On September 30th, Tom Glavine started the biggest game of the year for the Mets, a must-win for the Mets to have a real prayer at making the playoffs. And Tom Glavine proceeded to pitch the worst outing of his career, an affair where he gave up seven runs while recording just one out, effectively killing the Mets’ chances before the team had a chance to hit. While I doubt he wants that to be his last outing in the big leagues, he has almost certainly thrown his last pitch in a Mets uniform.

I won’t suggest that Glavine didn’t care or didn’t try, having already won his 300th victory. He’s too much the consummate professional for that; even if he didn’t particularly care about winning, there were those that did, and I don’t think Glavine would deny them their chance. Furthermore, in the seven starts immediately following his 300th win, he went 3-0 with a 2.54 ERA over 46 innings, arguably his best stretch of the season. What I think actually happened that fateful September afternoon was simple: he was tired. He had been pitching poorly for a month at that point and, at 41, just didn’t have much left in the tank so late in the season.

With Glavine probably on his way out of New York, it seems as good a time as any to look back at his career. And what a career it is, with 14 seasons of 200 or more innings, ten seasons with fifteen victories, and fifteen where he posted an ERA better than the league average. He’s earned a ticket to Cooperstown, and is 21st on the all-time wins list.

And you know what the most amazing thing of all is? He’s never seemed like he should be any good.

While Tom Glavine had decent stuff early in his career—a low 90’s fastball, a great circle changeup, and two solid breaking pitches he never used—as he’s gotten older he’s had to get by with much less than he used to have.

Josh Kalk has been doing a lot of phenomenal research with the PITCHf/x data currently available thanks to Major League Baseball. He’s spent a lot of time compiling the data into easy-to-read formats, such as player cards for every pitcher for whom there’s a sizeable chunk of data available (100 pitches). On these cards, you’ll find statistics about that pitchers pitches (like frequencies, average break, and average speed) and plots that show how much break he gets on his pitches. Tom Glavine’s card confirms what you can pretty much gather from watching any of Glavine’s outings. While it’s not perfect—at least a few of those slower sliders are really curveballs, and some of the faster ones are actually cutters—it does demonstrate a few fundamental truths about Glavine.

First, his pitches have no movement whatsoever. His fastball has little lateral break and not a lot of sink for an 85 mile-per-hour pitch. His changeup breaks almost identically. Second, he really doesn’t get a big difference in velocity from his fastball to his change. In all, the average difference is slightly less than seven miles-per-hour. Looking at a comparable pitcher, Jamie Moyer has roughly a 7.6 mile-per-hour difference. Finally, while Glavine may throw his breaking pitches more often than he used to, they’re still something of a last resort.

So what do we have? A 41-year-old pitcher who throws one of two pitches 92 percent of the time, and the only thing different about those two pitches, for the most part, is about seven digits on a radar gun. Of course, there are plenty of pitchers who succeed with sub-par stuff; they usually have outstanding control, some other means of missing bats (like a funky delivery), or extreme groundball rates.

Tom Glavine? He’s got none of these things. His control is good, but not great, his walk ratios over the course of his career being only ten percent below the league average. He doesn’t get strikeouts, posting above average ratios just twice in his career. In fact, given what we now know about Fielding Independent Pitching, Glavine becomes a huge anomaly: his strikeout-to-walk ratios are downright poor. In fact, of any pitcher who has won 200 games in the modern era, only three have worse ratios when compared to the league average: Wabash George Mullin, Joe Niekro, and Charlie Hough. The latter two make sense, being knuckleballers who didn’t always know where their pitches were going. Mullin wasn’t comparable to Glavine, either, being a fastball pitcher with a big curve in the early 1900’s. And none of those guys were ever as successful as Glavine.

He’s never gotten an extreme number of groundballs, either, despite not allowing too many homeruns. His career groundball-to-flyball ratio is 1.43, above average, but not tremendously so. In short, Glavine has had a career so far above and beyond what you could reasonably expect, it’s mind-boggling.

How has he done it? The easy answer is location. Since he doesn’t have top-notch stuff, and he doesn’t do a great job keeping batters off the basepaths, or do anything else good pitchers do, he must be doing something special with his location. That much is obvious. But I’m going to offer a second reason: timing.

Warren Spahn famously said, “Hitting is timing. Pitching is upsetting timing.” However, when I say timing, I don’t mean it that way, although I’m sure Glavine uses that seven mile-an-hour difference between his changeup and heater to as good effect as he can. No, Glavine upsets a team’s timing. Consider his career splits in various situations:

Split          BA    OBA    SLG
Total        .257   .318   .376
RISP         .250   .355   .351
- - -        .255   .303   .382
Men on       .259   .339   .367
1 - -        .271   .316   .385
- 2 -        .250   .388   .370
- - 3        .266   .399   .359
1 2 -        .234   .308   .339
1 - 3        .276   .327   .373
- 2 3        .240   .483   .330
1 2 3        .249   .272   .305

You’ll notice a few clear trends. When the bases are empty or there’s just a runner on first, Glavine is pretty stingy with the walk. He’ll give the hitter more to drive (as evidenced by higher slugging percentages), but he doesn’t like to give them a free pass to first. Second, when a runner is in scoring postion, Glavine reverts to damage control. If a base is open he’ll stay out of the strike zone and try to make the hitter chase. In these situations, Glavine avoids giving the hitter anything he can drive, likely by throwing fastballs out of the zone or changeups low and away. He’s always willing to give the hitter something, but he makes damn sure it’s something he can live with. In other words, he makes it exceedingly difficult for offenses to get into a rhythm.

And finally, and perhaps most impressively, Glavine is an incredible pitcher when the bases are loaded, limiting hitters to just a .577 OPS (and just a .305 slugging!). To put that into some perspective, only his former teammate, Greg Maddux—who’s nearly off the charts at .520—among recent pitchers is close to Glavine’s mark. Not Roger Clemens, nor Pedro Martinez, nor Randy Johnson, nor John Smoltz, nor anybody. Quite simply, if there’s a situation with the bases loaded and I need to get three outs, I’m taking Maddux, but Glavine’s my “Plan B.” Thumbing through historical splits since 1957, there are only two or three others I’d consider taking before him, too.

When the Mets first signed Glavine to a three-year, $35 million contract in November 2002, I was strongly against it. The Mets were signing a 37-year-old pitcher who had likely just finished the last great season of his career. He didn’t strike anyone out, his control was so-so, and there was no good reason for him to be doing what he was. And worst of all, he was a Brave. And you know what? Despite the fact that he ripped my heart out on the last day of the season, and despite the fact he never felt like a Met no matter how long he stuck around, I am sorry to see him go.

Tom Glavine is a truly unique pitcher in history. No pitcher has done so much with so many flaws, and we’ll probably never see another like him in a Mets uniform.

Notes and References

John Walsh of The Hardball Times made pitching plots for many major league pitchers using the PITCHf/x data, too. His results aren’t any different from Josh Kalk’s, but they might be easier to read for some. You can find Glavine’s on page two here.

Speaking of Josh Kalk and The Hardball Times, he just posted a new web tool that provides the location of every pitch, sorted by type and opposing batter, that a pitcher threw. For instance, you could look at the locations of all fastballs Glavine threw to Jimmy Rollins this season. I didn’t have time to really play around with it yet, but it’s neat and definitely relevant to the discussion at hand.