Wednesday, November 28, 2007

Surely Some Revelation Is At Hand

Tommy Glavine put the finishing touches* on the Mets historic collapse in the last game of 2007 with one of the most spectacularly bad starts of his long career, giving up seven earned runs in only a third of an inning. In a way, I can almost respect that. I mean, if you’re going to blow a seven-game lead with only 17 games left to play, you may as well finish it in the most magnificently horrible way possible.

But that begs the question, how did the Mets even end up in the position of having their entire season depend on the sulky Glavine? An easy (and popular) answer is that Guillermo Mota did it. And granted, Mota was terrible all year. But really, perpetrating a baseball crime this heinous is too great a task for any one man, no matter how hittable his fastball.

Some of the guys over at Baseball Prospectus who are way smarter than I am have a simulation that they used to estimate each team’s chances of making the playoffs based on current standings, the expected win percentage over the remaining games, and . . . I don’t know, logarithms and shit. The actual method isn’t important. Anyway, they figured that, at their highest, the Mets had a 99.8% chance to make the playoffs, behind only the 1995 Angels as the team most likely to make the playoffs but which ultimately failed to do so.

Those are 500-to-1 odds against. In order to miss out on the playoffs, the Mets would need to run into a perfect storm of bad luck, poor performance, and hot competition in order to complete the choke. Of course, that’s why they call ‘em the Amazin’s.

The Starters

On Monday, September 17th, the Mets woke up with a 3.5 game lead and an estimated 99% chance of making the playoffs, despite having just been swept by the Phillies . . . again.** Over the remaining two weeks of the season, the Mets would get just three quality starts, and only manage to win two of those, going 5-9 overall. For the last 14 games of the season, the Mets averaged less than five innings per start. That’s terrible. But it just completed a disturbing trend that’d been developing all year. Take a look at this:

In the NL this year, starting pitchers accounted for 64% of batters faced, or a little less than two thirds (I will admit to using BFP, or Batters Faced by Pitcher, rather than Innings Pitched, simply because it was easier for me to find a month-by-month breakdown, and I am a very lazy man). Translate that into a game, and that’s six innings for the starter, and three innings for the bullpen. As the season went on, though, the Mets bullpen was being asked to carry a progressively heavier and heavier load; culminating in a two-week period where they averaged four innings per game, and as many as five, five and a third, or ultimately eight and two-thirds innings a game. That’s too much.

The more innings that have to be pitched by the bullpen, the more a manager has to rely on the Guillermo Motas and Aaron Seles of the world; and Mota and Sele are terrible. This sets up a vicious cycle where relievers may have to come into a game early and without sufficient rest, then pitch badly because they are tired, only to be replaced by an even worse reliever who pitches badly because he is also overworked, and so on, and so on; until Billy Wagner openly complains about how badly you manage your bullpen and you just wish Flanders was dead.

It looks like a small difference, but during this stretch the Mets bullpen was pitching 20% more than NL average.

Bad Luck

Some commenter here on MetsGeek (I wish I could remember who) wrote during this bleak two-week stretch that it looked to him like the Mets were the victims of an unusually high number of bleeders, dying quails, dinks and dunks; and wondered what the Mets BABIP was (Batting Average on Balls In Play). Well, here you go:

The average BABIP in the NL for 2007 was .304 for starting pitchers, and .298 for relievers. Anything much different than about .300 is probably the result of luck. It looks like a lot of the starting pitchers’ woes were due to just that – bad luck. The Mets starters actually had pretty good peripherals during this time, too: 9.5 K/9, 3.4 BB/9, and 1.2 HR/9. So you wouldn’t expect them to have gotten as shellacked as they did. Sadly, this was just a regression to the mean from the ridiculously good BABIP the Mets posted in the first half.

The Schedule

This is also a bit of bad luck. The Mets were scheduled to play 16 games over the final 17 days of the season. A rescheduled rainout from June 28th against the Cardinals meant that the Mets would have to play 17 straight, without a single day off. Where’s Crash when you need him? The lack of a day off, combined with starters who couldn’t go past the fifth inning and Billy Wagner’s back spasms, meant that by the time the Mets got to the last week, the bullpen was reeling. Many people (me included) criticized Willie Randolph for using sub-par relievers like Mota and Schoeneweis in important game situations, but the truth is that he had to run somebody out there. L’il Pedro and Heilman couldn’t pitch every inning. Now, would a single day off have turned these Mets back into a 97-win team? Probably not. But it might have made the difference in a game or two, and two games would have given the Mets the division.

The Competition

Finally, for this historical collapse to be complete, three other teams all had to play well. Obviously the Phillies kept the pressure on by going 9-4 in the final two weeks of the season, and hats off to them (jerks). They played well when it counted. But they were still only one game better than the Mets. If they had played out their final 13 games closer to their established winning percentage of .540, the Mets could still have taken the division. Likewise, if either the Rockies or the Padres had played only a little bit worse, then the Mets could have ended up in the playoffs as the Wildcard. If the Rockies had “only,” won 11 of their last 14, instead of 13-of-14, the Mets would have played the Diamondbacks in the first round of the playoffs. If you look at three teams over a given period you can usually expect at least one of them to be less than awesome.

The End

There was a reason why the Mets still had an estimated 99% chance to make the playoffs with only two weeks to play – so many highly unlikely things would all have to happen for the Mets to miss out on the playoffs entirely. Not only would the Mets have to play badly (and unluckily) over 129 innings in just 14 days, but the Phillies would have to play well, the Rockies would have to play ridiculously well, and the Padres would have to bend but not break. What are the odds that all of those things would happen? One in a hundred? Well, there you go. I don’t think that the 2007 Mets were a deeply flawed team, or one that had a single, glaring weakness. If any one of these things had gone just a little bit differently, the Mets could still have made the postseason. Rather, it was the accumulation of small, unlikely events that combined to kill the Mets season. It was like being nibbled to death by mice.

At work the following Monday morning on October 1st, my friend Tom from Philadelphia came by my office to gloat/tempt me to kick him in the throat, and I’ll close with what I told him then. Yes, the Phillies won the division; and yes, the Mets missed out on the playoffs by collapsing in historic fashion to the jeers of the nation. But it’s not like the Phillies were such a shit-hot team – they won 89 games. Really, it isn’t so much that the Phillies won the division, as that the Mets lost it. Like, if I drop $20 on the ground and Tom picks it up, it’s not like he has a new job. He didn’t earn that 20 bucks; I just got careless. Next time, hopefully, maybe I won’t be so careless.

Thanks to Baseball Prospectus, David Pinto’s Baseball Musings, and for making my inaccurate and uninformative ramblings possible.

*This nightmare of a game represented the last of the partial season tickets I got with some buddies. As we were splitting up games after the All-Star break, I remember thinking that the Mets would have the division sewn up by then and would be resting guys for the playoffs, so it’d be a meaningless game; probably played by minor leaguers. Man, those were good times.

** I was at this game, too. Clearly, I need to go to fewer games.

Labels: , , ,

Thursday, November 08, 2007

Congratulations, David

Well, good job, David. Uh . . . and Beltran, too. Of course, Beltran winning a Gold Glove comes as no surpise. I think Wright is a pretty good defender, but Gold Glove? Of course, in the NL, who else is there at third? Rolen I guess, but he only played in 112 games this year, so . . . Zimmerman? Ramirez? Probably either one of those guys would be a better choice, but neither one of them had the year at the plate that Wright did.