Billy Wagner and the Myth of the Failed Starter
You can't assume a starting pitcher could equal Wagner as a reliever. Just ask John Smoltz.
When the 2025 Baseball Hall of Fame ballot was unveiled last week, one name in particular looked very familiar: Billy Wagner. It’s his tenth and final try to get in via the Baseball Writers’ Association of America’s annual vote. His support has more than septupled in that time, from 10% in 2016 to a tantalizing 74% in 2024 — I maintain it’s strange that a supermajority of Cooperstown’s expert electors think Wagner somehow got worthier several years into retirement — and this is his last shot to reach the 75% required for baseball immortality.
If you’re reading this, you probably know some of Wagner’s bona fides, but even many baseball fans don’t realize just how dominant he was. His 33% strikeout rate is the highest among the nearly 1,500 Major League players with at least as many innings pitched, and it isn’t particularly close. Ditto his .184 batting average against, in which his 16-point lead over runner-up Nolan Ryan is equal to the gap between second place and 20th. He pitched in 16 big-league seasons and had only one year with an ERA above three. (Comparing that to the three relievers who’ve been inducted in the last 15 years: Mariano Rivera had two. Trevor Hoffman had five. Lee Smith had 12.) And he posted these gaudy numbers in the height of the high-scoring Steroid Era, and while pitching left-handed despite being a natural righty.
Wagner’s heretofore exclusion from the Hall of Fame can be explained by two words: “failed starter.” It is universally understood that it is easier to throw one inning at a time than to maintain effectiveness deep into a game — hence what happened in this year’s postseason — and that a pitcher who can throw 180 innings a year is more valuable to a roster than one who can soak up 60. That means a pitcher coming out of the bullpen usually either failed to hold down a rotation spot or was never considered good enough to try. From this kernel of truth has sprung a mindset, shared by a horseshoe-theory alliance of old-school cranks who think elbow soreness builds character and statheads for whom player performance is merely a synonym for context-adjusted value,1 that presumes relievers’ inadequacy and does not recognize their greatness on its own terms.
The “failed starter” meme has created a uniquely and near-impossibly high Hall of Fame standard for the most-common position in baseball. Albert Pujols, who spent his early career tumbling down the defensive spectrum, was not derided as a “failed third baseman.” Voters did not dismiss Roberto Alomar’s 10 Gold Gloves by saying second base is for those who flunked off of shortstop. Ted Williams’ legacy is not tainted by his playing the easiest of the three outfield positions. It’s particularly frustrating when applied to Wagner, who didn’t actually fail as a starter! He had a 3.20 ERA in four years in the minors while pitching exclusively out of the rotation. Had he not proven his prowess out of the bullpen so quickly, perhaps the Houston Astros would have let him finish developing as a starter.
Unfortunately, it’s a hard talking point to rebut. Wagner’s skeptics would tell you that it doesn’t take an elite talent to pitch like he did as a reliever. I say he performed at damn near the upper limit of plausible sustained dominance in any role. We find ourselves at a rhetorical impasse. In order to prove it either way, you’d need to take a Cooperstown-quality starting pitcher from the same era, move them to the bullpen in the middle of his career, and see how their numbers compare to Wagner’s. There are only two ways to gain access to such a natural experiment. One is to get a time machine. The other is to look at John Smoltz.
Long before he found his true calling as a baseball broadcaster who openly despises the modern sport, John Smoltz was a bona fide ace. For an 11-season stretch from 1989 to 1999 he averaged 14 wins a year with a 3.29 ERA, earning four All-Star nods and a Cy Young Award while ranking in the top five for innings pitched and strikeouts. Then Tommy John surgery cost Smoltz the 2000 season and he struggled in his return the following year. So in the summer of 2001, he moved to the bullpen.
Smoltz took to the role with aplomb. He won Atlanta’s closer job within a month. He posted a 1.54 ERA in 36 appearances as a reliever through the rest of the season and the playoffs. In 2002 he led the Majors with 55 saves and finished third in the NL Cy Young voting.2 Over just four years in the bullpen he was named an All-Star twice and received MVP votes thrice. He returned to the rotation in 2005 and immediately had another All-Star season. (There’s a reason this guy is in the Hall of Fame.)
