Let's start by making one thing clear: Everyone wants what's best for Stephen Strasburg. He is the greatest collegiate pitcher of all time, and in his short major league career, Strasburg has averaged more strikeouts per nine innings than Randy Johnson. No matter where you stand on The Strasburg Rules, everyone wants to see him pitch for 20 seasons, strike out 5,000 batters, and make a pretty speech in upstate New York in the summer of 2039.
It's worth stating the obvious, because to hear some of the rhetoric, you'd think that there are only two positions on the subject — those who believe the Nationals are correct to shut down Strasburg and those who think it's fine to sacrifice Strasburg's arm on the altar of October baseball.
But there is a middle ground, which comes from understanding that the only sure way to eliminate injuries in baseball is to eliminate baseball. There is some baseline risk of injury that exists every time a player takes the field. That risk is elevated for pitchers, whose value derives solely from the kinetic chain that is centered in the shoulder and elbow of their pitching arm. But injuries are not 100 percent preventable. Just ask Brandon McCarthy: He left the ICU Sunday after brain surgery to remove an epidural hematoma, caused by a line drive that fractured his skull.
If the goal is the complete prevention of pitching-related injuries, the only answer is to turn over the job to an indestructible batting-practice machine (or Livan Hernandez, which is the same thing). But if the goal is to prevent pitchers from an elevated injury risk due to overuse of their arms, Major League Baseball accomplished that several years ago.
Stephen Strasburg was supposed to make his final start of the season tonight, but after he allowed five runs in three innings last Friday, the Nationals announced that he would be shut down immediately. Manager Davey Johnson, in what was not his finest moment, blamed "media hype" for not allowing Strasburg to "be totally mentally concentrating on the job at hand."
That's right — the Nationals ended Strasburg's season earlier than expected because of the media response to their decision to end Strasburg's season earlier than expected. How meta.
There are many reasons to question Nationals GM Mike Rizzo's premeditated decision to shut down Strasburg with his team almost certain to win the NL East. (Although, to judge from reading the D.C. press, anyone who questions the Nationals can, in the words of Tom Boswell, "kiss my press pass.") Many of those reasons have been discussed elsewhere, but perhaps the biggest reason to question Rizzo's decision hasn't received enough attention: The Nationals are making Strasburg pay for sins inflicted on pitchers from a different generation.
Until a quarter century ago, it wasn't clear that baseball teams realized it was possible to overuse a pitcher. Pitch counts weren't kept on a systemic basis in MLB until 1988, and they were originally kept by STATS Inc., a company that wasn't the league's official statistician.
Speaking of which, has anyone figured out why the Nationals kept Strasburg on an innings limit and not a pitch limit? There's a reason why teams keep their starters on a pitch count and not an innings count. A 100-pitch complete game puts less stress on a pitcher's arm than a 120-pitch slog through six innings.
Early on, pitch counts were kept mostly as a curiosity or a macabre way to measure the carnage endured by a pitcher's arm. They certainly didn't influence how teams handled their pitchers. In 1989, 23-year-old Al Leiter threw 163 pitches in his second start of the season. Three starts later, he went on the DL and pitched in just eight games over the next three years. Leiter would go on to win 162 games in his career and 155 of them after his 27th birthday.
Leiter's case was unusual but hardly unprecedented. In 1989 alone, starters threw at least 140 pitches 55 times. Leiter's wasn't even the worst case of pitcher abuse that season. In his final start of 1989, a game between two teams with no chance of making the playoffs, Orel Hershiser was allowed to throw 11 innings and 169 pitches in a win that prevented him from having a losing record that season. That capped a three-year stretch in which Hershiser led the National League in innings pitched each season. Four starts into the 1990 season, Hershiser tore his rotator cuff.
The abuse of starting pitchers was so rampant that not even Tim Wakefield was immune. Knuckleballers, by virtue of their low-stress and low-velocity delivery, are thought to be relatively safe from arm injuries. And they are — if you don't make them throw 172 pitches.
Wakefield was the rookie sensation of 1992, emerging from obscurity to give the Pirates a 2.15 ERA in 13 starts and then two complete-game wins against the Braves in the NLCS.1 But after gutting out 10 innings and 172 pitches in his fifth start of 1993, Wakefield fell apart; he finished the season with a 5.61 ERA. He spent all of 1994 in Triple-A with an even higher ERA of 5.84, prompting the Pirates to release him. Rumor has it the Red Sox gave him a second chance the following year.
