Filling the Gap: Reimagining Mike Mussina as John Smoltz

Tony Gwynn could not believe what he had just seen. He stared dumbfounded where the ball had fluttered past him and thudded into the fat part of Eddie Perez’s mitt. He turned toward the home dugout where his equally incredulous teammates were starting to crack up.

“Oh, my God, that’s a knuckleball.”

John Smoltz had just thrown Tony Gwynn a knuckleball. Then he threw him a slider, and Gwynn hit it off the wall for a double.

Like many pitchers, Smoltz had secretly experimented with the knuckler in the past. By late 1999, he needed to see if he could use it. He had already spent parts of the past two seasons on the disabled list with elbow issues. Now, the elbow was hurting again.

Less than a month later, in his playoff start against Houston, Smoltz dropped his arm slot from overhand to three-quarters in an attempt to mitigate the pain. He would allow 13 runs over 22.2 innings that postseason. The following spring, the pain returned. An MRI revealed a tear in Smoltz’s elbow requiring Tommy John surgery.

Meanwhile, Mike Mussina kept going.  He would pitch nine more seasons of quality baseball.  This year, his second of eligibility for Hall of Fame consideration, he was named on just under a quarter of the ballots.  Smoltz, chosen by five-sixths of the voters, was a first ballot shoo-in.  Let’s compare.

Even at the time of Smoltz’s injury, you already could have made the case that Mussina was catching up to him. Here are their numbers through 1999:

John Smoltz vs. Mike Mussina, Through 1999

Name	GS	Team W-L	IP	RA9	FIP*	rWAR	fWAR	K	uBB+HBP	HR
Smoltz	356	202-154	2,414.1	3.68	3.76	43.3	52.3	2,098	757	195
Mussina	254	159-95	1,772	3.71	4.13	42.0	39.5	1,325	426	182

*FIP includes HBP and excludes IBB, and is scaled to RA9 rather than ERA

Granted, Smoltz still had about 100 more starts than Mussina, but what Mussina was doing in those first 254 starts was phenomenal. Mussina was giving up runs at a nearly identical rate to Smoltz, only he was pitching in the higher-scoring American League. His team won more often when he was on the mound despite Baltimore being considerably worse than Atlanta over the course of the 1990s. And I know, a big part of Atlanta’s success in those years was Greg Maddux and Tom Glavine and pitching in general, which didn’t help Smoltz win his games. But still, Baltimore was winning over 62 percent of the time Mussina pitched, and the Orioles were under .500 without him.

That’s why Baseball-Reference has the two almost dead even in WAR in spite of the huge gap in games started and innings pitched. FanGraphs does give Smoltz a comfortable edge in WAR, but it’s entirely due to the difference in playing time. Mussina still compares favorably to Smoltz on a rate basis, averaging 4.98 fWAR per 32 starts compared to 4.70 for Smoltz. So when we’re talking about Smoltz having an extra 100 starts under his belt, we’re talking about whether what Atlanta got out of Smoltz was equal to what Baltimore got out of Mussina, plus:

A) another 100 starts it was already getting from guys like Rick Krivda, Scott Kamieniecki, or Arthur Rhodes* (B-R valuation), or;

B) another 100 starts from Mussina (FG valuation)

*Yes, Arthur Rhodes actually came up as a starter for Baltimore. Surprised me, too.

I know that’s not exactly a resounding case that Mussina had the better career through 1999, especially if you lean toward valuation B, because there’s a big difference between “Mussina” and “Mussina if he keeps this up another 100 starts.” Still, they were fairly comparable pitchers through 1999, at least as much as two pitchers with that big a gap in playing time can be.

With Smoltz’s injury, however, their careers began to diverge.

Smoltz missed all of the 2000 season. He began 2001 on the DL, started five games with a 5.76 RA9, and then went back on the DL with tendonitis in his surgically repaired elbow. Nothing ever came of the knuckleball experiment, but the Braves needed another contingency plan. When Smoltz came back off the DL that July, it was to the bullpen. He wouldn’t start another game until 2005.

