“If something seems too good to be true, it probably is,” cautioned National Post baseball writer John Lott in November 2012, as fans whipped themselves into a frenzy over the Toronto Blue Jays. With everyone in baseball talking about the team—which had just completed a massive deal with the Miami Marlins for two starting pitchers and an all-star shortstop—Lott wasn’t prophesying defeat; he was simply applying some sober second thought. “In this business,” he says, “you have to try to maintain some kind of equilibrium and not get swept up.”
Lott may have been the only one not predicting how the Jays would do. The team was, after all, the story of last year’s off-season. It garnered unprecedented coverage during spring training and beat reporters foresaw great things. The Globe and Mail’s baseball writers—Jeff Blair, Tom Maloney and Robert MacLeod—each predicted the Jays would finish first or second in their division. Richard Griffin of the Toronto Star, who has covered the team since 1995, predicted it would make the World Series, but lose to Atlanta.
Of course, Toronto did not win the World Series. Not even close. The Jays finished last in their division, with a record only marginally better than the year before. Local baseball writers admit that pre-season predictions are just a bit of fun—not the kind of thing anyone hangs his hat on. While advanced statistics have the potential to make forecasts more reliable, this past Jays season showed once again that there are limits to using the past to predict the future.
Predictions have long been a staple of sports coverage. On the eve of the 1908 season, a New York Times baseball correspondent wrote, “With a fair share of luck coming their way, the Yankees have as good a chance of winning the pennant as any other club.” They finished last in the American League, but that didn’t matter, just as last year’s faulty predictions haven’t hurt anyone’s reputation. “That’s part of the beauty of it, how wrong we are,” says Shi Davidi, who covers the Blue Jays for Sportsnet and co-wrote a book with Lott about the 2013 season. “If the outcomes were so predictable, who cares? Why would you watch?” Sports forecasts are good fodder for talk radio and social media—“a talker, more than anything,” says Maloney, who is now the editor of the Globe’s automotive section.
It’s not just fun, though; baseball writers say their guesses are more than gut feelings. “It’s statistical analysis,” says Maloney, “but it’s also knowing the players.” With the 2013 Blue Jays, the thinking was, “On paper, if everyone has an average year, a couple of guys have great years and—big, big, big asterisk—everyone stays healthy, then they ought to have been able to compete.” That asterisk is a constant wrinkle in forecasting.
Lott, who doesn’t make predictions and doesn’t consider himself an expert, says those who do should start with the pitching rotation and the understanding that “one or two of these guys you know is gonna get hurt for two to five weeks in the season, maybe longer.” Injuries turned out to be one of the biggest factors in the disappointing showing, with many Jays players—including ones acquired in the off-season spending spree—sidelined for weeks at a time.
The severity of injuries is nearly impossible to predict, but the statistics era is starting to reduce the number of unknowns in forecasting. New ways of measuring performance have made it easier to crunch the numbers and produce objective predictions for individual players. These somewhat obscure statistics include NASA-like acronyms like VORP (how many runs a hitter gets, compared to a low-cost substitute) and PECOTA (a complicated calculation that projects overall performance). One Baseball Prospectus writer has even tried to quantify the risk of injury.
Derek Carty, a fantasy baseball analyst who has worked for stats websites like The Hardball Times and Baseball Prospectus, thinks these advanced statistics can help old-school commentators look beyond batting averages or earned runs. There are just too many things that affect a player’s performance, and “the human mind can’t comprehend all those factors,” he says. The stats revolution is most prominent in baseball, but other professional sports are catching up, with advanced metrics such as Corsi (shot attempts) and PDO (save percentage plus shooting percentage) cropping up more in NHL coverage. One possible reason for the rise in popularity of these new measures is the fact that, as Gabriel Desjardins explains on his hockey statistics website Behind the Net, more traditional criteria “don’t tell us very much about a player’s true value.”
But Carty admits there are limits to what the numbers can tell a scout, bookie or writer. “The stats tell you the ‘what’ and the scouting can tell you the ‘why,’” he says. Data points don’t explain, for example, why a player has a banner year and whether he’ll have another, “whereas the scouting maybe can.” Scouting also has the advantage of being more current. As the Blue Jays showed last year, stats from past performance are not necessarily predictive: the team that hit the turf at the Rogers Centre did not play like the roster assembled on paper.
This is part of why sportswriters have travelled to the team’s training facility in Dunedin, Florida: to see how the Blue Jays look and ask them how they feel about the season. “Baseball beat writers spend anywhere from four to six weeks at spring training, and by the end of that time, you’ve had a great many conversations with managers and general managers and players,” Maloney says. So when it comes time to make predictions, “It’s not just looking on paper.”
The 2014 Blue Jays look a lot like last year’s team, with a few exceptions: small acquisitions and rumours that a couple of minor-league pitchers may be on their way up. But the hopeful predictions that crept into last year’s coverage, when the words “World Series” were nearly synonymous with “Toronto,” have been absent so far. With opening day just over a month away, the focus has been more on what the team lacks than on what it might be—and maybe that’s not a bad thing.