Those are the most dramatic examples of baseball's widespread indifference toward the most valuable commodities in the sport. In every year but one from 1988 through 1999, starters were allowed to throw 130 pitches in a game at least 100 times. Many pitchers held up under the strain; many others didn't. For every Randy Johnson, who racked up 105 starts of 130-plus pitches in his long career, there was a Bobby Witt. Witt threw 130-plus pitches 24 times between 1988 and 1990. He had his greatest season in 1990, with career bests in strikeouts (221) and ERA (3.36). In 1991, he blew out his shoulder. He pitched for another decade as an innings-eating junkballer; the ace was gone.
While Witt was in the majors — and running up his pitch count — when he was just 22, Johnson didn't make his major league debut until after his 25th birthday. This is relevant because evidence that had been around since the mid-1980s suggested that pitchers under the age of 25 were particularly vulnerable to overuse. A 25-year-old starter might hold up under the strain of a heavy workload, but a 22-year-old starter probably wouldn't.
(There's a litany of talented young pitchers from the late '80s and early '90s who blew out their arms from overuse: Edwin Correa. Juan Nieves. Floyd Youmans. Jose DeJesus.)
As the 1990s progressed, teams slowly reduced the number of 130-plus pitch outings to which they subjected their pitchers. Teams at the forefront of the sabermetric revolution (i.e., the Oakland A's) led this effort. In 1988, pitchers threw 130-plus pitches in
a game 236 times. By 1993, that number was down to 159; by 1998, it was down to 133. When a pitching phenom emerged, whether his arm would hold up was sometimes a function of when in the decade, and with what team, he emerged. Case in point: Ramon Martinez was supposed to be the best pitcher in his family. As a 22-year-old in 1990 he made the All-Star team and finished second in the Cy Young vote. Ramon pitched for the Dodgers and notorious arm destroyer Tommy Lasorda, and by the time he turned 23, Ramon had already thrown 130-plus pitches 12 different times. He never made the All-Star team again and was essentially done as a pitcher by the time he turned 31.
His younger brother Pedro Martinez was lucky to be used in middle relief as a rookie because Lasorda didn't think Pedro had the stamina to be a starter. He was then traded to the Montreal Expos, where manager Felipe Alou had a more enlightened approach to pitcher use. Pedro threw 130 pitches in a start just once before he turned 25. While Pedro was also overused in his prime — and was left in too long on one memorable October night — he had the Hall of Fame career that most scouts had expected from his older brother.
New York Mets fans surely remember "Generation K," a trio of top pitching prospects who were supposed to bring glory back to Flushing Meadows. But all three pitchers — Bill Pulsipher, Paul Wilson, and Jason Isringhausen — were allowed to throw too many pitches at too young an age, and all three got hurt. Isringhausen got lucky; he hurt his elbow and made it back as a closer after the Mets gave up on him. Pulsipher and Wilson blew out their shoulders and were finished before they ever got started.
When Baseball Prospectus launched in 1996, ending pitcher abuse was one of the first causes we championed. Beginning with this article in 1998, I introduced a statistic known as Pitcher Abuse Points (PAP), which attempted to quantify the risk being placed on starters' arms. This stat was refined over the years.
Then something funny happened: Baseball teams began to see the light. In 1998, Kerry Wood was a 20-year-old pitcher who threw perhaps the most dominant game in baseball history in his fifth start. He also threw 120 pitches eight times that season, punctuated by a 133-pitch outing on August 26. One start later, Wood developed an arm problem that forced him to be shut down, with the Cubs in a pennant race. Wood returned to throw five innings in an NLDS start before the Cubs were swept.
The following spring training, Wood tore his UCL and underwent Tommy John surgery. The question of whether a pitcher threw too many pitches at too young an age was finally being discussed at a baseball-wide level. That season, the Detroit Tigers announced that their prized young arm Jeff Weaver would be limited to 110 pitches in every start.