So how does Wagner’s performance stack up against a Cooperstown-worthy starting pitcher who took on and excelled in the same, easier role? For simplicity’s sake, I will focus on four statistics that are key to Wagner’s résumé: ERA, the most-universal indicator of run-prevention; strikeout rate and batting average against, good shorthands for how tough of an at-bat a pitcher is; and ERA-, which adjusts ERA for how the run environment varies over time and across ballparks (average is 100, lower is better). These metrics all have something in common: among the over-1,100 pitchers with at least as many innings pitched in the last century, Wagner either ranks at the top (K%, BAA) or he trails only Mariano Rivera (ERA, ERA-).
From 2001-04, when both Smoltz and Wagner were bullpen contemporaries, here’s how they each performed:3
John Smoltz: 260.1 IP, 2.35 ERA, 28% K%, .211 BAA, 56 ERA-
Billy Wagner: 272.0 IP, 2.32 ERA, 31% K%, .183 BAA, 53 ERA-
When a first-ballot Hall of Fame starting pitcher spent four years in the bullpen, the result was that he pitched almost as well as “failed starter” Billy Wagner.
While it’s effectively a tie in ERA, I’d argue Wagner’s line was more special. Smoltz’s 28% strikeout rate is quite good, especially for 20 years ago, but it’s a clear step down from Wagner’s 31%. To the extent that we care about opponent’s batting average, 28 points is a large gap. And ERA- reveals that Wagner pitched in hitter-friendlier parks, making his numbers all the more impressive.
Maybe comparing the two in that timeframe isn’t fair. Smoltz is four years older than Wagner. He was 34 when he made his big-league bullpen debut, while Wagner was about to turn 30. If you compare Smoltz’ relief numbers to Wagner’s in his seasons at the most-similar age range, 2005-08, the gaps in strikeout rate and batting average get a little narrower. But this time Wagner’s topline results are plainly better:
John Smoltz: 260.1 IP, 2.35 ERA, 28% K%, .211 BAA, 56 ERA-
Billy Wagner: 265.1 IP, 2.14 ERA, 30% K%, .195 BAA, 50 ERA-
Before anyone checks my math, if you go to your favorite statistics website, you’ll see that Smoltz’ 2001-04 seasons encompass ages 34-37, while the above years for Wagner are listed at ages 33-36. This is due to a quirk in how seasonal age is conventionally calculated; taking the years Wagner was listed at 34-37 effectively penalizes him by a year in terms of aging curves.4 But lest I be accused of cherry-picking, here’s how they compare during their nominally same-age years:
John Smoltz: 260.1 IP, 2.35 ERA, 28% K%, .211 BAA, 56 ERA-
Billy Wagner: 203.1 IP, 2.35 ERA, 31% K%, .203 BAA, 54 ERA-
Wagner threw a lot fewer innings, as he spent most of his ostensible age-37 season recovering from Tommy John surgery, and didn’t even pitch until after he turned 38. (Miraculously, he returned to the mound less than a year after the notoriously difficult procedure and posted a 1.72 ERA down the stretch — mostly for the Boston Red Sox, who acquired him a week after they released Smoltz.) But in terms of performance, their ERAs were exactly the same, with Wagner again looking better when considering secondary skills and the respective run environments.
Finally, to confirm that we’re not just looking at an unrepresentative slice of Wagner’s career, here are each’s overall regular-season relief numbers, including the entirety of Wagner’s 16-year career and Smoltz’ brief return to the bullpen in 2008:
John Smoltz: 261.1 IP, 2.41 ERA, 28% K%, .213 BAA, 57 ERA-
Billy Wagner: 903.0 IP, 2.31 ERA, 33% K%, .184 BAA, 54 ERA-
If one of the consensus-greatest pitchers of his generation couldn’t match Billy Wagner’s level of dominance, let alone exceed it, who’s to say any starter could?
There are two broad ways to interpret this comparison. The first, which I subscribe to, is that it doesn’t mean much. It’s a fun piece of trivia and a nice feather in Wagner’s cap, but we know that translating pitching performance from the rotation to the bullpen isn’t linear. It’s more like moving from second base to left field: sure, it’s easier, but you can’t assume that the average middle infielder would win a Gold Glove if they moved to a corner. Though if we can agree to evaluate relievers on their own merits, having a claim as the toughest at-bat in MLB history and boasting a combination of dominance and longevity that only Mariano Rivera has inarguably surpassed ought to have made Wagner a shoo-in.
