[h=1]Where have all the good relievers gone?[/h]
ESPN PLUS ($ MATERIAL)
Does baseball have a supply-and-demand issue when it comes to relief pitching? If so, what's behind it?
You wouldn't think this would be a possibility. For the past few years, as more and more innings have shifted from starting pitchers to relievers, we've heard about waves and waves of thunder-throwing young pitchers who have rendered the old models of pitching staff construction as obsolete as the rotary phone.
Yet there is a finite supply of everything in this corporeal existence of ours, and the consumption habits of humans being what they are, you had to figure we'd rub up against the outer limits of fungible -- but effective -- relief pitching. But these trends are so new and have accelerated so quickly that if the supposition presented in the first sentence above were to have some substance, it would be pretty surprising.
It's early. Every analysis piece with an April publish date must wear that caveat around its neck. Through Wednesday, we had played just under 15 percent of the 2019 season. That's right! We get to do this all 6.67 more times. Everything suggested here could be rubbed out by reality's eraser.
Nevertheless, here are some numbers comparing starting pitchers to relief pitchers in each of the past 10 seasons, with each season measured through the end of April.
<offer style="box-sizing: border-box;"></offer>
<aside class="inline inline-table" style="box-sizing: border-box; font-family: -apple-system, BlinkMacSystemFont, Roboto, Arial, "Helvetica Neue", Helvetica, sans-serif; height: 407px; border: 1px solid rgb(220, 221, 223); clear: both; margin: 6px 0px 18px; padding: 15px; width: 565px; color: rgb(0, 0, 0); font-size: 16px;">[h=2]Year-By-Year ERAs Through The End Of April[/h]
<thead style="box-sizing: border-box;">
</thead><tbody style="box-sizing: border-box;">
</tbody><tfoot style="box-sizing: border-box;">
</tfoot></aside>
What you should notice is that bullpen ERAs are generally better than starter ERAs, something that manifests in the opening weeks of the season and remains true through each campaign. So far in 2019, however, the starters are winning. In fact, it's the second time in three years that a season has started this way.
By the end of 2017, the relievers caught and surpassed the starters in ERA. That might be happening in 2019 already, as the gap has closed the past couple of weeks. But it's no secret that bullpens so far in 2019 have struggled across the big leagues. Teams have been forced to juggle roles and roster spots in an attempt to find the right mixes. Some have made progress; others haven't but will keep trying.
The question here: How long can clubs keep playing this shell game?
Let's consider some year-by-year reliever data downloaded from fangraphs.com. We'll work with fWAR for our purposes here.
In 2015, 97 big league relievers posted an fWAR of at least 0.5 -- a mere half-win -- for a combined fWAR of 102.5. The very next season, 93 of those pitchers logged big league innings. However, those 93 pitchers combined for just 55.5 fWAR, a drop of nearly 45 percent in value.
The trend only gets worse as the seasons go by. In 2017, those 93 pitchers were down to 81, and the fWAR was down to 45.2. Last season, it was 73 surviving relievers putting up 35.5 fWAR. So far this season, that original group of 93 is on pace for just 16.0 fWAR. Over the span of five seasons, this group of good relievers will have lost 84 percent of their collective value.
That is a lot of production to replace in a time frame that is less than the six-year window of controllable seasons for a player via baseball's service time system. And this rate of attrition might be accelerating.
Let's use 2016 as our starting point. That season saw 114 relievers with at least 0.5 fWAR. By the end of last season, those pitchers had lost 55 percent of their value. How about if we start with 2017, when there were also 114 relievers at 0.5 or better? Those hurlers, by the end of last season, had ceded 56 percent of their value -- from 131.0 fWAR to 57.8. So far this season, they are on pace for 38.0.
Last season, we saw 116 relievers reach 0.5 fWAR, with a combined total of 124.5. Those 116 pitchers are on pace to hit 59.4 fWAR. That's more than half the value of last season's best relievers gone in the span of part of one autumn, a long winter and a bit of spring.