Mussina, for his part, did keep it up for another 100 starts. Three years later, he was still allowing a similar number of runs to Smoltz in a higher-scoring league, putting up comparable fielding-independent numbers once you accounted for the league difference, and leading his team to a better record when he pitched:

Mike Mussina, 1991-2002

Name	GS	Team W-L	IP	RA9	FIP*	rWAR	fWAR	K	uBB+HBP	HR
Mussina	355	216-139	2,454	3.76	4.05	59.2	57.2	1,931	570	257

Smoltz was still pitching in the bullpen, of course, but for now, let’s just focus on their careers as starters. Smoltz eventually did return to the rotation in 2005, and he more or less picked up where he left off with three more really good years. Mussina remained an effective starter over the remainder of his career, putting up one of his better years (5.2 rWAR, 4.9 fWAR) in his final season before retiring.

Here are their career lines as starters (I excluded Smoltz’s five starts between injuries in 2001, just because it is simpler to treat that whole year as part of his post-surgery bullpen conversion, and because including them wouldn’t really help Smoltz anyway):

John Smoltz vs. Mike Mussina, Career Stats as Starting Pitchers

Name	GS	Team W-L	IP	RA9	FIP*	rWAR	fWAR	K	uBB+HBP	HR
Smoltz	476	271-205	3,286.2	3.67	3.74	59.1	70.1	2,784	932	267
Mussina	536	325-210*	3,559	3.94	4.05	82.7	82.5	2,812	816	376

*Mussina also pitched a tie game in 2003. Until 2007, tied games that were called due to weather after at least five innings were officially recorded as ties, with the game counting toward official stats but not the standings. The tie game was then replayed from the start at a later date. Now, such games are suspended and later resumed from the same point in the game instead, though ties can still occur if the suspended game can’t be finished in accordance with rule 4.12.

The gap in RA9 did open up a bit (though that is at least partially due to the Yankees being so much worse defensively than Baltimore had been), but the difference is still smaller than the effect of pitching in the AL versus pitching in the NL. Their fielding-independent performances stayed around the same, except now Mussina is the one with the big edge in games started.

As a starting pitcher, at least, Mussina got better results on a per-game basis, and he maintained those results over more starts. In the same terms as we looked at Smoltz’s games-started gap earlier, depending on whether you lean toward FanGraphs’ or Baseball-Reference’s valuation, Mussina’s career as a starter was somewhere between being roughly equal to Smoltz’s career as a starter, plus:

A) another 60 starts of vintage John Smoltz (FG valuation), or

B) another 60 starts of Pedro Martinez, circa 1999-2000 (B-R valuation)

…

It’s clear that Smoltz’s injury affected his career as a starter. We can see this in more detail by looking at his and Mussina’s careers season by season:

You can kind of see how the two compare, but it’s a bit hard to gauge with the numbers jumping around all over the place like that. Let me rearrange the seasons a bit to clean things up: here is the same graph, but with the points arranged so they mimic a more standard aging curve:

This really illustrates how the five seasons Smoltz spent either injured or in the bullpen fit into the picture. If you pretend the injury never happened and assume Smoltz stayed a productive starter that whole time, you can mentally fill in the gap in the graph, and you get a picture of two fairly similar pitchers: Smoltz becomes a slightly lesser version of Mussina, but with a bit more longevity.

Granted, eyeballing the graph to connect the dots is kind of a crude analytic exercise, especially after I rearranged the seasons to make it easier to follow. There’s really no way to know what Smoltz would have done had he stayed healthy and remained a starter. It does give us a general template to work with, though.

And even if we make some really optimistic assumptions about what Smoltz would have done–say that his elbow doesn’t really give him any trouble, and he maintains a high level of production the whole five years, and the extra wear doesn’t speed up the later shoulder injury that ended his 2008 season–it would take him averaging something like 4.5 rWAR per year over those five years to catch him up to Mussina’s total. That’s … really optimistic. Smoltz has exactly one five-year stretch (1995-99) when he averaged at least 4.5 rWAR. He’d have to repeat that just to get him level with Mussina.

fWAR, of course, rates Smoltz higher than rWAR. Repeating the above graph with its data:

We get a similar picture to what we saw from the Baseball-Reference data: Smoltz essentially looks like Mussina with a few years cut out of his career, except Mussina no longer has a noticeably better peak. Using this valuation, obviously you don’t have to be as optimistic with filling in the missing years to catch Smoltz back up to Mussina.

