clock menu more-arrow no yes mobile

Filed under:

Were the Best Closers Just Failed Starters?

A couple days ago, our friend Rob Neyer wrote a post about Mariano Rivera and his 600 saves. In the article, he references a post by Joe Posnanski where Poz states:

But [Mariano Rivera]’s also a failed starter.

Neyer quite eloquently debunked that thought:

Rivera got 10 starts in the majors, and yes: he struggled as a starter.

In the minors, though?

In 13 triple-A starts, he posted a 3.98 ERA with 53 strikeouts and 13 walks in 61 innings.

In 9 double-A starts, he posted a 2.27 ERA with 39 strikeouts and 8 walks in 63 innings.

In 17 fast-A starts, he posted a 2.25 ERA with 69 strikeouts and 17 walks in 96 innings.

It was an interesting approach—using numbers to show whether or not a pitcher actually was given the opportunity to fail. Neyer insists (and I agree) that Rivera never actually failed in the role. There may have been something about Rivera that made scouts believe he was better suited for closing (like that whole "one pitch" thing and perhaps some injuries). But when you get technical about it, he never actually failed.

A friend and I started looking at other closers to see if the numbers showed they were failed starters. The results were fascinating, so today I give you the top twenty relievers of all time by Baseball-Reference's WAR, and whether or not they were ever given the opportunity to fail as starters.

Mariano Rivera

In 1995 (his first big league season), Rivera posted a 5.51 ERA (84 ERA+) in 19 games (10 starts). He never started another big league game. In the minors, he started 68 of his 103 games. He produced a 2.35 ERA in all 103 games. He relieved in rookie ball (posting a 0.17 ERA!!!), then posted a 2.75 ERA in A ball (29 G, 15 GS). Across two levels in 1993, he posted a 2.08 ERA in 12 game (all starts). Across 3 levels in 1994, he posted a 3.09 ERA in 22 games (all starts). Finally, before getting the call in 1995, he had a 2.10 ERA in 7 starts.

While he didn't shine in his first ten big league starts, that's not a large enough sample size to say he failed. I'm sure not every elite pitcher pitched wonderfully in the first ten starts of his career. Rivera's minor league starts were impressive.

Verdict: Never failed.

Hoyt Wilhelm

Wilhelm started just 52 of his 1070 career games. Interestingly, his most valuable season (by WAR) was 1959. That year he was almost exclusively a starter (32 games, 27 starts) and he was 15-11 while leading the league in ERA (2.19). While his few big league starts were impressive, it's easy to see how he ended up in the bullpen. His minor league game started data isn't complete, but it's easy to see he was almost exclusively a starter in the minors. He posted a 3.55 minor league ERA, but the ERAs were higher the closer he got to the majors (including ERAs of 4.95 and 3.94 in his final two years at AAA). I think it's safe to say he was considered a failed starter and sent to the bullpen.

Verdict: Failed.

Goose Gossage

In Gossage's first 100+ IP season (1975 at age 23), he posted a 1.84 ERA in relief. The next year, he was a starter (29 of 31 games were starts). He posted a 3.94 ERA (91 ERA+). He never started again.

I'm not sure I'd call him a failed starter though. In 54 minor league games (43 were starts), he was 23-11 with a 2.88 ERA. In 1971 (the year before he got the call), he was a 19-year old with an 18-2 record and a 1.83 ERA in 25 games (24 starts). Tough to call that a failure.

Verdict: Tough to say (excelled in the minors, but didn't seize his one extended big league opportunity)

Trevor Hoffman

Hoffman pitched in 1035 big league games. He didn't start any of them. He started just 12 minor league games (but one was in a rehab stint at age 41). In 6 AA starts, he posted a 1.52 ERA. I can't tell how he did in his five AAA starts because his numbers are mixed in with 37 relief appearances (his combined ERA was 4.27). No matter how bad those six starts were, you can't call a guy a failed starter based on six starts.

Verdict: Never failed.

Lee Smith

Smith started just six of his 1022 major league games. That's not a big enough sample to get anything from. However, in the minors he started 83 of his 190 games. He excelled in 10 rookie league starts at age 17. In two years at A ball, he posted a 5.35 ERA (in 26 games, 18 starts) and 4.29 ERA (also in 26 games, 18 starts). In his first year at AA, he posted a 5.98 ERA in 30 games (25 starts). With control being a huge issue (128 BB in 155 IP), the conversion to a reliever began.

