Postseason predictions are a staple of sports pages. Everyone loves them, it seems. Sportswriters like them because, after all, they are in the business of giving their analysis of upcoming events. The sports media like them because the fans devour them. Fans like predictions because they provide unmissable reading pleasure.

Think about it: have you ever met a sports fan who read a postseason prediction in a newspaper and lost hope in their team’s luck? If the writer chooses the fan’s team, he gets validation of his hopes. If the writer doesn’t pick the fan’s team, then he creates an “us against the world” feeling. There is nothing that makes a fan feel like a part of the team more than feeling that “everyone” lacks respect for their team.

In a sense, therefore, postseason predictions don’t have to be right or wrong. That being said, how often do sports experts get these predictions right? Baseball’s recent Division Series offers a good opportunity to examine this. the websites of sports news and Sports Illustrated each provided the predictions of eleven of their baseball writers for all four American League and National League Division Series. With 22 writers giving four predictions each, there is enough sample available to form some conclusions about the accuracy of the experts’ postseason predictions.

Of these 88 predictions, 41 times a writer predicted the series winner. This is a 46.59% success rate, or slightly worse than what you would get if you flipped a coin for each set. The pundits did better when predicting the Red Sox-Angels series, where 15 writers (68%) picked the winner. The Yankees-Indians series was predicted correctly by only 10 (45%), and the Diamondbacks and Rockies were underestimated, only 8 writers (36%) opted for each. Take away the relative success the writers had in predicting a Red Sox win, and the pundits picked the right team only 40% of the time.

If we consider predicting not just the series winner, but the number of games, the results are even worse. Almost every writer tried to predict how many games each series would last, eg “Yankees in 4”. Only one writer, Sean Devaney of sports news, correctly predicted this outcome for any series. This may be because these division series were unusually short, with 3 sweeps and a four game series. Writers are probably hesitant to predict sweeps, and in fact, only five writers predicted a sweep in any series. Unfortunately for them, they chose the exact opposite of the actual results. Once again, a random selection would have predicted the correct result more often than the experts.

So what does this tell us? Not putting much faith in postseason predictions is one answer. That’s not much of a revelation, of course, but somehow you’re beginning to see just how wrong expert analysis can be. Sportswriters and analysts are also readers and viewers, and a bit of groupthink may set in. Everyone looked at the Yankees’ roster and late-season revival, and only 10 were willing to defy the consensus. Even fewer gave the Rockies and Diamondbacks their comeuppance, again overthinking the ways the Cubs and Phillies were bound to win.

Many sports fans believe that the sports media is biased towards the East Coast, and these predictions may confirm that as well. The only series in which the number of correct predictions exceeded chance was the Red Sox-Angels series, won by a team from the East Coast. In all other series, most writers chose the easternmost team, with the westernmost team winning.

Regardless of the reasons, and there are probably many, including ones not mentioned here, the lesson is clear: Sports experts have no better idea who will win in the postseason than the rest of us.