Death of the Long Man
Over the last 40 seasons, there have been 752 player seasons where a reliever pitched at least eighty innings out of the bullpen and averaged at least 1 1/3 innings pitched per appearances. Last year was the first and only season of the last 40 where not a single player met that criteria. Increased reliever specialization and larger bullpens have minimized the long reliever, and those who have been given the long reliever role tend to be the low man in the bullpen hierarchy. That was not always the case, and the decreased offensive environment could be a good opportunity to reintroduce the good long reliever to baseball.
In the not too distant past, long relievers were a regular fixture on teams. Relievers making regular appearances longer than two innings has always been a rarity, but some teams had relievers truly earning the the title of long relievers. From 1975-2014, just 110 relievers pitched at least 80 innings and averaged more than two innings per appearance, but those seasons have all but disappeared in the last two decades.
Strike seasons of 1981 and 1994 are omitted
Only knuckle-baller Steve Sparks and Minnesota Twins pitcher Anthony Swarzak have had reliever seasons totaling more than 80 innings averaging at least two innings per appearance in the last two decades. Expecting relievers to average more than two innings per start when it was never really prevalent is unrealistic. However, setting the target a little lower brings forward a role that was common in prior years. Here are the players per year with at least 80 innings and an average of 1 1/3 innings pitched per appearance.
Teams regularly used at least one reliever who would appear consistently and go at least an inning. The numbers dropped to roughly half the teams during the 90s, and dropped even lower over the past decade. Last season, no reliever averaged more than 1 1/3 innings and pitched at least 80 innings. The swingman role, where a pitcher might move back and forth between the bullpen and the rotation is another effective strategy to get better pitchers more innings, but it is not common either. Over the last five years, just 22 players have made at least 40 appearances,more than five starts, and totaled more than 80 innings pitched. The swingman is barely alive, and the effective long reliever appears to be dead.
The long man at one point time was a solid player, around league average, but the quality of players used in that role has declined over time. As the number of long relievers has declined, so has their effectiveness. The chart below shows the average number of innings by long relievers (relievers averaging more than 1 1/3 IP), the average WAR, and the average number of innings per win.
| WAR | IP | IP/WAR | |
|---|---|---|---|
| 1975-1984 | 0.96 | 100 | 104 |
| 1985-1994 | 0.93 | 96 | 103 |
| 1995-2004 | 0.73 | 92 | 126 |
| 2005-2014 | 0.59 | 86 | 146 |
Up through the strike, the pitchers utilized in a long reliever role were roughly average, but in the last 20 years their performance has declined. Pitchers are rarely asked to get important outs and pitch more than one inning, decreasing their high leverage situations and their importance to the team. Using a good pitcher, perhaps someone who just missed out on the rotation, should provide the manager with more opportunities to give a long reliever important outs.
An alternative to taking a player nearly starter-worthy and putting him in the bullpen is the six-man rotation. While it does have its proponents, there are several problems with its implementation. First, it removes established veterans from their routine and potentially gives them fewer starts. Even if only used for lesser starters, it still asks pitchers to make outings at least as long they are currently making, potentially leaving the bullpen short a man and exposing them to hitters more than two times per game.
Piggybacking a starter could work, allowing all players to work on normal rest, and have one pseudo-starter relieve one of the starters for multiple innings. Unfortunately, that plan still leaves the bullpen short a player for most games. Getting a multi-inning performance twice a week eases some of the pressure on the rest of the bullpen. Given an average of three innings per night for the bullpen and 6.2 games per week over the 26-week season, the bullpen needs to cover around 19 innings per week. Giving one good reliever even 3 1/3 innings per week in two of those games ends up with one of a team’s better relievers pitching 90 innings a season in higher leverage situations than the modern long man. The remaining six members of the pen need to average roughly 66 innings per season, making five one-inning appearances every two weeks.