In its own way, the Tigers' decision to limit Weaver was as radical as anything the Nationals are doing with Stephen Strasburg. Pulling a starter when he reached his pitch limit, regardless of the situation in the game, was just as unprecedented as shutting down Strasburg when he reached his innings limit, regardless of the situation in the season. But the Tigers sucked, pitch counts were just entering the baseball consciousness, and no one really cared.
Did it work? Weaver never developed into more than a no. 3 starter, but he stayed in a major league rotation and was virtually injury-free for a decade. To invert Neil Young's lyric, Weaver chose to fade away rather than burn out.
Jeff Weaver was a top prospect but not a phenom. Rick Ankiel was both, and he had the agent to match. As a 20-year-old in 2000, Ankiel struck out 194 batters in 175 innings and nearly won Rookie of the Year honors. He was also the source of some mid-season controversy, when his agent — the one and only Scott Boras — publicly berated the St. Louis Cardinals for breaching an agreement to keep Ankiel on the same 110-pitch limit that Weaver had been on. (Ankiel exceeded 110 pitches three times that May, topping out at 120 pitches.)
Now this was a story — the Cardinals were contenders, Ankiel was a big part of it, and Boras knows how to stir up a media storm. For the first time pitch counts were a subject of discussion on SportsCenter. Boras got his way — Ankiel didn't exceed 111 pitches the rest of the season. And pitch counts were suddenly everywhere — listed next to the radar gun at the ballpark and updated every few minutes on TV and radio broadcasts.
Strict pitch limits are no cure for Ankiel's Steve Blass disease, but a generation of pitchers benefited. In 2000, major league starters reached 130 pitches 77 times. The following year, after the Ankiel controversy had sunk in, they did so just 27 times — a 71 percent drop in the span of a single offseason. By 2004, it happened just 14 times, and by 2006, just seven times all year. There have been fewer 130-plus pitch outings in the past 12 years combined than there were in every single season before the 1994-95 strike.
This is the single greatest change in the way baseball is played in the 21st century, and it isn't close. Every other significant change — the proliferation of seven- and eight-man bullpens, the endless shuttling of relievers into the game to gain the platoon advantage, the disappearance of the pinch-hitter because teams would rather carry another reliever instead — stems from this industry-wide decision to embrace pitch limits as a way to protect starting pitchers.
Along the way, the PAP statistic became obsolete, and I, for one, couldn't be happier. Pitcher abuse, as we understood it 15 years ago, simply doesn't exist anymore.
So are starters today — particularly young starters — less likely to get hurt? It's hard to say, because in any given year, the starters who throw the most pitches are asked to do so because they're the elite starters in the sport. The pitchers with the most high-pitch-count outings were routinely Roger Clemens, Randy Johnson, and Pedro Martinez; if you analyzed the data that way, you'd think that throwing more pitches prevented injury. In the past four years, the pitchers with the most 130-pitch outings are Justin Verlander and Roy Halladay, who were allowed to throw so many pitches precisely because they were thought to be indestructible.
Enough time has passed that we can now compare the best pitchers across eras — Verlander and Halladay, for instance, have both hit the 130-pitch mark just three times in the past four years, which was a bad month for Randy Johnson in his prime. But since we're talking Strasburg, let's focus on young starting pitchers.
Here's a sample comparison, between the two biggest pitching phenoms of the past 40 years, the only two pitchers in that time to strike out 75 batters in a season while still in their teens: Dwight Gooden and Felix Hernandez.
Gooden ranks behind only Bob Feller in the pantheon of teenage pitchers. He was Rookie of the Year in 1984, at age 19, leading the NL in strikeouts. He won the Cy Young Award the next year with a season for the ages: 24-4, 1.53 ERA. He also threw 16 complete games and 277 innings that season. Gooden never came anywhere near that level of success again. After 1985, he never had an ERA+ of higher than 127.2
It's easy to blame drugs for the destruction of Gooden's career — but he actually pitched well in 1987, when he returned from rehab, and again in 1988. But in 1989, he tore his rotator cuff and missed half the season; he was just 24 years old. Gooden returned in 1990 but was never the same.
