchase the consumer would be less likely to fall for the trick.

From The Top F2P Monetization Tricks by  Ramin Shokrizade. F2P stands for free-to-play and refers to games where you can start playing the game without paying. The most popular of such games, at least it certainly seems that way, is Candy Crush Saga. I know a few people who are struggling with a borderline addiction. When you add up the man-hours that have been dedicated to the game, it might be the most potent destroyer of productivity in recent human history.

Shokrizade makes a convincing case that Candy Crush Saga lulls the user into thinking they're playing a game of skill when in reality it's a game of luck (Shokrizade calls it a game of money, as noted in the excerpt below). Unable to make the distinction, players stuck on the more difficult levels later in the game end up making in-app purchases to keep feeding their addiction.

A game of skill is one where your ability to make sound decisions primarily determines your success. A money game is one where your ability to spend money is the primary determinant of your success. Consumers far prefer skill games to money games, for obvious reasons. A key skill in deploying a coercive monetization model is to disguise your money game as a skill game.

King.com's Candy Crush Saga is designed masterfully in this regard. Early game play maps can be completed by almost anyone without spending money, and they slowly increase in difficulty. This presents a challenge to the skills of the player, making them feel good when they advance due to their abilities. Once the consumer has been marked as a spender (more on this later) the game difficulty ramps up massively, shifting the game from a skill game to a money game as progression becomes more dependent on the use of premium boosts than on player skills.

Note that the difficulty ramps up automatically for all players in CCS when they pass the gates I discuss later in this paper, the game is not designed to dynamically adjust to payers. 

If the shift from skill game to money game is done in a subtle enough manner, the brain of the consumer has a hard time realizing that the rules of the game have changed. If done artfully, the consumer will increasingly spend under the assumption that they are still playing a skill game and “just need a bit of help”. This ends up also being a form of discriminatory pricing as the costs just keep going up until the consumer realizes they are playing a money game.

It's tough not to admire the skill with which Candy Crush Saga was built. Something about destroying tiny colored candies satisfies some reptilian instinct in people in the most addicting way possible. The guile with which the game entices users into forking over for in-app purchases makes me a bit squeamish, but no more than a casino should.

If you're going to give in to the temptation of CCS, just know that it's largely a game of luck, and resist the temptation to fork over dough when you get stuck unless you know what's being done to your brain and you're okay with it. 

Big Bang of body types

Tennis players have been getting bigger for years. When Lendl emerged as a force in the early 1980s, rare was the top-ranked man who was more than 6 feet. A decade later, Andre Agassi, at 5-11, and especially Michael Chang, at 5-9, were considered undersize overachievers.

In those days, Boris Becker, at 6-3, contended that men’s tennis was not only trending tall but would eventually be dominated by players well over 6 feet. While part of his prediction has not come true, it may be premature to say he was wrong.

Heading into United States Open, 9 of the top 32 men’s players were at least 6-5 — the tallest being Karlovic and the American John Isner, at 6-10. During this summer’s hardcourt season leading into the United States Open, del Potro defeated Isner in the finals in Washington. Raonic, a 22-year-old Canadian, made the final in Montreal, where he lost to Rafael Nadal. Isner upset No. 1 Novak Djokovic in the quarterfinals in Mason, Ohio, and outlasted del Potro in the semifinals before losing in two tiebreakers to Nadal.

“Every sport is going up and up,” Lendl said. “Look at basketball — and I don’t understand basketball — but I do know that the guys who were playing center before are playing wings now, or whatever you call them.”

More by Harvey Araton on how the average height of top men's tennis players has been creeping up. Patrick McEnroe discusses how he believes the sweet spot is from 6' 1" to 6' 4" just given the importance of agility and balance and footspeed and the difficulty of reaching low balls for someone of John Isner's height (6' 10").

A few years back a few of us from Hulu got tickets from one of our advertisers to go to the ESPY's in Los Angeles. At the event, Isner received an ESPY for Best Record Breaking Performance for his 11 hour 5 minute match against Nicolas Mahut at Wimbledon earlier that year. As Isner walked past us after he left the stage, I couldn't believe he was a tennis player, he was one of the tallest people I'd ever seen, as if Jeff Goldblum in The Fly had instead been mistakenly fused with a praying mantis or a giraffe. To serve from that height must be glorious, like hitting an overhead into a swimming pool.

In his great book The Sports Gene, David Epstein notes the gradual shift in each sport towards the optimal body type for that sport and away from a single ideal for the best human configuration. Sports scientists Kevin Norton and Tim Olds call it the "Big Bang of body types."