Using a potential starter as long reliever also serves to keep that reliever more stretched out in case the need for a starter should arise without having to sacrifice his innings to the minor leagues. For organizations worried about increasing innings loads for young pitchers, using a pitcher as a long reliever ensures that the innings jump moving from reliever to starter is lessened. As starters feel the drop in effectiveness as they face the third time through the order penalty, a good long reliever can provide important, valuable innings bridging the gap to the late innings. Not every organization is equipped to use a reliever in this way. Looking at the FanGraphs Depth Charts for pitchers who might not win a starting spot out of Spring Training, but have good projections as starters reveals potential long reliever candidates. The list is not exhaustive and some players could still be starters come April, but here a few candidates for long reliever roles.
| K/9 | BB/9 | HR/9 | BABIP | LOB% | ERA | FIP | |
|---|---|---|---|---|---|---|---|
| Tanner Roark | 6.5 | 1.9 | 0.9 | 0.295 | 73.1 % | 3.47 | 3.67 |
| Marco Gonzales | 7.8 | 3.0 | 0.9 | 0.298 | 74.0 % | 3.62 | 3.84 |
| Adam Warren | 8.1 | 3.1 | 0.9 | 0.298 | 73.3 % | 3.67 | 3.79 |
| David Phelps | 7.7 | 3.4 | 0.8 | 0.304 | 72.0 % | 3.82 | 3.83 |
| Yusmeiro Petit | 8.6 | 2.0 | 0.9 | 0.300 | 73.9 % | 3.32 | 3.30 |
| Chase Anderson | 8.0 | 2.8 | 1.1 | 0.304 | 72.8 % | 3.92 | 3.95 |
Roark is getting pushed out of the rotation through not fault of his own. After signing Max Scherzer, the Washington Nationals have a starting five that is tough to crack. Roark could find a home as a set-up man, but his lack of strikeout stuff would be unusual for that role. Using him more innings at a time would keep him ready to start and get him more innings for the Nationals. Marco Gonzales could also find himself outside of the rotation for the St. Louis Caridnals. If Jaime Garcia manages to stay healthy for another few weeks and wins a rotation spot, the Cardinals might have two candidates in Gonzales and Carlos Martinez. With a rotation full of age and injury question marks, a multiple-inning reliever could be very useful for the team. Gonzales pitched out of the bullpen in the playoffs last season and getting him multiple innings at a time at the big league level could help one of the Cardinals’ top prospects develop while still contributing to the major league team.
Petit pitched well as a swingman last season, and Phelps could further feel the squeeze when Jose Fernandez returns making both players good candidates to pitch in long relief. The good long reliever has virtually disappeared from the majors, replaced by the mop-up man at the end of the bullpen. Using potential starters as long relievers helps the MLB team and spurs development for young players. The long reliever might be gone, but hopefully it has not been forgotten.
Craig Edwards can be found on twitter @craigjedwards.
Pirates are dealing with this issue now with Locke and Worley Fightng for the last rotation spot and both of them out of options
I think this kind of misses the point. The long reliever role is largely dead because there’s no real need for it. In a close game, having a fresh guy every inning, or playing the matchups is likely more effective. In one-sided games, if you’re winning by a lot, your starter is usually doing well, so you won’t need too many innings, and a mop-up guy can do the work anyway. If you’re miles behind, it’s not worth using one of your better arms, so, again, the mop-up guy is what you need. Obviously that’s a simplification, and there will be some occasions where a competent long man is useful, but, basically teams don’t use them because they are a poor use of resources.
I would have thought that a longman could be more useful in preventing the best relievers from getting overworked over the course of the season from making multiple appearances. I would think having a pitcher that can go 2-3 innings once per 3-4 games or so would go a long way towards protecting short relievers.
I would add Miguel Gonzalez as a potential candidate to play this role this season. The Orioles are unlikely to cut bait on Ubaldo Jimenez in spite of his struggles, and we could see Gonzalez play the swingman role he was quite good at when he first came up.
I disagree. There is only no need for a long man because teams carry 12 or 13 relievers. If teams carried 11 relievers, they could use a long man to eat up those extra innings. That frees up an extra hitter or two for the bench. More pinch hitting opportunities and more platoon opportunities. You’re not giving up your two best relievers to fill those spots. You’re giving up your two worst. In my opinion, it’d be worth the trade.
No one has 12 relievers – A bullpen is 7 or 8.
Using a long man could give you the benefit of saving your other guys for other/better opportunities, but that would be assuming that the long man is worth the innings that you’re giving him. That might be a hard guy to find.
