October 5, 2009
Tossing out the Sunday night opener, MLB teams play 162 games over 182 days, with no scheduled doubleheaders and a three-day break in the middle of the season. That means roughly 19 games every 21 days, with some 21-day stretches that include 20 games. There aren't a whole lot of breaks, and as the schedule has evolved over time, with planes replacing trains, and doubleheaders going the way of pullover jerseys and Ladies' Days, player usage has evolved. The five-man rotation is a rational approach to the elongation of the schedule, and while there's a case to be made for the four-man rotation, this adaptation seems to be permanent.
The postseason circa 2009 is a different animal from that long slog. If, say, the Red Sox were to go the distance in their Division Series matchup, then do the same in the ALCS, they would play 12 games in a 23-day span from today through October 27. Advancing while playing fewer games is, of course, possible, which would make it an even less trying schedule. At no point in there would they play on more than two consecutive days.
Since the 1970s, MLB teams have generally adapted to the postseason by reducing their starting rotations to four men, and on occasion using three. When you look at the modern schedule, though, there is very little reason to concern yourself with a fourth starter until the World Series. As it stands now, a team can maximize the use of its top two starters in most scenarios while rarely asking them to start on fewer than four days' rest. Given the limited role of a fourth starter-occasionally serving as a long man in the first two games of any series or very last game of a best-of-seven in addition to the one start-it's not clear that the roster spot is better used on him rather than on a player who can be deployed in other ways, in more games, in high-leverage spots.
Take the Red Sox as an example. They will most likely open play on Thursday in Anaheim. They can roster Daisuke Matsuzaka with an eye towards having him start the fourth game of the series in Boston on Monday, and have him available in long relief-the need for which is highly unlikely-in the first two games. Or, they can plan to bring back Jon Lester on short rest, one time, and use Matsuzaka's roster spot on a tactical reliever or, better still, a pinch-runner for one of their many slow hitters. If there is a Game Five, Josh Beckett would be pitching on full rest Wednesday. For a team playing the expanded Division Series, the one scheduled over eight days, the choice is even more clear-the first and second starters can pitch games four and five, if necessary, on full rest. Most teams have made this adaptation, as we'll see with the Yankees this year.
Extend this discussion to the ALCS, and you see that the need for a fourth starter becomes even less apparent. The Game One starter would have to come back and throw Game Four on short rest. That's two short-rest starts in a span of about two weeks, which is non-standard usage, but far from abusive. If you win the Division Series in four games, you can have the ALDS Game Two starter start ALCS Game One and take on the short-rest role, spreading the burden between two pitchers. A Game Four LCS starter will get four days of rest before a Game Seven due to the insertion of a non-travel off day between Games Four and Five. This is why there is only one possibility of a short-rest start in the LCS.
Over the first two series, there would be a maximum of two short-rest starts, and they would be given to a team's number-one starter twice, or a number-one starter and a number two. The gain would be in getting more starts from your very best pitchers as opposed to your fourth guy, as well as the roster value in not carrying a pitcher whose job is to start once, and who is expected to go six innings at most. That roster spot can be put to better use.
If you reach the World Series, a three-man rotation becomes a bit more problematic, as all three of your top starters would be needed on short rest, due to the merciful absence of the non-travel offday. At that point, however, you can re-roster that extra starting pitcher to allow your top three full rest if you so choose. Or, you can go to a three-man rotation knowing that your pitchers get four months' rest once the series is over. Flags fly forever, and the pursuit of a championship over a time period measured in days is a time when you can ask more of your players, particularly veteran players. Protecting the arms of pitchers aged 24 and under has been and should be an industry goal, but over-worrying the workloads of those who remember Cosby sweaters is a misapplication of principles. Doing so in October is just silly.
Now, back to the original issue, and let's squeeze it a bit and look at the NL, which this year is the "early" series. An NL team going the distance would play its first 12 games in 20 days, from Wednesday through scheduled NLCS Game Seven on October 24. At no point will a National League team play on more than two straight days, which means that no reliever will ever be asked to work more than two days in a row, or four days in five.