What's going on here? To know, we'd have to burrow in a lot deeper than we will today. But I wonder if we're starting to see the makings of a supply problem. It's not just that the attrition rate of relievers is so high; it's that their production depreciates at an even faster rate. Teams bring back a lot of the same names, but their stat lines don't necessarily come with them.
We've known these things for a long time -- that reliever performance is difficult to predict. The trends toward bullpen domination have not been built on certainty at the individual level because it barely exists. Heck, in 2017, 13 relievers posted 2.0 fWAR or better for a combined total of 32.3 fWAR. That benchmark -- 2.0 fWAR -- is elite stuff for a relief pitcher. In 2018, those 13 bullpen aces combined for 9.2 fWAR. In 2019, they're on pace for 5.3.
So, yeah, individual relievers are flaky as heck, and that's why teams are so careful about overpaying for them. That's arguably why the best reliever of the 2010s, Craig Kimbrel, is currently unemployed. But teams have dealt with this extreme variability by cycling through arms like a long roll of lottery tickets. Eventually, good teams can hit upon enough positive producers to get through a season.
This season might be no different. Teams certainly show no signs of shifting bullpen innings back to the rotation, though that might change in 2020, when we begin the era of three-batter minimums. But you have to wonder if there will be a limit to how many competent pitchers teams will be able to burn through.
You also have to wonder how fair all of this is to relief pitchers. Relievers are promoted and asked to max out in terms of effort, spin and velocity. Then they will almost invariably burn out before they reach a level of service time that translates into big money.
But that's how this contemporary model of pitching staff construction is fueled -- on an endless supply of young power arms. What happens if that supply is not truly endless, if the burn-out rate surpasses the replacement rate? How will this model hold up when there aren't enough quality arms to cover for starting pitchers collectively asked to throw around five innings on average?
If we reach that point, then the model is broken. All things considered, that might not be the worst thing for the game -- or for the pitchers.
[h=3]Extra innings[/h]
1. You've probably seen this number going around: Cody Bellinger's double on Thursday at Wrigley Field moved him within one of tying former teammate Chase Utley for the most total bases by any player by the end of April. Bellinger had 84 total bases after that two-bagger against the Cubs.
Teams used to open their seasons later than they have in recent years, so all the leaders on the early-season total-base chart are recent. It's a good list to be on, and it might be predictive of postseason honors. Here are the leaders, other than Bellinger, with their total bases at the end of April and where they finished in the MVP voting:
Chase Utley, 2008 (85 total bases, 14th in NL MVP)
Jermaine Dye, 2000 (83, no MVP votes)
Alex Rodriguez, 2007 (82, won AL MVP)
Larry Walker, 1997 (82, won NL MVP)
Alfonso Soriano, 2003 (81, 20th in AL MVP voting)
Ken Griffey Jr., 1997 (81, won AL MVP)
Also, if you haven't noticed, Bellinger leads the majors with a ridiculous .426 batting average. His average on balls in play is .397. Of course, these numbers won't last. They can't. But ...
According to baseball-reference.com, which has splits of this type mostly complete back to 1908, Bellinger would become the 14th hitter to post a .400 average through April over at least 100 plate appearances. (Again, because teams start earlier these days, the leaderboard is titled toward recent decades.) In fact, .426 ranks fifth on that list.
Best March/April batting averages
Minimum: 100 plate appearances
1. Larry Walker (.456, 1997)
2. Darin Erstad (.449, 2000)
3. Ivan Rodriguez (.442, 1998)
4. Hal Morris (.427, 1998)
5. Cody Bellinger (.426, 2019 through Thursday)
2. Generally speaking, strength of schedule isn't a huge factor at the team level by the end of the season. Teams competing for the same thing -- division titles, namely -- pretty much play the same schedules as their competitors. Wild-card races can be a little skewed, but it's not a huge thing. However, early in the season, the standings can be greatly affected by scheduling quirks.
