Scorecasting: The Hidden Influences Behind How Sports Are Played and Games Are Won

by L. Jon Wertheim

  Of course, punters and kickers are just two players, and you could question whether either has the ball long enough to be affected by a rabid crowd. Fair enough. Unfortunately, there isn’t another isolated activity within the game of football we could point to in order to measure crowd influence outside everything else going on in the game. Instead, we could look at a number of offensive and defensive statistics to see where home teams fare better than visitors, recognizing that these advantages could come from a host of sources, including the crowd. It turns out that NFL teams rush better at home. They rush more—no surprise there since home teams are ahead more often—but also gain more yards per rushing attempt. Visitors, by contrast, pass more than the home team because they are usually behind and need to make up points in a hurry. But interestingly enough, visiting teams pass slightly better than home teams. (Who knew? NFL quarterbacks are a little better on the road than at home.) Though we might speculate that extreme crowd noise distracts visiting quarterbacks and makes their commands inaudible to their teammates, it doesn’t seem to affect their performance. Thus, at least with respect to this aspect of the game, it’s hard to say crowd noise contributes to the home team’s success.

  In baseball, the closest we can come to measuring the crowd’s influence is to examine the pitcher. Not his ball-strike count—influenced, as it is, by the batter, the umpire, and the game situation—but his velocity, movement, and placement. We got the data thanks to the MLB.com technology Pitch f/x. A computer generates the location of the pitch, the height of the ball when released from the pitcher’s hand, the speed at which the ball travels when it leaves the pitcher’s hand as well as when it crosses the plate, and the degree to which the ball’s direction changed or diverged from its path to the plate, both horizontally and vertically. Baseball researchers and Sabermetricians have been busily gathering and applying the data to answer all sorts of intriguing questions: Who has the nastiest sinker? How is Red Sox ace Josh Beckett’s fastball different from his breaking ball? (For the record, his changeup has as much movement as his fastball, a big factor in his success.)

  We obtained the last three years of these data, covering more than 2 million pitches, to answer a different question: Do pitchers actually pitch differently at home versus away? Before the Pitch f/x data existed, you couldn’t really answer this question. You could only ask if the outcomes—balls or strikes—from pitchers were any different at home versus away. But again, outcomes are dependent on pitch selection, hitter reaction, umpire response, and game situation. Now, with the Pitch f/x data, we can simply ask: When Josh Beckett throws a fastball at Fenway, does it have more velocity, more movement, and better placement than it does when he is on the road?

  Much like NBA players shooting free throws, on average Major League pitchers are equally accurate at home as on the road, throwing a ball within the strike zone 44.3 percent of the time at home and 44.5 percent of the time on the road. Pitches more than 1.5 inches outside the strike zone occur just as frequently at home as on the road (46.5 percent of the time). Even the most extreme pitches, those way out of the strike zone or those that hit the dirt, occur no more frequently on the road than at home.* In addition to having identical accuracy at home and on the road, pitchers throw with the same velocity—87 mph on average when the ball crosses the plate—and movement no matter where they play. We repeated these numbers for the same kind of pitch, categorized by Pitch f/x into changeup, fastball, curveball, four-seam fastball, split-fingered fastball, cut fastball, sinker, slider, and knuckleball, and found no difference in speed, movement, or placement for the same type of pitch by the same pitcher at home versus on the road. We tested the first inning versus later innings. Again, there was no difference. We tested different pitch counts. No difference. We even tested different game situations. Again, no difference. Pitchers appear to pitch no differently along any dimensions we can measure at home versus on the road, suggesting that neither the crowd nor the optics of the stadium influences their performance.

  We can also use the Pitch f/x data to help gauge whether playing at home has any impact on batters. Controlling for all the factors that might make a batter hit better at home is nearly impossible, but we can control for a lot by looking at identical pitches—same speed, location, movement—in identical situations, home versus away. Does David Wright, the New York Mets’ star third baseman, hit a waist-high, middle-of-the-plate, 90-mph fastball on a two-two count better at home, where the crowd is supporting him, than on the road, where he might be getting booed?