Now, all of this is comparing the two only as starting pitchers. Which, I know, ignores a pretty significant part of Smoltz’s career. And actually, it’s comparing the two only as starting pitchers during the regular season, which, again, ignores a significant part of Smoltz’s career. Of course, I can take Smoltz’s career, throw out everything he did as a closer and ignore what he did in the playoffs, and then say “see, Mussina is clearly better,” but that is a pretty hollow argument.

The point of all this is not to say that Mussina is clearly better. The point is to establish a baseline: Mussina’s career (at least in the regular season) is basically Smoltz’s career if Smoltz never got hurt. With that as a reference point, now we can go back and look at what Smoltz actually did over those five years following his surgery, and we can weigh that in with his record as a starter to get a full view of his career.

First, let me get this out of the way. Smoltz was outstanding in the postseason. He was beyond outstanding. He threw 209 postseason innings allowing just 2.89 runs per nine innings. That’s another full season worth of Smoltz at his very best. Mussina also was very good in the postseason: 139 innings, 3.61 RA9. That’s not Smoltz good, but we’re talking about Hall of Fame-caliber careers here, and Mussina in the postseason was still allowing fewer runs than Smoltz or Mussina in the regular season. I’m not going to get into how to weigh their respective postseason performances against the rest of their careers, but it’s something to keep in mind. Make of it what you will.

Now, the five years following Smoltz’s surgery.

The first year, he didn’t pitch at all. The first half of the second year was spent between the DL and a relatively ineffective stint in the rotation. That leaves three-and-a-half years of bullpen work to fill in the lost value.

You’ve probably heard at some point that Smoltz is the only pitcher in major league history with 200 wins and 150 saves (though Dennis Eckersly has 197 wins and 390 saves). And, well, that’s cool, but it’s not specifically why Smoltz belongs in the Hall of Fame. After all, Charlie Hough is the only pitcher with 215 wins and 60 saves, and no one cares*.

*I mean, people do care about Charlie Hough–I care about Charlie Hough. But no one cares, from a historical standpoint, that he is the only pitcher with 215 wins and 60 saves.

Milestones aside, Smoltz was legitimately one of the best relievers in the game from mid-2001 through 2004. That makes up some of the value lost as a starter. Let’s think for a minute about why Smoltz is the only pitcher with 200 wins and 150 saves, though.

Mussina picked up his 200th win in his third start of the 2004 season. If he had converted to a full-time closer at that point, could he have picked up 150 saves over the remaining five years before his retirement? Todd Jones saved 135 games from 2004-08, and he wasn’t even a closer in 2004, or for half of 2008. Francisco Cordero saved 186 games over that span. Nothing against Jones or Cordero, but Mussina was a much better pitcher than either of them.

That’s moot, of course, because Mussina didn’t convert to closer. He didn’t because no team in its right mind would take an established, Hall of Fame-caliber starter who is still pitching effectively and move him to the bullpen. It would kill his value. The same is true of practically every pitcher who has blown past the 200-win mark.

When the Yankees signed Mussina as a free agent before the 2001 season, they offered him $88.5 million guaranteed over six years. That same offseason, they offered Mariano Rivera four years and just under $40M, covering his final arbitration year and first three free-agent years. That’s Mariano Rivera, the greatest closer in the history of the game, and they didn’t offer him close to what they offered Mussina.

To this day, no team has offered a reliever anything close to the $88.5 million Mussina got as a starter 14 years ago. While starting pitcher contracts have now climbed to over $200 million (Clayton Kershaw), Jonathan Papelbon’s four-year, $50 million deal remains the most guaranteed money ever given to a reliever. Rivera, who in 2008 signed a three-year, $45 million contract, is the only reliever who has signed for as high an average annual salary as Mussina got, and even that was for half the length of Mussina’s deal and after league-wide payroll had risen nearly 40 percent since 2001.

The amount of money a player makes is obviously not a perfect analysis of how good he is, but at the same time, nobody is going to take a player whose contributions are valued at ~$90 million, and who is living up to those expectations, and then move him into a role where the best in the game don’t make close to that unless there is a reason to believe he no longer can succeed in the starting role.