Which brings me to the St. Louis Cardinals. The Redbirds currently rank second in the majors in run differential, behind the Tampa Bay Rays. However, St. Louis has compiled that result against the fourth-toughest schedule in baseball, according to baseball-reference.com. Early as it is, the Cardinals have already swept a series against Milwaukee (albeit after dropping two prior series to the Brewers), swept a four-game set against the Dodgers and taken two of three against the Mets.
It's too early to declare anything, but the Cardinals are shaping up to be one of the beasts in this year's National League playoff chase.
<article class="ad-300" style="box-sizing: border-box; clear: both; overflow: hidden; position: relative; z-index: 1000026; margin: 0px 0px 10px 10px; float: right; color: rgb(0, 0, 0); font-family: Georgia, "Times New Roman", Times, serif; font-size: 16px;">
</article>3. Recently, I was talking to a team official who told me I should dig in and do a feature on how big league baseball sets up its season schedule. The beef was that some teams haven't gotten to host Mike Trout as much as others. Obviously, this was a National League city, where Trout might appear only once in a blue moon. The spirit of the concern is well taken: It would be nice if fans in every stadium could get a regular dose of the game's greatest player.
Alas, that's not realistic, at least not without greatly expanding the breadth of the interleague schedule. If you do that, then you start to infringe on the integrity of the pennant races to a degree even greater than interleague play already does. Besides, there is almost certainly no bias when it comes to the Trout scheduling.
Through the early part of this season, here is how many road games Trout has played against National League teams:
15: Dodgers (natural rival)
7: Rockies
6: Brewers
5: Nationals
4: Cubs
3: Braves, Padres, Mets, Reds, Pirates, Giants, Marlins
2: Diamondbacks, Phillies
As the seasons pass, the numbers will even out, and cities such as Atlanta and San Diego will get another turn in the rotation. Trout recently missed a weekend series at Wrigley Field, which was a shame. But if there were bias at work here, these numbers wouldn't look like this. In other words, Trout would have played more than two career road games against his hometown Phillies.
ESPN PLUS ($ MATERIAL)
Does baseball have a supply-and-demand issue when it comes to relief pitching? If so, what's behind it?
You wouldn't think this would be a possibility. For the past few years, as more and more innings have shifted from starting pitchers to relievers, we've heard about waves and waves of thunder-throwing young pitchers who have rendered the old models of pitching staff construction as obsolete as the rotary phone.
Yet there is a finite supply of everything in this corporeal existence of ours, and the consumption habits of humans being what they are, you had to figure we'd rub up against the outer limits of fungible -- but effective -- relief pitching. But these trends are so new and have accelerated so quickly that if the supposition presented in the first sentence above were to have some substance, it would be pretty surprising.
It's early. Every analysis piece with an April publish date must wear that caveat around its neck. Through Wednesday, we had played just under 15 percent of the 2019 season. That's right! We get to do this all 6.67 more times. Everything suggested here could be rubbed out by reality's eraser.
Nevertheless, here are some numbers comparing starting pitchers to relief pitchers in each of the past 10 seasons, with each season measured through the end of April.