Verdict: Failed.

Billy Wagner

Wagner never started a big league game. Wagner started 73 of his 83 minor league games, however. He posted a 3.10 ERA while fanning 501 in 415 innings. He was 11-5 in 25 AAA starts with a 3.23 ERA. That doesn't look like failure to me.

Verdict: Never failed.

John Hiller

Hiller, easily the most underrated relief pitcher in history, was actually a good starter when called upon. In the minors, he started 53 of his 176 game, mostly concentrated into his first two years. In A ball at age 20, he was 14-9 with a 4.03 ERA in 29 games (22 starts). The next year, he posted a 3.46 ERA in 33 games (22 starts). Perhaps his weak record (10-16) triggered the move to the bullpen, because he started just nine more times in the minors (versus 105 relief appearances).

In the majors, Hiller sprinkled 43 starts over his career. I pulled the data for those 43 starts. He threw 279 innings with a 3.10 ERA, allowing 96 walks and 28 homers while fanning 205. He was better as a reliever (career ERA is 2.83), but you can't call those numbers failure. I think Hiller's rubber arm was just too valuable in the bullpen.

Verdict: Never failed.

John Franco

Franco never started in the majors, but he started 55 times in an unimpressive minor league career. Overall, he pitched in 82 minor league games with a 4.80 ERA. His AAA numbers include 23 starts (among his 47 games) with a 5.29 ERA. It's easy to see what happened here.

Verdict: Failed.

Bruce Sutter

Sutter never started in the majors. He started just two of his 116 minor league games. Those games came in a season where he posted a 1.38 ERA in 65 innings.

Verdict: Never failed.

Kent Tekulve

It's horrifying how long Tekulve was kept in the minors. He never started in the majors either, but barely started in the minors. He made 18 minor league starts. In Low A, he threw 9 games (7 starts) with a 1.70 ERA. Six years later, he started five games (his second most in any given season). He had a 1.77 ERA that year. Overall, his minor league ERA was 2.25. He didn't get the call until he was 27, but he still managed to appear in 1050 games.

Verdict: Never failed.

Rollie Fingers

Fingers started 37 games in his big league career. 35 of them happened over a three year stretch. In 1969 (age 22), he appeared in 60 games (8 starts) with a 3.71 ERA (93 ERA+). The following season, he appeared in 45 (19 starts) with a 3.65 ERA (97 ERA+). In the third, still just 24 years old, he pitched 48 games (8 starts) with a 2.99 ERA (113 ERA+). While he pitched well those seasons, his performances as a starter yielded a 4.32 ERA (196 IP, 63 BB, 116 K, 26 HR).

He was an excellent starting pitcher in the minors, going 35-30 with a 2.78 ERA in 83 games (76 starts). While he was given a few opportunities to perform well in the rotation, I'm not convinced he was given enough of an opportunity to cite failure.

Verdict: Tough to say (excelled in the minors, but didn't seize his brief big league opportunities)

Dan Quisenberry

Quiz started one game in his professional career. That came in A ball in his first professional season.

Verdict: Never failed.

Lindy McDaniel

The only six games Lindy McDaniel ever pitched in the minors were starts. He went 4-1 with a 3.64 ERA while averaging seven innings per start. Because he was a bonus baby, those starts actually came in his fourth pro season. As a 19-year old rookie, he pitched in four games (two starts). The next season, he performed well in a swingman role, posting a 3.40 ERA (111 ERA+) in 39 games (7 starts). He started a career high 26 games (of his 30 appearances) as a 21 year old in 1957, going 15-9 with a 3.49 ERA (114 ERA+). The following season, he slumped to 5-7 with a 5.80 ERA (72 ERA+) in 26 games (17 starts). That's the same year he ended up in the minors for a short time.

In 1959, he started seven games, but he also made 55 relief appearances. He posted a 3.82 ERA (111 ERA+) and won 14 games with a league-leading 15 saves. Apparently the Cardinals liked him in that role, because he reprised it the following season with exceptional results (65 games, 2 starts, 12-4 record, league-leading 26 saves). He started just 13 more games in his 15 seasons after that. His combined ERA for all of his starts was 4.65.