So what the hell is Tony La Russa thinking? Per Rick Hummel of the St. Louis Post-Dispatch, the Cardinals are leaning towards carrying 12 pitchers for the first round of the postseason. The Cardinals' championship hopes, mind you, are pretty much entirely reliant on their top three starters, two of whom will finish in the top three in the Cy Young voting, being effective and working deep into games. But come Wednesday, La Russa plans to have nine guys ready to pitch behind them, including potential Game Four starters John Smoltz and Kyle Lohse. That's nine relievers in the bullpen, which would mean having just five position players on the bench.
This is patently insane. Having 12 pitchers on the roster is overkill in the regular season, when you're playing every night and carrying five starters. Most teams end up not getting their back-end relievers enough work, or worse, changing pitchers' roles on a week-to-week basis as a reaction to the last three outings. In a postseason, however, when you'll never have to use guys for more than two straight days, this is madness. So you have a closer, in Ryan Franklin. You have four set-up or situational matchup guys in righties Kyle McClellan and Jason Motte and lefties Trever Miller and Dennys Reyes. You have the two Game Four options as long men. That's 10 pitchers, none of whom could possibly end up worn out by playing five games in seven days. That's 10 pitchers backing up three starters who averaged just shy of seven innings per start.
What in god's name is Brad Thompson adding to that mix? When are you ever going to bring Mitchell Boggs into a game? And Todd Wellemeyer? And Blake Hawksworth? If one starter gets hammered early, well, you have two starters in the pen as long men. If two starters get hammered early, you're not advancing no matter what, anyway. You can't justify this many pitchers based on playing National League baseball, because you don't have enough hitters on the roster with which to pinch-hit or double-switch anyway. Meanwhile, you won't be able to hit for the many crappy OBP guys on your roster because you were too concerned about what would happen to your bullpen in a second straight 17th-inning game.
Maybe La Russa will rethink this. Maybe Hummel has it wrong, although I don't think that he does, based on other places I've seen this notion. Carrying a dozen pitchers to play five games in seven days, or seven games in 10 days-when all the games are high-leverage and you never play more than two days in a row-is the kind of roster management that should get you fired on the spot. There is no possible use for that many pitchers, even granting La Russa's matchup-centric approach to the late innings. Even a team with lesser starting pitching would be hard-pressed to justify carrying that many relievers. No team has a dozen pitchers good enough to be used in post-season games. Some don't have a half-dozen, but we'll let one in tomorrow anyway.
There are real costs to this kind of elective post-season roster construction. Teams have given away the ability to counter bullpen moves by having quality pinch-hitters on their bench. They have lost the opportunity to take advantage of high-leverage situations on offense because they have to let glove men bat. Managers seem to have forgotten that pinch-hitters can be used for position players as well as pitchers; it's as if they've calculated that the potential offense taken by the player means more than the potential offense taken away by the player. In post-season games often decided by one run, one plate appearance, one base, that's malpractice. In building a post-season roster, teams have to stop worrying about what will happen in one highly unlikely situation, some series of injuries and/or extra-inning contests that leaves them short an arm, and focus on what will happen in almost every game: they'll have a bad hitter due up in a situation where a run will be needed. Win that battle, win it because you did a better job constructing your roster than the other guy did, and you might just win the war.
I'm late getting to this, but I want to make mention of it while we're still a month away. As many of you know, I go to the Arizona Fall League every November as part of Baseball HQ's First Pitch Arizona program. Ron Shandler has been inviting me for eight years now, and it's a pleasure and privilege to attend. The program consists of panel discussions and seminars with an all-star team of baseball minds, tickets to a number of AFL games, and the opportunity to spend four days with people who love baseball.
This year, the schedule provides a bonus in that the weekend, November 5-8, coincides with the AFL All-Star Game, which will no doubt be the greatest concentration of prospects this side of the Futures Game.
As an added bonus, Ron has generously offered a discount to people who sign up through Baseball Prospectus, making what is already a weekend worth the money that much more so. To get more information or to sign up, follow this link to the Baseball HQ site and read all about it. If you have any questions for me, you can click the link at the bottom of this page to ask me directly. Thanks, and I hope to see you out there next month.