One reason some people don't like statistics -- at least those fancy ones -- is they sometimes produce results incongruous with popular belief.
For example, this:
San Francisco Giants pitching staff 2012: 8.1 WAR (Baseball-Reference)
Kansas City Royals pitching staff 2012: 12.7 WAR (Baseball-Reference)
Giants pitching staff 2012: 13.3 WAR (FanGraphs)
Royals pitching staff 2012: 14.8 WAR (FanGraphs)
Giants pitching staff 2012: 56 runs above replacement (Baseball Prospectus)
Royals pitching staff 2012: 104 runs above replacement (Baseball Prospectus)
The three sites may disagree on the particulars, but they all agree on this: In 2012, the Kansas City Royals had a better pitching staff than the San Francisco Giants. The World Series champion Giants.
The above numbers do include bullpen results, and the Royals did have an outstanding bullpen in 2012. But even in pulling out just the starters, FanGraphs rated the Giants' starters at 11.7 WAR and the Royals' starters -- Bruce Chen and Luke Hochevar and Will Smith and Jonathan Sanchez and all the rest -- at 8.0 WAR. Not that much of a difference there.
The Royals had so much confidence in that staff that they traded for three new starters in James Shields, Ervin Santana and Wade Davis. The Giants were so discouraged by their results that they brought back the entire rotation.
So what gives? Is this a case of sabermetrics run amok? I don't believe so.
A player's home park can, of course, have a huge impact on his statistics. In the case of AT&T Park, the run-scoring environment is so low that the advanced metrics penalize Giants pitchers for their home-park advantage. In 2012, the Giants and their opponents scored only 580 runs at AT&T Park, but combined to score 787 runs on the road. In the case of the Giants' pitchers, they allowed 3.3 runs per nine innings at home, but 4.8 on the road.
The Royals, meanwhile, play in a more neutral park, but their pitchers allowed 4.7 runs per nine innings on the road (and remember, that's in the American League, where scoring was slightly higher overall -- 4.4 to 4.2 runs per game). When you factor in the advantage of pitching at AT&T, it decreases the value of the Giants' pitchers, and thus their WAR totals aren't as impressive as their ERAs may indicate. It's no different than it's easier to hit .300 at Coors Field.
Let's examine the Giants starters more closely. Below, you'll find their runs allowed per nine innings at home and the road over recent seasons.
2012: 2.1 home, 3.9 road
2011: 3.1 home, 3.6 road
2010: 2.9 home, 3.8 road
2012: 2.5 home, 5.0 road
2011: 3.4 home, 3.8 road
2010: 4.8 home, 2.2 road
2012: 4.3 home, 6.6 road
2011: 3.2 home, 2.9 road
2010: 3.9 home, 3.3 road
2012: 3.1 home, 4.2 road
2011: 2.2 home, 4.5 road
2012: 4.2 home, 4.8 road
2010: 3.8 home, 5.1 road
Last year produced more extreme results, but other than Lincecum in 2010 and 2011, the Giants' pitchers generally receive a nice boost at home. And while the Giants do have to play road games in Colorado and Arizona, that's balanced by the more pitcher-friendly parks in San Diego and Los Angeles. It's also worth noting that in broad terms all pitchers (and hitters) have a home-field advantage.
Does this mean that the rotation that has led the Giants to two World Series titles is overrated?
I think it's important to separate the regular season from postseason here. There's no denying the excellence of the Giants' starters in their 2010 and 2012 playoff runs. In 31 postseason games, they've allowed only 59 runs in 187 innings and posted a 2.70 ERA. Reputations are often created -- or sealed -- with October performance and there's no denying what Cain, Bumgarner, Lincecum and Vogelsong have done in the playoffs. Even Zito, after being left out of the rotation in 2010, stepped up last year, beating the Cardinals in the NLCS and then Justin Verlander in Game 1 of the World Series.
The tougher question to answer: What would these guys do on a different team, in a different park? Cain, Bumgarner and Lincecum have pitched only for the Giants. While the numbers suggest they would be worse, that may not necessarily be the case. Maybe pitching in an extreme environment -- like hitting in Coors Field -- produces extreme results. Just as Rockies hitters have trouble adjusting to life on the road, maybe Giants pitchers have simply adapted in unique ways to their home park and if they pitched in a more neutral environment, they would adapt differently and their home/road splits would even out.
In general terms, however, I would say the rotation is a little bit overrated (and the Giants' offense, which scored the second-most runs on the road in the majors last year, underrated), although you can't underestimate the value of their durability through the years.
That said, I'm pretty sure the Giants don't win the World Series with Hochevar starting instead of Cain.