But, as Norton and Olds saw, as winner-take-all markets emerged, the early-twentieth-century paradigm of the singular, perfect athletic body faded in favor of more rare and highly specialized bodies that fit like finches' beaks into their athletic niches. When Norton and Olds plotted the heights and weights of modern world-class high jumpers and shot putters, they saw that the athletes had become stunningly dissimilar. The average elite shot putter is now 2.5 inches taller and 130 pounds heavier than the average international high jumper.

When they connected the dots from 1925 to the present for each sport, a distinct pattern appeared. Early in the twentieth century, the top athletes from every sport clustered around that "average" physique that coaches once favored and were grouped in a relatively tight nucleus on the graph, but they had since blasted apart in all directions. The graph looked like the charts that astronomers constructed to show the movement of galaxies away from one another in our expanding universe. Hence, Norton and Olds called it the Big Bang of body types.

It's not just body shapes but individual body parts that matter in particular sports. For some sports, it's better to have long legs and short torsos (basketball and volleyball). For other sports, it's good to have long arms and short legs (boxing). 

Body type can affect performance in different temperatures. Paula Radcliffe defies the generally accepted optimal body type for endurance runners. Most world class long distance runners are small and thin, giving them a larger skin surface area versus their body volume, helping them dissipate heat more efficiently. Radcliffe was a great runner in cool temperatures, but in her two highest profile races in heat, the 2004 and 2008 Olympic marathons in Athens and Beijing, held in 95 and 80 degree heat and humidity, Radcliffe was never in contention.

If you want your child to be world class in a sport, it really matters what type of body they have. You can read Epstein's book for a richer documentation of what some of those might be, but here's one hint: if your child is over 7 feet tall and reasonably coordinated, put a basketball hoop in your driveway. 

The pitcher who conquered MLB's best hitters

In 2004 and '05, Finch hosted a regular segment on Fox's This Week in Baseball in which she traveled to major league training camps and transformed the world's best baseball hitters into clumsy hacks. "Girls hit this stuff?" asked an incredulous Mike Cameron, the Mariners' outfielder, after he missed a pitch by half a foot.

When seven-time National League MVP Barry Bonds saw Finch at the Major League All-Star Game, he walked through a throng of reporters to talk trash to her. "So, Barry, when do I get to face the best?" Finch asked.

"Whenever you want to," Bonds replied confidently. "You faced all them little chumps.... You gotta face the best.

"You can't be pretty and good and not face another handsome guy who's good," Bonds added, spreading his peacock feathers. He then told Finch to bring a protective net because, he said, "you're going to need it with me.... I'll hit you."

"There's only been one guy who touched it," Finch replied.

"Touch it?" Bonds said, laughing. "If it comes across that plate, believe me, I'ma touch it. I'ma touch it hard."

"I'll have my people call your people, and we'll set it up," Finch said.

"Oh, it's on!" Bonds said. "You can call me direct, girl. I take my challenges direct.... We'll televise it too, on national television. I want the world to see."

So Finch traveled to Arizona to face Bonds in spring training, and after he watched several of her pitches fly by, the raillery stopped. He insisted that the cameras not film him batting against her. Finch shot pitch after pitch past Bonds as his Giants teammates pronounced them strikes. "That's a ball!" Bonds pleaded, to which one of his teammates replied, "Barry, you've got 12 umpires back here."

Bonds watched dozens of strikes go by without so much as swinging. Not until Finch began to tell Bonds what pitches were coming did he tap a meek foul ball a few feet. He taunted her, "Go on, throw the cheese!" She did, and blew it right past him.

Finch visited Alex Rodriguez, who was then starring for the Rangers, at another spring-training park, in 2003, and Rodriguez watched over her shoulder as she threw warmup pitches to a Texas bullpen catcher. The catcher missed three of the first five throws. Before Rodriguez stepped into the batter's box, he made it clear he wouldn't dare swing at any of Finch's pitches. He leaned forward and told her, "No one's going to make a fool out of me."

An excerpt from David Epstein's new book The Sports Gene, which I'm about a third of the way through and enjoying quite a bit. 

One of the first mysteries he tackles is this one: why did MLB's best hitters, who have to hit baseballs that travel to home plate in about 400 milliseconds from just under 60 feet 6 inches away, struggle to even make contact with a much larger softball traveling to home plate in the same amount of time (thrown slower at 68mph, but from a shorter distance).

Epstein notes that the average time for a major league hitter to initiate muscular action is about 200 milliseconds, meaning baseball players must decide to swing at a baseball before it's even halfway to home plate. To make that decision, baseball players try to anticipate the pitch being thrown by looking at the pitcher's delivery motion. 

This explains why certain pitches and pitchers are so effective. Take Mariano Rivera, for example. For much of his career, he threw just one pitch, the cutter. How can a pitcher survive throwing one pitch such a high percentage of the time? In fact, this season he's thrown the cutter 89% of the time. 