  Location may be the key to real estate, but when it comes to hitting a baseball, for the average Major Leaguer geography doesn’t matter. When a player swings at a pitch in the strike zone, his probability of hitting the ball is exactly the same at home and away. For pitches outside the strike zone, batters also fare equally at home and on the road. They are no more likely to swing at the same pitch at home than on the road and demonstrate no greater ability to swing at better pitches at home.

  Of course, it could be the case that the crowd influences other aspects of the game—pumping up home teams to play better defense, encouraging them to greater effort. We can’t rule this out, nor can we measure or isolate the crowd effect on these aspects of the game from other things going on at the same time. What we can observe is that in situations in which many of these other influences are “turned off” or controlled for, the crowd seems to have no effect. If the crowd is ineffectual in these isolated situations, it is at least questionable how much of an effect it could have in other situations.

  Hey-batter-batter-batter-swing? Sorry, but he’s going to do it equally well whether you’re chattering or not. Just as he’s going to shoot free throws, kick field goals, and deke the goalie comparably well whether you’re encouraging him or cursing him.

  Conventional Wisdom #2: Teams win at home because the rigors of travel doom the visitors.

  By the end of the first quarter, the Spurs led comfortably, 29–20. The Blazers were shooting poorly, were committing scads of mental errors, and were conspicuously sluggish on defense. They were late to rotate, made only halfhearted efforts to block shots, and generally treated the lane as if it were a zebra-striped crosswalk: No, really, after you. Go right ahead! Portland’s coach, Nate McMillan, would later observe that his players were “a step slow.”

  Who could blame them? The Blazers had played in Houston the previous night, boarded a plane, landed in San Antonio after midnight, and then taken a bus from a private airstrip to the hotel. Some players reported that they didn’t fall asleep until dawn. As anyone who’s taken a red-eye flight can attest, the grind of travel—the sitting, the dislocation, the time changes—can exact a steep price on the body. All the more so when the itinerary mirrors that of an NBA team pinballing randomly to Indianapolis one night, to Phoenix the next, and over to Dallas a few nights later. Your circadian rhythms are thrown off; your immune system can betray you. At the hotel, everything from the lumpy pillows to the inevitably ill-timed knock from the minibar stocker to the inadvertent wake-up call militates against a good night’s sleep. It makes sense that the athletes who slept on their favorite pillows and woke up in their own beds that morning will outperform the ones who flew in earlier that morning. Particularly so for a team such as Portland, which—thanks to the Seattle Sonics relocating to Oklahoma City—is a significant flight away from every other team in the league.

  We submit, however, that the travel doesn’t much matter. The rigors of the road exist, but they don’t underpin the home court advantage. Why do we say this? Consider what happens when teams from the same (or a nearby) city play each other, when the Los Angeles Lakers play the Los Angeles Clippers in the NBA—the two teams share the same arena—or the New York Rangers play against the New York Islanders or the New Jersey Devils in the NHL. For these games, the “rigors of travel” are nonexistent. Everyone is in his natural time zone and sleeping in his own bed. Yet if you look at all these “same city” games, you find that home te
ams have the exact same advantage they do in all the other games they host. Likewise, road teams don’t lose more often when they travel greater distances. Controlling for the quality of the opponent, the San Antonio Spurs, for example, fare no better when they take puddle-jumpers to play the Dallas Mavericks and Houston Rockets than when they make longer trips to Boston, Toronto, and Miami.

  We can take this one step further in the NHL, looking at games that involve not only long travel distances but also border crossing—which can require negotiating customs and other procedures that generally increase the pain of transit—by examining U.S. teams that play in Canada and vice versa. Yet we find no abnormal home ice advantage for U.S. teams visiting Canadian teams or vice versa, even for those farthest from home.