Which brings us back to Smoltz. It’s not so much that he reached both the 200-win and 150-save milestones because he is uniquely talented in major league history. It’s that he was uniquely positioned, among pitchers of his caliber and talent, in a situation where a move to the bullpen during his prime years might make sense. No one else of his effectiveness has made that move because, even if Smoltz was the best reliever in the game for those three-and-a-half years (which he may well have been–he’s at least in the discussion with Rivera and Eric Gagne), he still wasn’t worth as much to a team as a good starting pitcher.

Atlanta made that move only because of Smoltz’s injury. Or not necessarily only because of the injury–things like the Braves’ already strong rotation and their pressing need to get rid of John Rocker were also factors–but the option was on the table only because of the injury. Moving Smoltz to the bullpen was a Plan B, an attempt to salvage something out of the realization that his elbow might not hold up to pitching at an elite level for 200-plus innings a year.

If we elevate Smoltz’s standing specifically because he succeeded as both a starter and a closer, we are implicitly suggesting that his injury actually improved his career. I think that is clearly not the case: the injury definitely hurt his career. If it didn’t, we probably would see more ace starters switched to relievers. And it’s not that teams don’t do that because they don’t know if their ace starter could succeed in the bullpen–they know they can. We see it in critical postseason games, à la Randy Johnson or Madison Bumgarner. Then, once the postseason heroics are done, those pitchers always go right back to the rotation.

Once Atlanta was confident that the old Smoltz was back, he went back to the rotation as well.

…

Let’s go back to our graphs from earlier and see what they look like if we fill in Smoltz’s actual production from the five missing years:

It is certainly impressive that Smoltz was able to return to pitching at such a high level after his injury, and those seasons do help his case. It’s just that the injury concerns meant that production was limited to about a third the number of innings, which in turn limited the amount of impact he could make on a season compared to what he did as a starter.

So, while it’s an interesting side effect of the injury that Smoltz ended up with his dual career as a starter-closer, I don’t think there is any question it detracted from his overall value. Atlanta would have preferred the more valuable Smoltz-as-a-starter the whole time to the possibly more interesting Smoltz-as-a-closer, just as the Braves liked Maddux and Glavine way too much to ever even consider what they might be able to do from the bullpen. And even if you hold closer-Smoltz in as high a regard as starter-Smoltz, don’t forget there is still the first year-and-a-half after the injury that was completely lost before he re-emerged in the bullpen.

This gets back to what I posited earlier about Mussina: Mussina is basically Smoltz if Smoltz never blew out his elbow. When we compare him to real-life Smoltz, we’re talking about taking five years of Smoltz as an ace starter and distilling that down to three-and-a-half years of relief work.

I don’t think everyone necessarily will accept that perspective on Mussina, especially because there are a lot of little things working against his public perception that make his value a little harder to see than Smoltz’s. Things like pitching in the AL versus the NL, or Mussina having collectively worse defenses behind him over his career (thanks, again, to New York’s defense completely undoing the benefit Mussina received in Baltimore), which means Mussina’s ERA looks higher on paper. Things like the strike cutting short two of Mussina’s best seasons, which meant, among other things, that when he finally won 20 games in his last season, people were talking about “Mussina never won 20 games before?” rather than potentially his third 20-win season.

I don’t think everyone will accept it, but I do think it is a valid perspective. Not the only one, certainly, but a valid one. And, if you can accept that, it’s a pretty big difference in favour of Mussina. Does Smoltz’s extra postseason success cover that? Does being more interesting make up for being less valuable?

It was a great career from John Smoltz. A Hall-of-Fame-worthy career, and that’s in spite of him getting hurt. But if he hadn’t gotten hurt? He could have been even better. He could have been Mike Mussina.

References & Resources

• Savannah Morning News, “Smoltz’ knuckleball just an experiment”
• Tony Gwynn, ESPN.com, “Smoltz’s repertoire makes him game’s top closer”
• Stephania Bell, ESPN.com, “Bell: This could be it for John Smoltz”
• The essentials in life: FanGraphs, Baseball-Reference, Cot’s Baseball Contracts, Retrosheet and the Lahman Database