Skip to main content

Phillies' much-hyped rotation even better than expected

The closest that has ever come to happening was in 1998, when the Braves' Tom Glavine, Greg Maddux and John Smoltz each finished in the top four places, but among five pitchers, with Maddux and Smoltz tied for fourth behind Glavine and Padres' Trevor Hoffman and Kevin Brown. The 2005 Astros had three of the top five, with Roger Clemens finishing third, Roy Oswalt fourth and Andy Pettitte tied for fifth.

When Lee signed with the Phillies last December, I was tasked with projecting where the Phillies' rotation would rank in baseball history. Limiting my view to the top four men in the Phillies rotation, I concluded that the Phillies' top four could well be the best quartet of rotation-mates since those 2005 Astros, but, including multiple iterations of rotations such as the Braves' in the 1990s or the Dodgers' in the 1960s, it didn't project to crack the top 70 from the last 56 years. With the season nearly over and the Phillies having performed so well, now seems like a good time to revisit that projection and attempt to find that Philadelphia quartet's proper place in baseball history.

The statistic that was the primary focus of my December piece was Baseball Prospectus's win-expectancy-based SNLVAR. Unfortunately, BP has discontinued that stat, believing that it was redundant given that both it and their primary total-value statistic, WARP (Wins Above Replacement Player) were both measured in wins against a replacement-level baseline. That requires a bit of a adjusting of my initial projections, but not much given the straight-forward manner in which I assembled them. For reasons that I elaborated on in the original piece, I used the following methods to project Phillies top four starters:

Halladay: Average of 2008 to 2010 seasons

Lee: Average of 2009 and 2010 seasons (leaving out what looked like a career year in '08)

Hamels: Average of 2008 and 2010 seasons (leaving out his inferior and somewhat unlucky 2009 campaign)

Oswalt: Average of 2008 to 2010 seasons, then slight reduction in wins above replacement to compensate for his luck on balls in play in 2010 and the degenerative disk in his lower back

That gave me the following (now with Baseball-References WAR, also wins above replacement, in place of SNLVAR):

With no more than two starts left in the season for each pitcher, this is how they have performed thus far:

My projections totaled 19.85 wins above replacement. Thus far the actual total has been 20.6 wins above replacement. Prorating that latter figure over the remainder of the season, which should yield at most five more starts for that quartet, we get 21.4 wins above replacement.

The quick verdict on my projections is that Lee has had a much better season than anticipated, perhaps because I failed to adjust sufficiently for his move to the weaker league because he didn't show much effect during his prior stint in the NL in 2009 (though, in retrospect, those dozen starts down the stretch may have been too small a sample for me to write off that effect as completely as I did). Oswalt, meanwhile has had a much worse season than expected but for an entirely expected reason: that degenerative disk in his lower back cost him about 10 starts via two stints on the disabled list. Hamels has been better than expected, but not by a lot, and I'm going to say that I nailed Halladay, even if the numbers aren't perfect matches. As a result, Lee (roughly two wins better than expected) and Oswalt (roughly two wins worse) largely cancel each other out, Hamels rounds things up slightly, and my total projection for the four captured 92 percent of their actual value. Not bad.

So how does that stack up against history's greatest starting quartets? Using SNLVAR, I identified the 1966 Dodgers and 1997 Braves as having had the best top four starters since 1954 (which is as far back as that stat went). Here are those two rotations again with their bWAR totals:

At 20.5 wins now and a pro-rated 21.4, Halladay, Lee, Hamels and Oswalt are right there with those two quartets. However, per bWAR, which covers all of major league history, those were not only not the two best rotation quartets in history, but weren't even the two best of those last 56 years. Here's my best attempt at a bWAR leader list from the modern era (1901 to present):

That list is a reminder of how long and rich the game's history has been. (In case you're wondering, the 1971 Orioles staff that is the only one in history to feature four 20-game winners -- Jim Palmer, Dave McNally, Mike Cuellar and Pat Dobson -- compiled a bWAR of just 13.0) Some of those quartets are more top-heavy than others. For example, Bob Gibson's 1968 season and Dwight Gooden's 1985 were responsible for more than half of the wins above replacement compiled by the top four men in the '68 Cardinals' and '85 Mets' rotations. There are also four rotations that effectively occur twice (the Gibson/Steve Carlton Cardinals, the Fergie Jenkins/Bill Hands/Ken Holtzman Cubs, the Christy Mathewson/Joe McGinnity Giants, and the Glavine/Maddux/Smoltz Braves). Also, the 1969 Cardinals and 1974 Mets didn't have a proper fourth starter and their bWAR totals thus include the combined wins above replacement in starts and relief of their top swing man.

One could thus argue that the '69 Cardinals, who received 18.2 wins above replacement from Gibson and Carlton alone, don't belong on this list. That would allow the Phillies' pro-rated total of 21.4 wins above replacement to slip past every version of the Glavine/Maddux/Smoltz Braves into the top 10 in the modern era (since 1901), the top eight in the live-ball era (since 1920), and the top seven since integration (1947), and you can shave an extra place off of each of those by treating the 1969 and 1970 Cubs as a single team.

That's a much more impressive showing than my SNLVAR-based projection from December, and makes the 2011 Phillies rotation legitimately one of the greatest baseball history.