The Moralgorithm

At the height of performance, great athletics seems inexplicable, like something conjured. Even the athletes are at a loss to explain what goes on in those moments that are always vaguely articulated as “flow,” or being “in the zone,” or what Zen students of samurai swordsmanship called mushin. Ineffable or not, the effect of deep physical focus is apparent enough to spectators, and nowhere more so than in basketball, where the initiative changes every 24 seconds at most, and where pace and team play make the game a sequence of runs and rallies, charges and comebacks. Or lopsided meltdowns, as when the San Antonio Spurs knocked the Detroit Pistons on their heels in the opening minute of Game 2 of the NBA Finals, 8-0. One game down already, the Pistons turned cold, and San Antonio eased into mushin, winning by 21 points. This came as a surprise to those who thought the series would expose the Spurs’ soft spots and reconfirm the Pistons’ supremacy. But not to me. I’d simulated the game in advance several times on NBA Live 2005 for Xbox. (I like to play the Spurs because I like the Spurs: Their team feels like being at a discotheque on Corfu or something, with Tony Parker’s supercilious Parisian charm, Manu Ginobili — a.k.a. the ballerina from Argentina — cutting up the rug, and soft-spoken Tim Duncan, the Caribbean expat, quietly running the bar.) And I knew the nuanced Spurs had it in them to beat the hard-driving Pistons, because I’d seen Parker and Ginobili’s Xbox avatars shut down Chauncey Billups and Rip Hamilton before. It may always be a challenge — near impossible sometimes — to defeat the champs on NBA Live; but it can be done, especially if you can discombobulate them early. I’ve noticed the same thing with any strong team in the game: If the CPU gets off to a solid start, you’re in trouble. The only hope then is to disrupt their flow. That assumes that the CPU has some flow to disrupt. “It’s in the program,” I theorized about this to BACOBIT, a friend I sometimes play against on Xbox Live. The flow factor plays a role against other players as well. I squandered a 10-point lead with the Spurs to BACOBIT’s Pistons and my guys started growing left feet for all their limbs. There must be some kind of algorithm that generates morale, I suggested, a moralgorithm — and my team’s was off. “It’s not the computer,” said BACOBIT, “it’s your moralgorithm. When you’re off, you hit the buttons too late, judge the plays wrong, lose patience and shoot funny. Don’t blame the game, bro. Yours is the algorithm that’s losing.” Could be that too. NBA Live gives the player extremely detailed control, so your shot timing or passing or playmaking is affected if you get emotional or anxious or have clammy hands. “But that’s a good question,” said BACOBIT, reconsidering. “Is it in your head — or in the Xbox? Just where is the algorithm?” At least some of the ghost is in the machine. Gathered on one floor of the Electronic Arts campus in Vancouver are the 53 people responsible for building new versions of the NBA Live game each year. Over the years, explained the game’s producer, Todd Batty, dozens of programmers have indeed tried to quantify the intangible by writing several hundred lines of code that govern what Batty calls “in-game momentum.” Dozens of variables affect the game’s computational assessment of flow. Home crowds, big baskets, surprise bench scoring, time-outs, even the coach’s known record in difficult situations, can drive momentum. Individual players, too, go on hot and cold streaks. Three badly chosen, off-balance leaning jump shots that miss will wear down a player’s confidence, and reduce his percentage of hitting the next shot. And vice versa: A string of nice baskets will give the next one a little boost. “Perform well, and it feels like the stars are aligning — your guys jump a little higher, or dribble a little faster,” Batty said. “Or sometimes you’re just digging yourself a hole, and nothing works right. We tried to create that to make the game play like real basketball.” The programmers have not yet interviewed neurologists or looked into Mihaly Csikszentmihalyi’s Flow, but they do go through reams of data to assign all the intricate multipliers to the moralgorithm. “We watch lots of game footage and collaborate with NBA statisticians,” said Batty. “We need to see how much better a team actually shoots when they’re on fire. And what the stats are for teams that step up in the last two minutes of a tight game. We analyze a lot of information.” I told Batty that I’d occasionally seen the moralgorithm needle hit the red; that every so often the game goes supernatural. Like one matchup, where I was playing the Sonics against the Heat on Xbox Live, and I could tell something really clicked for me when Jerome James, my oafish center, hit two three-pointers. Then Vladimir Radmanovic, whose little polygonal computer face seemed to look as surprised as I was, stole the ball from 10 feet away and passed to Antonio Daniels on a fast break that drew so much moralgorithmic momentum, I swear Ray Allen levitated up court for the most improbable alley-oop of all time. A glitch? Perhaps. Over time, Batty says, the programmers have fixed most real glitches in the game. It could also be that those improbable pyrotechnics are accidentally reflecting that fundamental intangibility in real basketball. What are the flashes of mushin in sports if not human glitches — momentary suspensions of reality? Maybe spontaneous supernatural power is not just a digital artifact. Afer all, wasn’t Dr. J’s famous floating reverse adjusted no-look layup in the 1980 finals an error of some kind, a brief, one-in-a-million window into the realm of impossibility? “One more step towards human irrelevance,” said BACOBIT when I explained that NBA Live’s moralgorithm is both real and remarkably well-developed. But it’s not (yet) irrelevance, so much as technological entanglement — the mushin belongs to neither man nor the machine alone. It’s a recombinant program, a digital-analog continuum of flow residing somewhere in between. If it’s disconcerting that we’re headed into a future where computers supply half the emotion, there may be some consolation. Batty admits that there are still plenty of intangibles that the Electronic Arts programmers have yet to translate into code. “Like team chemistry,” Batty said, “which is now quantified only very simplistically.” A good example of that, I offered, was the Phoenix Suns, whose cumulative rating in the game did not correspond to their actual playing this season. (NBA Live computes team ratings by combining the players’ characteristics for a score out of 100; the Suns’ NBA Live percentage was never the highest despite this year’s record — the best in the league and the most dramatic turnaround in history.) “We noticed that too,” Batty said. Then again, he observed, even in NBA Live, the Suns outperformed their actual stats. “When we were finishing the 2005 version, we updated the rosters, with Quentin Richardson and Joe Johnson, tweaking Amare Stoudemire’s settings and so on. And in testing, Phoenix became everyone’s favorite team to play with.” Fun, quick, hard to beat — just like the real players — the digital Suns formed a whole greater than the sum of the parts. And don’t forget, Batty noted, that the NBA Live production team discovered the Suns’ power before the regular season even started. They knew then that the team was going places. Apparently an emergent property of the game’s overall intricate detail, chemistry seemed to function anyway. And without any special coefficients, the game still provided an uncanny predictive power in the case of the Suns. That didn’t surprise Batty, who said that NBA Live also forecast how the playoffs would shake out. The production team can simulate a set of games thousands of times over and average the results for an expected outcome, and as much as Batty wanted to see the Suns make the Finals (Steve Nash is from nearby Victoria), their playoff model always pointed toward a Spurs-Pistons face-off. “What did the eight ball say about that outcome?” “Spurs are going to take it.” NBA LIVE 2005 | Electronic Arts | $20–$30