rec.autos.simulators

>Fortune 500, Meet Daytona 500
>Slate ^ | 17 feb 03 | Charles Duhigg

>When Michael Waltrip feinted briefly, sped in close behind another
>car, and then whipped past Jimmie Johnson to win the rain-shortened
>Daytona 500 on Sunday, the crowds surrounding the track, screaming
>through mouthfuls of $7 chili dogs and a sea of Confederate flags,
>cheered with wild glee. The E***ment! The Noise! The Complexity
>Theory!

>So an economics professor goes to the Daytona 500 It sounds like the
>beginning of a bad joke. But for game theorists, stock car racing is
>emerging as an unlikely laboratory for understanding the world. A 2002
>Rand Corp. paper titled "Social Science at 190 MPH on NASCAR's Biggest
>Superspeedways" suggests that NASCAR should interest academics because
>it offers an opportunity to study a complexity rarely seen in other
>sports but much evident in the real world: the tension between
>cooperation and competition that is necessary for modern victory.

>The NASCAR laboratory is a product of simple physics. Unlike other
>sports, which are largely determined by individual athletic ability or
>team strength, NASCAR requires its competitors to cooperate in order
>to win. The forced cooperation is a product of two factors. First,
>NASCAR imposes engineering restrictions that prevent any driver from
>attaining a major equipment advantage over rivals. And second,
>"drafting" allows cooperating cars to go faster. In the 1960s, drivers
>discovered that if one stock car closely follows another, drafting in
>its wake, both cars increase speed. The lead car benefits from a drop
>in air resistance that comes when the slight vacuum at its rear wheels
>is filled. And the second car speeds up because it's partially
>protected from the wind. The more cars that are lined up, the faster
>each car goes. Hence the monotonous rows that develop around the oval
>track at Daytona.

>Drafting raises all sorts of delightful game theory possibilities. As
>long as two racers stay in a partnership, they can catch up with or
>pass other cars that are not drafting. Partnering doesn't bring home
>the trophy, though. To win, a racer must defect and pass his
>opponents, but as David Ronfeldt, the author of the Rand paper
>explains, when "a racer in a line wants to break out and get ahead he
>needs at least one partner. If he swings out alone, he is bound to
>lose momentum" and lose the race, leading to the NASCAR axiom that "it
>takes two to pass one." So, cars engage in a delicate dance of game
>theory and false allegiances, forming partnerships at 190 mph to
>overtake the leader until a betrayal or defection leaves one car
>falling back while the other partners with a new rival.

>The gamesmanship gets even more complicated. When two cars begin
>drafting, the second driver gains an advantage if he (and all NASCAR
>drivers are "he") drops back a length or two and then re-enters the
>draft zone and uses the tiny extra momentum to slingshot past the
>leader. Additionally, following cars can "fan the tail" of the leader,
>moving their bumper within inches of the lead car and fanning back and
>forth to disrupt the airflow over the vehicle. As the aerodynamics
>scatter, the lead car loses downward grip and the driver must slacken
>acceleration to avoid sliding sideways. So lead drivers will slow when
>they see followers begin to fall back or "mirror" the drivers behind
>them to prevent attacks.

>This tension between cooperation and competition is precisely where
>Ronfeldt sees applicability to the real world, because it turns out
>that similar dynamics exist all around us. When Microsoft and Cisco
>developed partnerships during the '90s with potential competitors such
>as Intel, Compaq, and Dell, they formed "drafting lines" that
>weathered the booms and busts that damaged lone drivers like Apple.
>Sen. Trent Lott's fall from grace and Sen. Bill Frist's ascendance can
>be explained in part by a series of defections of potential rivals
>from Lott to Frist that, once started, built a momentum that became
>impossible for Lott to slow. At the core of these events is a tenet of
>the emerging science of complexity: Cooperation can sustain a
>surprisingly high degree of competition, as long as the participants
>agree to some basic rules.

>But why do some partnerships work while others end in disarray? Again,
>NASCAR teaches lessons. Frist and Microsoft attract drafting partners
>for the same reason that Dale "The Intimidator" Earnhardt won races:
>They have enough "social capital" to convince competitors they will
>honor their partnership contracts, even if just for a while. In
>NASCAR, some of this social capital is achieved through agreements
>struck prior to a race or by radio communication during competition.
>(NASCAR's utility as a laboratory is enhanced by our ability to
>retroactively review communications and compare them to tapes of the
>race.) But more frequently it is a product of the NASCAR ethic of Hard
>but Fair. "I'll race clean because I know the rules," one driver
>explained. "But if you get too greedy, it's payback time." Put another
>way, even though we may be rivals in the future, we'll agree to trust
>each other right now.

>NASCAR also serves as an interesting model for what causes certain
>kinds of aggressive behavior. Economists at the University of Chicago
>made a surprising discovery when they began examining NASCAR
>accidents. While they initially assumed that aggressive driving and
>the attendant crashes occur when racers are eager to pass cars in
>front of them, it turns out the opposite is true. Drivers are more
>likely to crash when someone is about to overtake them. In other
>words, racers become more aggressive not when they think they can win,
>but when they're afraid they're going to lose.

>And so does everyone else. Research suggests that investment bankers
>are more prone to commit fraud when they feel the competitor at their
>heels. Students in school cheat not to get the "A," but to avoid the
>"C." From Indonesian***fighting to technological change,
>recklessness stems more from the fear of falling than the e***ment
>of new heights. "Don't be greedy" is how Cisco's CEO explains his
>strategy for transforming competitors into docile partners. The
>cooperation of NASCAR, or any other system it turns out, persists only
>when everyone believes he has the opportunity to win.

>These implications may be significant. America stands at an
>interesting crossroads: The drafting lines that developed in the wake
>of the Gulf War have fractured. Our social capital in the global
>arena, which carried us through much of the last decade, may be on the
>wane. Last year, when President Bush invited Tony Stewart, the 2002
>NASCAR Winston Cup champion, to the White House, some cynics suggested
>he was pandering for Southern votes. But maybe the president was just
>looking for a few tips.