I really wonder if the “poor use of resources” bit is true. The death of the long man basically co-occured with an increasing offensive environment. So the 1IP or less reliever usage only and 12 man bullpens was probably born out of need to adapt to that particular environment. Now we’re seeing the lowest run scoring levels in a couple decades, so a new adaptation may be more effective. Runs in general are so much harder to come by that once you have 2-3 run lead its now a safer lead than it was 10 years ago. So, it might make more sense to throw the same guy out there for 2-3 innings and save your bullpen aces for a tied/1-run game that might happen tomorrow. And in the process you get an extra roster spot, even 2, to draw from for platoons or pinch hitters to help gain back some of that offense that is so hard to come by today.
Where are you guys getting “12 or 13 relievers” and “12 man bullpens” from? Watching too much spring training?
I think they mean 12 man pitching staffs
He CLEARLY means 12 man pitching staffs dude. Pretty obvious. Sigh.
I would love to see another Mike Marshall or Kent Tekulve in the game. Funs times.
Every reliever these days seem to throw 95 mph or more. With such an emphasis on velocity and strikeouts, it may just be that pitchers (other than elite starters) can only maintain such velocity for an inning at a time.
Side note – Yusmeiro Petit is a very good pitcher. His combination of a high K%, low BB% and reasonable HR% is really impressive. He does all this without extreme velocity. My guess is that he will start (his competition is zombie Matt Cain, zombie Jake Peavy, zombie Big Time Timmy Jim, zombie Tim Hudson and regular Ryan Vogelsong. Petit cannot possibly be the 6th best pitcher in this crowd of has-beens.
Petit is fun to watch. I’m not sure I would throw Cain and Peavy under the bus yet.
“…and regular Ryan Vogelsong.”
Made my morning!
Todd Redmond was very close… 75 innings all out of the pen over 42 games.
Craig Stammen is another “close” candidate. 2012-2014, IP/G: 88.1/59, 81.2/55, 72.2/49
I wonder how much of this is due to expansion.
You said that “The good long reliever has virtually disappeared from the majors, replaced by the mop-up man at the end of the bullpen.” This is a little confusing. Why wouldn’t the mop-up man qualify as a long reliever? Isn’t the mop-up man averaging more than 1 1/3 innings? Are his opportunities to pitch fewer because games are closer and home runs more frequent (thereby making even blow-outs less out of reach)?
When you began by talking about the death of the long man, I thought that this would have included the mop-up duty pitcher. Am I wrong?
The mopup man is usually the worst reliever on the team. I’m advocating for a better pitcher to be used in a long reliever role different from the current iteration. That’s what I meant by “good” long reliever.
Why did you choose 80 IP? I suspect the underlying issue is more about the reduction of average IP by all pitchers rather than a particular role being eliminated.
You could do a similar article about the death of the workhorse (defined as a 250 IP starting pitcher).
Fair, but the 250 IP pitcher is basically non existent while around 30 relievers have gone over 80IP in the last five years. I suspect more could if they had longer outings over fewer appearances.
Good article, I agree it seems like relievers are turning into 1 inning specimans, in most cases not not even allowed to come back for the next inning. In September we get to watch Match-up-Pallooza with all the pitching changes the expanded rosters allow. The 2 inning relief appearance has mostly died except in playoffs and extra innings. Mariano Rivera pitched 60+ games and 100+ innings in relief early on and he seemed to have an OK career, and what a weapon that gives a manager.
I entirely doubt the entire Nationals rotation will be able to pitch every game they are scheduled. Having a guy like Roark back there (that could be a starter on most teams) has to be better than throwing out your Jason Marquis-Roberto Hernandez types.
And, why you wouldn’t use a 6 man rotation (right handed) against a team that has all righties, for example, or a 3 game series with a team with a lot of lefties, still seems like a strong market efficiency if people like matchups so much. (as they should)
roark will get some starts… sucks to be blake treinen, though… he’d be a starter on at least a dozen teams, too…
You can thank the Dbacks for having not a single pitcher better than Josh Collmenter as rotation options.
I’m surprised Betances didn’t qualify, he made over 90 innings in 70 appearances which averages out to 1.29 innings; might as well give it to him because that is like one inning shy of averaging 1 and a third for a season?
Did you eliminate spot starters or somehow filter out for Innings pitched in relief?
As a Mets fan, my immediate thought was Darren Oliver in 2006. I figured he had to have had one of the last good “long man” seasons.
