As we embark upon another postseason, there will be a rush to predict. Everyone has a different take as to what wins in October, what will make one team stand out from another, why the teams they pick are sure to go deep into the playoffs.
The truth is, none of us really knows. A big part of the fun of postseason baseball is the unpredictability. Five Wild Card teams have won the World Series. In 14 of the past 17 years, the team with the best record has not won the World Series. We're all trying, but none of us have figured it out.
To name one example, the smart folks at Baseball Prospectus introduced a formula a few years back called "secret sauce," which seemed to provide at least a blueprint for how to win in October. But four years later, the site retired the formula after a series of results that seemed to indicate it was less effective than originally believed. That's an indication of what we're up against: even some of the best number-crunchers don't have a model in which they have much confidence.
It all goes to show that baseball's a funny game, and just about anything can happen in five or seven games. So rather than pretend that there's any really reliable method for predicting a champion, let's take a look at a few things that recent World Series champions had in common.
The research only goes back to 1995, the first year of the three-tiered postseason, because it's a different challenge winning three series than two or one. It's reverse engineering, rather than anything scientific, but it's fun to take a look.
The front of the rotation: This is old baseball wisdom, and the recent numbers bear it out. Rotation depth is nice, but championships are won with top-tier starters.
The past 17 World Series winners had an average of 2.18 pitchers who ranked in the top 20 in wins, 1.65 who ranked in the top 20 in ERA, and 1.65 who ranked in the top 20 in strikeouts. Now, there's no denying that wins just aren't a good measure of a pitcher's effectiveness -- after all, it makes sense that good teams would have pitchers who won a lot of games.
But as part of a bigger picture, it's clear that the teams that won in October had pitchers at the front of their rotations who were able to provide high-quality innings. It's hard to match up with an opponent in a series if you give away the pitching advantage in the first two games.
Catch the ball: Defense matters. A lot. It matters in the regular season, and it matters in the postseason. As the margins get narrower in October, it becomes more essential not to give away outs and bases, and to make sure that when pitchers make good pitches, the batted balls turn into outs.
This is evident in the numbers. The average league rank of the past 17 World Series winners in Park Adjusted Defensive Efficiency, a number calculated by the folks at Baseball Prospectus, is 4.47. Three teams have ranked lower than seventh, while four have finished first.
Get on base: A team can score a lot of runs by hitting for a high average, but that kind of attack is susceptible to running into a really good defense. Hitting for power is great, and in fact the "short-sequence" offense is a very good way to go, as last year's Cardinals demonstrated. But even more reliable in recent years than power or batting average has been on-base percentage. As the saying goes, "OBP is life, and life is OBP."
The past 17 World Series winners finished with an average rank of 4.47 in OBP in their respective leagues. Four teams that led their league in OBP won titles, plus three runners-up. No team finished worse than 11th, and only four were lower than sixth. OBP reflects the ability to hit for average and to control the strike zone, and if you can do both, you're a lot better off than just doing one or the other.
Shut the door: There are all sorts of sophisticated ways to measure relief work, but let's look at a simple one. Teams that convert their chances to win games in the regular season tend to do well in October. Those 17 World Series winners ranked, on average, 4.41 in their leagues in save percentage in the regular season. Three teams ranked first, eight ranked in the top three, and only one team ranked worse than eighth.
And just for fun, here's one trait that's not as important as you might expect ...
Team walk rate: No team that led its league in walks allowed per nine innings has won the World Series in the Wild Card era. Only four teams that have finished third or better have won it all in that time. Teams ranking 14th, 11th, seventh, 14th, and eighth are among the champions in the past 17 years. Recent champions have been much better at racking up strikeouts and preventing home runs than at avoiding walks.
A much better indicator is strikeout-to-walk rate. The last 17 Series champs have finished with an average rank of 6.24 in walk rate. In K/BB ratio, the average is 4.47, including three teams that led their league, three more that finished second or third, and only two that finished worse than seventh (ninth and 10th for the 2006 Cardinals and 1997 Marlins, respectively).
Matthew Leach is an editor and reporter for MLB.com. Read his blog, Obviously, You're Not a Golfer and follow him on Twitter at @MatthewHLeach.