  In Major League Baseball the rigors of travel aren’t a significant issue, either. Just as in the NBA and NHL, for games involving teams from the same metro area—interleague play between the Chicago Cubs and White Sox, New York Yankees and Mets, Los Angeles Dodgers and Angels, San Francisco Giants and Oakland A’s—the home teams win at exactly the same rate at which they normally do. We also know that home field advantage has been remarkably constant over the last century; it was virtually the same in MLB from 1903 to 1909 as it was from 2003 to 2009. This suggests that the teams jetting on chartered flights with catered meals, high-thread-count linens, and flat-screen televisions have no more success than did the teams that traveled to games in Pullmans and buses. (Either travel isn’t causing the home advantage or teams need to rethink their jet purchases and the on-flight catering.)

  Nor do the rigors of travel play much of a role in the NFL’s home field advantage. Teams play only one game per week and in fact usually depart for a game a few days in advance to acclimate. As with the other sports, when nearby teams play—Oakland Raiders versus San Francisco 49ers, New York Giants versus New York Jets, Baltimore Ravens versus Washington Redskins—the home field advantage holds firm at its normal level.

  Finally, we noticed that home field advantage in soccer is the same in countries such as the Netherlands, Costa Rica, and El Salvador, where travel distances are minuscule, as it is in countries as vast as the United States, Russia, Australia, and Brazil. It is yet another indication that travel isn’t much of a factor.

  Conventional Wisdom #3: Teams win at home because they benefit from a kinder, gentler schedule.

  By halftime, the Blazers had whittled the San Antonio lead to three points. If both teams appeared fatigued—lacking “fresh legs,” to use the basketball vernacular—it was with good reason. They had both played a game the previous night.

  Because of the physical demands of running up and down a court for 48 minutes, it’s exceedingly difficult to compete at full strength on consecutive nights. On the second night of back-to-back games, NBA teams win only 36 percent of the time. It was Charles Barkley who once referred to second games as “throwaways.” In his inimitably candid way, he once explained, “You show up because they pay you to show up. But deep in your belly, you know you ain’t gonna win.”

  Okay, we’ve discounted the effect of the “grueling” travel. But what about the fact that visiting teams play the vast majority of back-to-back games? Could that influence the home court advantage in the NBA? We think it does. And this particular Spurs-Blazers game notwithstanding, the vast majority of back-to-back games are played by road teams. Of the 20 or so back-to-back games NBA teams play each season, an average of 14 occur when they’re on the road. That alone affects the home court advantage in the NBA. By our calculations, you are expected to win only 36 percent of those 14 games relative to your normal chances of winning on the road when you aren’t playing back-to-back games. That translates into one or two additional games you will lose each season on the road because of this scheduling twist. In other words, home teams are essentially spotted an advantage of one or two games relative to road teams just from the NBA’s scheduling of consecutive games.

  It’s not just the back-to-back games. Home teams not only play fewer consecutive games but also play fewer games in general within the same time span, such as the last three days or the last week or even the last two weeks. All this takes its toll on visitors. We estimate that about 21 percent of the home court advantage in the NBA can be attributed to the league’s scheduling. Adjusting for this scheduling effect, the home court advantage drops down to 60 percent. So part of the explanation for the very high NBA home court advantage is the way the league is arranging the schedule.

  Recently, there were internal discussions—ultimately fruitless—in the NBA about reducing the season from 82 games to 75 games. The first games that would have been cut? The notorious back-to-back road games.

  The league will tell you that the bunching together of a team’s road games in as few days as possible is done to economize on travel. But by accident or design, it has the effect of working to the detriment of teams on the road. And it’s not the only evidence suggesting that the league prefers to see home teams win. For instance, in home openers, the majority of teams start the season against weak opponents who had either inferior or similarly poor records the previous season. Just look at the Sacramento Kings and Washington Wizards, the two worst teams in 2008–2009; they started the 2009–2010 season playing five road games between them.