Verdict: Failed.

Tom Henke

Henke never started in the major leagues and he only started 18 of his 233 minor league games. He only made one start after posting a 2.93 ERA in 28 A ball games (8 starts) in 1981. I'm not sensing any failure here.

Verdict: Never failed.

Stu Miller

By the time he made the majors, Miller had already started splitting time between the rotation and bullpen in the minors. It is a role that he kept in the majors for the first half of his career. He started 93 games, spread between 1952 and 1960. He made a career high 20 starts in 1958, the same season he led the league in ERA (2.47, 154 ERA+).

In his starts, he posted a 3.79 ERA (compared to 2.94 as a reliever). In fact, in his final eight seasons (the eight in which he didn't make any starts), he posted a 2.72 ERA. He wasn't terrible as a starter, but he was also much better as a reliever—and his managers realized this.

Verdict: Failed.

Francisco Rodríguez

After posting a 5.38 ERA in 20 High A starts at age 19, K-Rod was moved to the bullpen. He posted a 2.27 ERA across two levels and got the call. He never started again.

Verdict: Failed.

Joe Nathan

Nathan was a starter throughout his minor league career (94 starts in 129 games) and, quite frankly, wasn't a very good one. He posted a 4.90 ERA in his minor league career. His first two big league seasons as a starter weren't anything special either. He posted an ERA+ of 103 in his rookie year, then 82 the next. After spending the majority of the next two seasons starting in the minors, he re-emerged in the bigs in 2003 as a reliever. He went 12-4 with a 2.96 ERA. In 2004, he became the closer of the Twins and never looked back.

Verdict: Failed.

Dave Righetti

Righetti started 89 games in the majors, all by his age 24 season. In those four seasons, he was 33-23 with a 3.29 ERA (117 ERA+). He was moved to the bullpen because of a surplus of starters in 1984. He excelled in the closer's role, even setting the saves record with 46 in 1986.

He was also an excellent starter in the minors. In 107 games (99 starts), he was 40-29 with a 3.15 ERA and just 0.5 home runs allowed per nine innings. Given plenty of opportunities, Righetti never failed as a starter. He was just a really good closer, too.

Verdict: Never failed.

Doug Jones

Jones made 107 minor league starts in his career. He started off pretty well, following up an excellent A ball season in 1979 (25 games, 20 starts, 1.75 ERA) with a combined 14-7 record and 2.97 ERA across A ball, AA and AAA in 1980. However, he regressed and bounced between AA and AAA for the next seven years.

The last three of those season were in the Cleveland organization, where he was converted to a reliever. His ERAs dropped to 3.65, then 2.09, then 2.04 before finally making it to the majors to stay at age 29. He started just four of his 846 big league games.

Verdict: Failed.

Bob Stanley

Stanley made 85 starts in his career, most of them at the beginning of his time in the bigs. In his first four seasons, he started 63 of those 85 games (and appeared in 185 games overall) and posted a 3.55 ERA (123 ERA+). He started 20 games in 1987 as a veteran, but didn't fare nearly as well (5.01 ERA, 90 ERA+ in 34 total games). Overall in his starts he posted a 4.40 ERA as a starter. Take out that 1987 season and it drops to 4.20.

While the seasons in which he had some starts weren't typically bad, he did fare much better in a relief role than he did as a starter.

Verdict: Failed.

To me, that's nine pitchers who never actually failed: Rivera, Hoffman, Wagner, Hiller, Sutter, Tekulve, Quisenberry, Henke, and Righetti. On the other hand, nine cleary failed as starters: Wilhelm, Smith, Franco, McDaniel, Miller, Rodriguez, Nathan, Jones, and Stanley. There are a couple pitchers where it's really hard to say: Goose Gossage and Rollie Fingers.

Of course, it's not responsible to say these nine pitchers would have made great starting pitchers. Scouts are smart. They probably rescued some of these guys from situations where they felt they would fail. Some, however (like Righetti and Hiller), were just so good in their relief roles that the team probably just didn't want to mess with a good thing (think Jonathan Papelbon for a more modern example).

It's also not responsible to say all closers are failed starters (and I've heard that a lot). In my opinion, you can't call someone a failure if they were never given the opportunity to fail.