King Felix was 19 when he came up in mid-season in 2005, and fashioned a 2.67 ERA in 12 starts. He wasn't nearly as effective in 2006, with a 4.52 ERA, but he still took the ball every fifth day. The difference is that, whereas Gooden threw 277 innings as a sophomore, Hernandez threw 191. We don't have pitch counts from Gooden's first four seasons, but anecdotally we know that Gooden frequently exceeded 140 pitches when his fastball was humming and his curveball was darting. Hernandez, on the other hand, never threw more than 120 pitches until the end of his fifth season in the majors, in 2009, when he threw 121. To this day, Hernandez has never thrown more than 128 pitches in a game.
Whereas Gooden peaked at 20, Hernandez was just beginning his climb. His ERA improved from 4.52 to 3.92 to 3.45 to 2.49 in 2009, when he finished second in the AL Cy Young vote. In 2010, he led the league with a 2.27 ERA and 250 innings pitched, and despite a 13-12 record deservedly won the Cy Young. Hernandez was 24, the same age Gooden was when he blew out his shoulder. Hernandez has thrown at least 30 starts every year since his partial rookie season; this year he has a perfect game, five shutouts, and is a leading contender for another Cy Young Award. While Gooden never had an ERA+ of higher than 127 after age 20, Hernandez is working on his third ERA+ of 140 or higher in the past four years.
The Doc and The King offer just one comparison. Let's look at some more comprehensive data. Going back to 1984 — Gooden's rookie season — I found every pitcher who (1) made 25 or more starts, and (2) was no more than 22 years old. The point was to isolate pitchers who were in a major league rotation at a precocious age — these are the pitchers whom teams want to protect.
I then made a note as to whether, five years later, that pitcher once again made 25 or more starts. If he did, he was a "survivor."
Here's the chart:
(1994-95 are missing from the list because of the strike, and 1989-90 are missing for the same reason — it's not fair to evaluate how a pitcher in 1989 did five years later when that season ended in August.)
You'll notice that starting in 1999, the year the Tigers limited Jeff Weaver's pitch counts and other teams quietly followed suit (like the Expos with Javier Vazquez), the number of pitchers who "survived" to pitch five years later ticked upward. The change becomes more obvious if we combine the numbers:
From 1984 to 1998, one out of every two young starters was still starting regularly five years later. From 1999 onward, two out of every three young starters have done so. These are the best young pitchers in the game — and their failure rate has been cut by a third. That ain't beanbag. You can credit some of this to improved surgical techniques, perhaps, but it's worth noting that the success rate in the 1980s (13-of-24) was actually higher than it was from 1991 to 1998 (7-of-16), even though Tommy John surgery, for instance, was commonplace in the 1990s. More to the point, almost none of the 19 "survivors" from 1999 to 2006 overcame major arm problems through surgery or rehab — nearly all of them avoided injury in the first place.
That's the take-home point here: Major league baseball teams have dramatically altered the way they handle starting pitchers — and in doing so, they have significantly reduced the risk of injury to those pitchers.
If you look at it superficially, you could argue that the way MLB has improved the way it handles young pitchers proves the Nationals' point — that reducing a starter's workload reduces his risk of injury, so shutting down Strasburg will keep him healthy.
But the sabermetric argument is the opposite: That precisely because the industry has already reduced the risk of pitcher injuries significantly, there is less to be gained by further reducing Strasburg's workload. Strasburg, like every pitcher of his generation, has had his pitches monitored from the moment he signed a pro contract. Strasburg never threw even 100 pitches in a game as a rookie before he blew out his elbow, nor did he throw even 100 pitches after returning from Tommy John surgery last year. This year the gloves have come off a little; he threw 119 pitches in one start, 111 in another, and no more than 108 pitches in any other start.
Strasburg is the product of an era in which a pitcher's well-being already comes first. This approach has succeeded in reducing injuries significantly, but not entirely. There's only so much risk that can be squeezed out of the equation, no matter how much you protect a pitcher's arm. Some pitchers will get hurt no matter how well you protect them. You know which pitcher illustrates this best? Stephen Strasburg.