The key is late movement. The batter has to decide to swing at the ball and where to swing before the pitch is halfway to home plate, but the pitch tends to move laterally and downwards, away from the arm side, very late in its path to the plate. By then it's too late for a hitter to adjust his swing path.  The same principle applies to the slider which may be the single pitch most responsible for the rise in strikeouts in the modern era. It comes in looking like a fastball, and the best sliders move both sideways and away from the arm side very late, too late for a hitter to do anything about his swing path.

Why is Yu Darvish such a tough pitcher to hit? Part of it is his filthy and broad repertoire of pitches, but another is the fact that he manages to deliver every pitch to multiple locations with the same exact motion, leaving the hitter with fewer cues to try to guess which pitch is coming. This animated GIF that circulated earlier this year illustrates this deception beautifully. 

Even skills that appear to be purely instinctive, such as jumping to rebound a basketball after a missed shot, are grounded in learned perceptual expertise and a database of knowledge about how subtle shifts in a shooter's body alter the trajectory of the ball. Without that database, which can be built only through rigorous practice, every athlete is a chess master facing a random board, or Albert Pujols facing Jennie Finch: He is stripped of the information that allows him to predict the future.

Since Pujols had no mental database of Finch's body movements, her pitch tendencies or even the spin of a softball, he could not predict what was coming, and he was left reacting at the last moment. And Pujols's simple reaction speed is downright quotidian. When scientists at Washington University in St. Louis tested him, perhaps the greatest hitter of his era was in the 66th percentile for simple reaction time compared with a random sample of college students.

It's true, some hitters will guess that a certain pitch is coming, and if they guess incorrectly they can look silly, letting a fastball right down the middle go by without lifting the bat off their shoulders. Being labeled a "guess hitter" was a stigma.

From Epstein's summary of how hitters operate at a neurological level, however, it turns out almost every hitter is guessing in some ways, from the moment the pitcher starts his motion to just after the baseball is released.

Asiana Flight 214

Patrick Smith of Ask the Pilot  fame provided an informed critique of many of the hasty judgments people grasped at wildly after Asiana Flight 214 crashed at SFO. Among those:

Lastly, we're hearing murmurs already about the fact that Asiana Airlines hails from Korea, a country with a checkered past when it comes to air safety. Let's nip this storyline in the bud. In the 1980s and 1990s, that country's largest carrier, Korean Air, suffered a spate of fatal accidents, culminating with the crash of Flight 801 in Guam in 1997. The airline was faulted for poor training standards and a rigid, authoritarian cockpit culture. The carrier was ostracized by many in the global aviation community, including its airline code-share partners. But Korean aviation is very different today, following a systemic and very expensive overhaul of the nation’s civil aviation system. A 2008 assessment by ICAO, the civil aviation branch of the United Nations, ranked Korea's aviation safety standards, including its pilot training standards, as nothing less than the highest in the world, beating out more than 100 other countries. As they should be, Koreans are immensely proud of this turnaround, and Asiana Airlines, the nation's No. 2 carrier, had maintained an impeccable record of both customer satisfaction and safety.

Whatever happened on final approach into SFO, I highly doubt that it was anything related to the culture of Korean air safety in 2013. Plane crashes are increasingly rare the world over. But they will continue to happen from time to time, and no airline or country is 100 percent immune. 

Malcolm Gladwell arguably did more than anyone to popularize the theory that Korean culture lay at the root of Korean Air's poor safety record in the 1980's and 1990's. In this interview with CNN Money, he summarized his theory from his book Outliers:

F: You share a fascinating story about culture and airline safety.

G: Korean Air had more plane crashes than almost any other airline in the world for a period at the end of the 1990s. When we think of airline crashes, we think, Oh, they must have had old planes. They must have had badly trained pilots. No. What they were struggling with was a cultural legacy, that Korean culture is hierarchical. You are obliged to be deferential toward your elders and superiors in a way that would be unimaginable in the U.S.

But Boeing (BA, Fortune 500) and Airbus design modern, complex airplanes to be flown by two equals. That works beautifully in low-power-distance cultures [like the U.S., where hierarchies aren't as relevant]. But in cultures that have high power distance, it's very difficult.

I use the case study of a very famous plane crash in Guam of Korean Air. They're flying along, and they run into a little bit of trouble, the weather's bad. The pilot makes an error, and the co-pilot doesn't correct him. But once Korean Air figured out that their problem was cultural, they fixed it.