But if you see even elite relievers as strictly inferior to and more replaceable than starters — a view that is at least implicit in most arguments I’ve heard against Wagner, often couched in questions of value that make far more sense in contemporary roster and transaction analysis than reflecting on an athlete’s legacy — this case study is important. If working out of the bullpen is so predictably easier, isn’t it a huge statement about Wagner’s deservingness that he outpitched a first-ballot Hall-of-Famer in the same role? And doesn’t that equivalency go both ways? Under such a rigid view of pitcher roles, since he was a better reliever than Smoltz, we should also expect him to be a better starter than Smoltz. Why wouldn’t we assume that Wagner would have been on a clear Cooperstown track had he remained in the rotation?
I’m bullish on Wagner’s chances this year, even amid a stacked class of new candidates competing for voters’ attention. No one who’s come as close as Wagner has ever failed to gain induction one way or the other. He has momentum, which for some reason matters for an athlete who’s been retired for 14 years, and holdovers usually get a boost in their last year of eligibility. I suspect many Wagner skeptics don’t want to be the reason why he falls short. If he flips the voters who checked his name in 2022 or 2023 but not 2024 or said they would have picked him last year were it not for the 10-vote limit, that alone would get him over the hump. But it’s a pock on the sport that a player as special as Wagner wasn’t ushered in on his first ballot. As we prepare for his hopeful enshrinement next summer, and look ahead as the next generation of elite relievers wrap up their careers, this dismissive attitude towards greatness out of the bullpen ought to be a non-starter.
I say this as someone who has done it professionally.
Finishing just behind Smoltz in fourth place, another newly converted closer who was starting to make a name for himself in his new role: Eric Gagné.
Not counting Smoltz’ appearances as a starter in 2001, which were considerably worse.
Seasonal age is typically based on how old the player was on July 1, the symbolic midpoint of the regular season. Smoltz’ birthday is May 15, so this method reflects his post-birthday age. Wagner was born on July 25, so he is listed as his pre-birthday age. Wagner was therefore 10 months older than Smoltz for any pair of seasons in which they are listed as the same age, making Wagner’s age-minus-one year a better comparison.
Great article and analysis! I never felt strongly either way, mainly because I didn’t know much about his candidacy or career stats, but you have persuaded me that he was very deserving of prior if not first ballot election!
That’s always been my unhappiness with voting such as this, that there were obvious players who deserved first ballot election and yet there are idiots who would not vote the player in (for me, how could Willie Mays not be unanimous first ballot? Plus, I’m of the belief that Bobby Bonds deserved to be in the HOF as well, though I understand the negatives).
The baseball Hall of Fame is an odd thing, with voters incentivized to make odd choices in an effort to game the system, instead of just voting for who they think is a Hall of Fame player. The football Hall of Fame gives their voters infinite votes (yes or no on all candidates), and the world has not exploded. Baseball can surely do the same thing, and it'd help people like Billy Wagner.
It must be delineated that it's not the small sample size that causes baseball writers to discredit relief pitchers. If this were the argument (that no pitcher with 900 IP is Cooperstown worthy), it's possible I could get behind it. However, and ironically, all the talk around Billy Wagner exposes that of being quite false, because of the choice of many of his critics to hyper focus on about 12 of his career innings. The postseason ones. Quite frankly, I don't care about 12 inning sample sizes, but I have a feeling this impossibly small sample is what's held Billy out of the Hall for so long. Since evidently the writers care a lot about small sample baseball, they ideologically forfeit their right to hold out on voting for Billy simply because he's a relief pitcher.
What you address in this article is the perception (not fact, but perception) that starters would be better at being relief pitchers than the relief pitchers, and at least in the case of John Smoltz, prove it to be untrue. I'm a big fan of analyses like this, because although you cannot truly prove anything with a single natural experiment like this, it does plant the seed that a Hall of Fame starter being allowed to use full effort for every single inning they pitch like relievers do does not automatically put them at the very top of the RP ranks. It's important to know things like this, because as we advance further and further into baseball history, relief pitchers (the important ones) are going to become a bigger and bigger part of the story, and the myth of the failed starter will either cast a shadow over the entire sport, or it will die.
I'm in the camp of helping this myth die, for the good of the sport, because who wants to continuously hear that the nastiest pitchers in the world are actually kind of mid, and couldn't make it at the more valuable position? That doesn't help anybody, almost certainly doing much more harm for the sport than good. You've gone a great way towards combatting this failed starter myth with me, and for that I thank you.
Great stuff here!