Would further reducing young pitchers' workloads have made the overall survivor rate even higher? Let's look at the eight pitchers — two of them qualified twice — from 1999 to 2006 who didn't survive to make 25 starts five years later:
- Rick Ankiel (2000) — developed "Steve Blass disease"
- Dontrelle Willis (2003 and 2004) — also stopped throwing strikes entirely; went on the DL with anxiety
- Jeremy Bonderman (2003 and 2004) — developed thoracic outlet syndrome, a circulatory problem that reduces blood supply to the shoulder
- Ryan Dempster (1999) — was never very effective as a starter; was moved to the bullpen, where he found success (and eventually moved back to the rotation, where he is still effective)
- Oliver Perez (2004) — very wild at his best, developed knee problems in 2009 and stopped throwing strikes; has made it back this year as a middle reliever
That leaves just three pitchers whose failure can be tied to arm problems likely caused by overuse. Scott Olsen, a talented lefty with the Marlins in 2006, tore the labrum in his shoulder in 2009. Ben Sheets, a rookie in 2001, missed half the season in 2006 with shoulder tendinitis — but bounced back to pitch well in 2007 and 2008. He missed all of 2009 and 2011 with elbow issues, but surprised everyone by signing with the Braves this July and returning to the majors.
The third pitcher is Mark Prior. That's a total of three pitchers who had injury problems that can be linked to overuse, and in Prior's case, he pitched for the Chicago Cubs, the one team in baseball that sat out the pitch-count revolution.
Prior was Stephen Strasburg before Stephen Strasburg. As a rookie he struck out 147 batters in 117 innings; as a 22-year-old sophomore he threw 211 innings with a 2.43 ERA and pitched the Cubs to within five outs of the World Series.
Along the way, however, he threw a ridiculous number of pitches. In 2003, the Cubs were managed by Dusty Baker, who at the time was a holdout against the new approach to starting pitchers.3 That season, Cubs starters threw 120-plus pitches in a game 29 times, the most by any team in the past 12 years. Prior threw nine of those games — five of them in September alone. Here are Prior's pitch counts in September: 131, 129, 109, 124, 131, 133. In his first playoff start, he threw 133 pitches. In his second start, despite the Cubs being up 11-0 after five innings, he was left in to throw seven innings and 116 pitches. In his third playoff start, he tired in the eighth inning, and everyone blamed Steve Bartman.
Prior was never the same. He missed the first two months of the 2004 season with tendinitis in his ankle, but wasn't nearly as effective when he returned. He missed part of 2005 with inflammation in his elbow. And in 2006, after missing part of the season with shoulder inflammation, he came back and was shelled, posting a 7.21 ERA in nine starts. After multiple shoulder surgeries, he has never pitched in the majors again, though he's still trying, God bless him; Prior pitched for Triple-A Pawtucket this season.
Mark Prior threw as many games with 130-plus pitches in September and October of 2003 as every pitcher in the major leagues combined in 2012. He may be the precautionary tale of what could happen to Stephen Strasburg — but the Nationals have already taken the necessary precautions to prevent the same fate. Regardless of whether or not they shut down Strasburg, the way the Nationals have used Strasburg is nothing like the way the Cubs used Prior.
Major League Baseball before the turn of the century was like a highway with a speed limit of 80 mph. Baseball today has a speed limit of 55 mph, seat belts are mandated, and air bags are standard. What the Nationals are doing is lowering the speed limit to 40 mph and arguing that it will reduce car accidents further.
They might be right, but given that the injury risk has already been reduced so significantly, it's likely that any further benefit to shutting down Strasburg will be minuscule. Meanwhile, the risk that shutting him down costs the Nationals the NL pennant or a world championship is a lot more than minuscule. The point of having a pitcher like Stephen Strasburg is to help you win a championship. Preventing Strasburg from helping you win a title this year — so that he might be more likely to help you win a title in the future — is causing certain harm to your team in the present for a theoretical benefit in the future. That is, in a word, dumb.
That leaves the issue of Strasburg's recovery from Tommy John surgery, which the Nationals will tell you is the entire reason behind their plan to shut him down. They'll tell you that Strasburg is on the same plan that teammate Jordan Zimmermann was on, and that Zimmermann was limited to the same number of innings in 2011, and that Zimmermann has pitched just fine this year. (Never mind that the Nationals were miles from a pennant race last year, or that since August 1 Zimmermann has a 5.54 ERA, so it's not like being babied last year has kept him from being gassed this year.)