I pulled pitchers with your stats since 2005 from Baseball Reference, put a cap on starts at 4, filtered out War<1 and sorted by WAR. There were a few pitchers' seasons that stood out:
2011 Alfredo Aceves 2.7 WAR
*2008 Hung-Chih Kuo 2.6 WAR
2012 Craig Stammen 2.0 WAR
Aceves and Kuo both recorded multiple starts, but were awful in them (5.14 and 5.06 ERA respectively). Kuo's innings as a reliever were under the 80 inn total (69.1), so he gets and asterisk, but still outstanding relief stuff. Aceves pitched 90+ innings in relief, so he gets bonus props for that.
Darren Oliver was 11th of 12 on the list with a 1.1 WAR, but 3rd if you eliminate the spot starters, so I was not imagining it when I thought that he really did have one of the last good "long man" seasons.
It would be a “poor use of resources” to intentionally carve a roster-spot for a mop-up man, but it seems more like the conventional wisdom was that a 25-man roster should have, in a pitching staff of 11-13, at least one guy who can (a) eat low-leverage innings in relief and (b) make a spot start, which can be acquired for a player earning at or near league minimum. In today’s environment, there are constant transactions, and teams seem to keep lots of players with options and lots of AAAA players, and roster competition that goes deep into the season. So if a starter gets chased early, instead of looking to a dedicated “mop-up” or “long guy”, managers will use the opportunity to let guys who are competing for more, higher-leverage, innings get a chance to sharpen up, and demonstrate their value. If they get overused, or a spot starter is needed, someone who can be optioned gets sent down, to make room, or someone who hasn’t been used in a while finds his way to the retroactive DL. Having a mop-up man means taking away a revolving door spot that can be used to keep the set-up men sharp. I may be biased by how my homer club does things, but it sure looks like the people who have the set of skills, and the value to be “long men” often are AAAA players. Frankly, I was happy when my home team abandoned the practice. Terry Mulholland coming out of the bullpen was about the last thing I ever wanted to watch. And, speaking of ole Terry, the author makes another great point: the long man role is inherently self-sabotaging. Few players relish the role. Most of the time if a team has a guy who only gets used in low-leverage innings, it’s because that guy has lost the confidence of his manager, and the GM hasn’t quite caught up to making a move… and not because there’s a mutual coincidence of the team and player wanting to be in that role.
There is a class of pitcher that can excel being used once-though-a-lineup and I wouldn’t call him a mop-up man. Think of all of the failed starters turned specialist
(or relegated to AAAA status) because they couldn’t quite go seven innings, or twice through an opposing line-up. Teams are giving up quality innings by refusing to let pitchers be good for 50 pitches at a time. It must be 100 pitches or 25 pitches or whatever it takes to get one hitter out.
And one or two of those type of guys combined with your 4th/5th starters would probably mean less runs scored in innings 1-7 of games started by the back of your rotation. Instead, teams throw the 4th/5th guy out there, basically wait until he fails, then call on a sub-par guy from the bullpen because the team is now behind in the 5th inning.
I’d love to see a comeback of the long reliever. I’m somewhat dubious this will happen in the short-term.
First, most of the candidates for the long reliever role are going to either be (a) veterans trying to hold on and (b) young players trying to break in. The former tend to be more of the mop-up model. If they were still decent starters they would not accept the role and/or would be traded for something more useful. The latter have the problem of free agency out on the horizon. Even a below average starter has quite a bit of value and to use up a year of team control for a lower value relief role is not something that many teams can risk. Rich teams can afford the luxury – the Yankees seem to come up with someone in this vein every couple of years – but they also can fix the problem by throwing money at better starters.
Second, I think the problem is less about having long relievers than limiting existing relievers to the one inning or less workload. Teams have become obsessed with not using relievers for more than an inning. Does this really make sense? We know that at least some relievers like Betances can be very effective over multiple innings. We know that relief aces in the past could be very effective over multiple innings. We know that relief aces can be very effective over multiple innings in the playoffs. If relief pitchers, especially relief aces, are used for multiple innings at a time without significant decline in effectiveness that would change bullpens a whole lot. Even if the the pen is, say, 90% as effective when used over multiple innings, that probably is more than sufficient to get rid of the last guy or two in the pen and replace them with useful bench players. If you want your revolution, that’s probably where it will be.
Of course, if bullpens shrink that would make a long reliever more useful as to not use up the best relievers in low leverage situations.