A fairly thorough rebuttal to Gladwell's theory was posted at the blog Ask a Korean:

First, the way in which Gladwell quoted the transcript is severely misleading. This is the full transcript, which goes from pp. 185 to 187 of the NTSB report:

CAPTAIN: 어... 정말로... 졸려서... (불분명) [eh... really... sleepy... (unintelligible words)]

FIRST OFFICER: 그럼요 [Of course]

FIRST OFFICER: 괌이 안 좋네요 기장님 [Captain, Guam condition is no good]

FIRST OFFICER: Two nine eighty-six

CAPTAIN: 야! 비가 많이 온다 [Uh, it rains a lot]

CAPTAIN: (unintelligible words)

CAPTAIN: 가다가 이쯤에서 한 20 마일 요청해 [Request twenty miles deviation later on]

FIRST OFFICER: 네 [yes]

CAPTAIN: ... 내려가면서 좌측으로 [... to the left as we are descending]

(UNCLEAR SPEAKER): (chuckling, unintelligible words)

FIRST OFFICER: 더 오는 것같죠? 이 안에. [Don't you think it rains more? In this area, here?]

(emphases mine)

Note the difference between the full transcript, and the way Gladwell presented the transcript. Gladwell only quoted the first two lines and the last line of this sequence, omitting many critical lines in the process. In doing so, Gladwell wants to create an impression that the first officer underwent some period of silent contemplation, and decided to warn the captain of the poor weather conditions in an indirect, suggestive manner. 

The full transcript reveals that this is clearly not the case. The first officer spoke up directly, clearly, and unmistakably:  "Captain, Guam condition is no good." It is difficult to imagine how a person could be more direct about the poor weather condition. Further, there was no silent contemplation by the first officer. Nearly three minutes elapse during this sequence, during the captain and the first officer chatted constantly. And it is the captain who first brings up the fact that it is raining a great deal: "Uh, it rains a lot." In this context, it is clear that the first officer is engaged in some friendly banter about the rain, not some indirect, ominous warning about the flight conditions.

To be fair to Gladwell, when asked about whether he thought his theory from Outliers came into play in the case of Asiana Flight 214, he did not bite. 

We asked Malcolm Gladwell for his thoughts on the use of his essay in the particular context of the Asiana crash. "I can understand why my Outliers chapter has been of interest, given how central cockpit communication issues are in plane crashes," Gladwell told The Atlantic Wire in an email, adding, "My sense is that we should wait for the full report on the crash before drawing any conclusions about its cause." As for the applicability of his work to the recent Asiana crash, Gladwell noted that his essay was specific to the problems (and solutions) of one airline — Korean Air, "which I think did an extraordinary job of addressing the cultural issues involved in pilot communication. This was a crash involving a completely different airline," he said. 

The NTSB has yet to issue any formal assessment of what happened that day. I happened to arrive at SFO for a flight to Paris just a few hours after the crash occurred, and we could see the wreckage in the distance from our gate. Our flight was delayed by 9 hours, and we finally took off at around 9pm that night.

Our plane turned onto the runway and accelerated towards takeoff. The moment just before our wheels left the ground, I saw, just out the window to the left, the wreckage of Asiana Flight 214, sitting just off our runway, illuminated by some giant spotlights, like a giant burned out metallic skeleton that had just been dug out of the earth.

Racial diversity in American cities

I killed a good half hour playing with this racial diversity dot map. It visualizes some of the spatial racial distribution of cities that you can only intuit from ground level as a resident. While a city may seem quite integrated and diverse, it isn't until you zoom in that you see that what looks like a diverse blob of people is really a series of segregated neighborhoods. 

One personal hunch, though, is that how diverse a city feels is not just based on how segregated the population is based on their place of residence but how much those populations interact in day to day life, and that is a function of city density and the developmental maturity of the city's public transportation. While New York City looks like Chicago or San Francisco or other big cities in being a collection of segregated neighborhoods, it felt like the most diverse city I've ever lived in because those populations crossed paths on the city streets and subways every day in high numbers. 

Thomas Schelling's Segregation Model, one of the more powerful agent based models I've ever studied, shows how the extreme segregation of American cities might arise from much milder racial preferences. It's a critical model to study, one that is useful to keep in mind when trying to avoid a variant of fundamental attribution error when trying to explain how something that builds over time ended up in a bad state.

This report (PDF) from 2011 studies long term racial segregation trends in America and comes to this conclusion:

The 2010 Census offers new information on changes in residential segregation in metropolitan regions across the country as they continue to become more diverse. We take a long view, assessing trends since 1980. There are two main findings: 1) the slow pace of lowering black-white segregation has continued, but there is now some change in the traditional Ghetto Belt cities of the Northeast and Midwest; and 2) the rapidly growing Hispanic and Asian populations are as segregated today as they were thirty years ago, and their growth is creating more intense ethnic enclaves in many parts of the country.