  Says one NBA owner: “If only [fans] knew how the NBA scheduled games. Teams submit blocked dates for their arena [i.e., dates when the circus is using the building or the NHL team is using the facility]. The NBA picks ‘marquee TV match-ups,’ and then one guy figures out the rest with marginal help from software. Teams kiss his ass because we know he can throw more losses at us than Kobe can!”

  To test the role of economic incentives, the most valuable NBA franchises, according to numbers from Forbes, are afforded a slightly stronger scheduling bias, as are teams in big markets. Yes, all teams play more consecutive games on the road than at home, but it’s less so for the most valuable franchises in the biggest markets.

  Remember how in most seasons every NBA team, even the Clippers, fares better at home than on the road? One look at an NBA schedule, and it starts to make sense why that is the case. When teams leave the comforts of home, they get hammered by the schedule makers, playing as many as three games in four nights, seldom in any logical order.

  Seen through this lens, the action unfolding during the Portland–San Antonio game becomes clearer. The Blazers not only had played the night before at Houston (and lost) but had played four games in the week leading up to their game with the Spurs with only one day of rest between games. The Spurs meanwhile had played only three games the previous week and hadn’t traveled in the last four days. If you take our numbers on back-to-back road games and factor in that Portland had also played one more game than San Antonio in the last week, the Blazers’ chances of winning fall to less than one in three, which puts it in the vicinity of “ain’t no way we’re winning this motherf——” territory.

  What about other sports? As in the NBA, teams in the NHL are brutalized by back-to-back games, which also occur disproportionately on the road. As physically demanding as the NBA is, the NHL may be even more taxing. In a typical season, road teams will play six more consecutive games than home teams do, which translates into about one or two extra home victories per team per season.

  Scheduling is less of an issue in baseball; the 162-game schedule is set up so that teams play in three-, four-, and five-game series. When teams travel, they get to stay put in the visiting city, and the consecutive games have less of a physical effect on the athletes. The player who exerts himself the most—the pitcher—plays only once every five games.

  In the NFL, the one league that publicly and unapologetically strives for parity—“any given Sunday” your team could beat the other team—there is virtually no evidence of scheduling bias. Even in home openers, the most successful NFL teams are not favored. In fact, we find the opposite: NFL teams that did well the previous season are more likely to face a better opp
onent in their opening home game than they are to face a team that did poorly last season.

  By contrast, in college sports, scheduling plays a huge role in the home team advantage. College boosters would have you believe that the exceptionally high winning percentage in NCAA sports is a consequence of rabid school spirit, the pep bands and cheerleaders, and those exuberant undergrads annoying opponents with witty cheers and taunts. But most followers of college sports are likely to guess what’s really driving a large part of the high home field advantage. It’s the scheduling of weak opponents—cupcakes, patsies, sacrificial lambs, road kill, call them what you will—early in the season.

  Although the NCAA and the conferences set the schedule for most “in-season” games, the individual schools are generally free to negotiate their own preseason schedules. At large schools, there is an incentive to pad teams’ records early in the season. Stacking the scheduling deck in their favor, teams from the six “big” football conferences—the Big Ten, Pac-10, SEC, ACC, Big 12, and Big East—win almost 90 percent of their home openers. In addition to pleasing the crowd, especially those cotton-head donors sitting in the prime seats, early success bolsters the team’s chances of reaching the postseason bowls and tournaments, which come with a direct financial payoff and, generally, a spike in alumni contributions.

  At small schools, there are incentives to play along. One is to raise revenue. Often, playing at Big State U ensures a monetary reward far superior to what the team could have earned playing a smaller opponent. In 2006, for instance, the small-time football program at Florida Atlantic University was paid $500,000 to play at Clemson in the season opener. FAU then reportedly made an additional $1.325 million playing its next three games at Kansas State, Oklahoma State, and South Carolina. From those four road games alone, they covered a sizable chunk of the annual operating expenses for their entire athletic department. But they lost the four road games by a combined score of 193–20.