There are a couple of holes that need poking in this position. The first is that pitchers have been having Tommy John surgery for decades, and once they're healthy enough to return to the mound, they've typically been healthy enough to pitch a full season. You know who wasn't limited to 160 innings in his first season after Tommy John surgery? Tommy John.
Patient Zero missed the entire 1975 season after Dr. Frank Jobe replaced his ulnar collateral ligament. John returned in 1976, made 31 starts, and threw 207 innings. He was so devastated by this flagrant overuse that he made at least 30 starts every year (not counting the 1981 strike) until 1984, when he was 41 years old.
John was 33 years old when he returned from surgery, 10 years older than Strasburg. The Nationals have argued that there are no real precedents for Strasburg — a pitcher so young and talented who has returned from Tommy John surgery — so they had to be extra cautious with him. The Nats are right that it's unusual for a pitcher of Strasburg's caliber to already be a Tommy John survivor in his early 20s. Really, the most comparable is Kerry Wood. And as with Prior, the way the Cubs handled Wood is nothing like the way the Nationals have handled Strasburg.
Wood returned from Tommy John surgery in 2000 and was handled reasonably at first, only reaching 120 pitches twice in his first season back. And here's the thing: Wood recovered nicely — after a 4.80 ERA in 2000, Wood had a 3.36 ERA in 174 innings in 2001, then a 3.66 ERA in 2002, then a 3.20 ERA and a league-leading 266 strikeouts in 2003.
But once he showed he had returned to form, his workload ramped up quickly. In 2001 he threw more than 100 pitches in 21 consecutive starts. (Strasburg's highest such streak is three.) And in 2003, with Baker at the helm, Wood was sent to the galleys. He threw 120-plus pitches in a game 13 times. On May 10, he threw 141 pitches, one of the 15 highest totals of any pitcher this century. In his last six starts, he went 125, 120, 122, 114, 125, 122. He threw 124 and 117 pitches in his two divisional series starts, 109 in Game 3 of the NLCS, and then 112 in Game 7 before he was pulled, having allowed seven runs in 5⅔ innings.
And like Prior, he broke down. Wood made just 22 starts in 2004, then 14 in 2005 and 2006 combined before he moved to the bullpen full-time for the rest of his career. But what destroyed Wood wasn't how the Cubs handled him after Tommy John surgery — it was how they abused him four years later. The lesson of Wood's demise isn't that you can't throw more than 160 innings after Tommy John surgery — it's that you can't throw 120 pitches every single time you take the mound. Given that Strasburg hasn't thrown 120 pitches any time he's taken the mound, this shouldn't be a problem. Yet the Nationals have turned it into one.
There are so many reasons why shutting down Strasburg is a mistake. Having made the decision to limit his innings before the season, the Nationals' refusal to skip the occasional start early in the year so that Strasburg would be fresh for October is mind-boggling. (The only explanation is that they didn't think they'd actually be in contention this year. In which case, may we suggest they read Grantland more often?)
The decision to limit Strasburg's innings instead of his pitches suggests a team that has been suckered by the Verducci Effect, named for the Sports Illustrated writer Tom Verducci, who posited years ago that young pitchers who exceed their previous career high in innings by more than 30 are likely to get hurt. It's a pretty theory, but it has been debunked by analysts many times over.
Or maybe they actually allowed themselves to be influenced by Scott Boras, Strasburg's agent, who has openly supported the decision to shut down his client. Boras's position is that a risk of injury to his client, no matter how small, reduces his future earning potential, while the benefit of allowing Strasburg to pitch — a possible world championship — doesn't increase Strasburg's income by more than the amount of his playoff share. That's a legitimate position for an agent. But for the Nationals? Why should they care what payday awaits Strasburg in free agency?
That Boras's opinion is taken seriously on this subject is mystifying. I'd love to see Mike Rizzo tell Boras: "Since you want us to put Strasburg's future ahead of our franchise's chances to win in the present, I'm sure you want us to have an investment in Strasburg's future as well. How about a long-term contract?" I suspect it would be a short conversation.
But the main reason the Nationals are wrong to shut down Strasburg is simply this: The risk they're trying to mitigate has already been mitigated for them. Major League Baseball has changed the way it uses starting pitchers, and has succeeded in reducing pitcher injuries. The Nationals' failure to recognize this is putting them at needless risk for something else — a quick exit this October.