<offer style="box-sizing: border-box;"></offer>
<aside class="inline inline-table" style="box-sizing: border-box; font-family: -apple-system, BlinkMacSystemFont, Roboto, Arial, "Helvetica Neue", Helvetica, sans-serif; height: 407px; border: 1px solid rgb(220, 221, 223); clear: both; margin: 6px 0px 18px; padding: 15px; width: 565px; color: rgb(0, 0, 0); font-size: 16px;">[h=2]Year-By-Year ERAs Through The End Of April[/h]
| YEAR | OVERALL | STARTER | RELIEVER | REL-STRT |
|---|---|---|---|---|
| 2010 | 4.20 | 4.29 | 4.01 | -0.29 |
| 2011 | 3.90 | 4.06 | 3.59 | -0.47 |
| 2012 | 3.84 | 3.89 | 3.73 | -0.16 |
| 2013 | 3.93 | 4.07 | 3.66 | -0.41 |
| 2014 | 3.82 | 3.88 | 3.70 | -0.18 |
| 2015 | 3.93 | 4.15 | 3.54 | -0.61 |
| 2016 | 3.98 | 4.10 | 3.76 | -0.35 |
| 2017 | 4.09 | 4.04 | 4.18 | 0.14 |
| 2018 | 4.12 | 4.20 | 3.99 | -0.21 |
| 2019 | 4.36 | 4.34 | 4.39 | 0.05 |
| Source: TruMedia |
<thead style="box-sizing: border-box;">
</thead><tbody style="box-sizing: border-box;">
</tbody><tfoot style="box-sizing: border-box;">
</tfoot>
What you should notice is that bullpen ERAs are generally better than starter ERAs, something that manifests in the opening weeks of the season and remains true through each campaign. So far in 2019, however, the starters are winning. In fact, it's the second time in three years that a season has started this way.
By the end of 2017, the relievers caught and surpassed the starters in ERA. That might be happening in 2019 already, as the gap has closed the past couple of weeks. But it's no secret that bullpens so far in 2019 have struggled across the big leagues. Teams have been forced to juggle roles and roster spots in an attempt to find the right mixes. Some have made progress; others haven't but will keep trying.
The question here: How long can clubs keep playing this shell game?
Let's consider some year-by-year reliever data downloaded from fangraphs.com. We'll work with fWAR for our purposes here.
In 2015, 97 big league relievers posted an fWAR of at least 0.5 -- a mere half-win -- for a combined fWAR of 102.5. The very next season, 93 of those pitchers logged big league innings. However, those 93 pitchers combined for just 55.5 fWAR, a drop of nearly 45 percent in value.
The trend only gets worse as the seasons go by. In 2017, those 93 pitchers were down to 81, and the fWAR was down to 45.2. Last season, it was 73 surviving relievers putting up 35.5 fWAR. So far this season, that original group of 93 is on pace for just 16.0 fWAR. Over the span of five seasons, this group of good relievers will have lost 84 percent of their collective value.
That is a lot of production to replace in a time frame that is less than the six-year window of controllable seasons for a player via baseball's service time system. And this rate of attrition might be accelerating.
Let's use 2016 as our starting point. That season saw 114 relievers with at least 0.5 fWAR. By the end of last season, those pitchers had lost 55 percent of their value. How about if we start with 2017, when there were also 114 relievers at 0.5 or better? Those hurlers, by the end of last season, had ceded 56 percent of their value -- from 131.0 fWAR to 57.8. So far this season, they are on pace for 38.0.
Last season, we saw 116 relievers reach 0.5 fWAR, with a combined total of 124.5. Those 116 pitchers are on pace to hit 59.4 fWAR. That's more than half the value of last season's best relievers gone in the span of part of one autumn, a long winter and a bit of spring.
What's going on here? To know, we'd have to burrow in a lot deeper than we will today. But I wonder if we're starting to see the makings of a supply problem. It's not just that the attrition rate of relievers is so high; it's that their production depreciates at an even faster rate. Teams bring back a lot of the same names, but their stat lines don't necessarily come with them.
We've known these things for a long time -- that reliever performance is difficult to predict. The trends toward bullpen domination have not been built on certainty at the individual level because it barely exists. Heck, in 2017, 13 relievers posted 2.0 fWAR or better for a combined total of 32.3 fWAR. That benchmark -- 2.0 fWAR -- is elite stuff for a relief pitcher. In 2018, those 13 bullpen aces combined for 9.2 fWAR. In 2019, they're on pace for 5.3.
So, yeah, individual relievers are flaky as heck, and that's why teams are so careful about overpaying for them. That's arguably why the best reliever of the 2010s, Craig Kimbrel, is currently unemployed. But teams have dealt with this extreme variability by cycling through arms like a long roll of lottery tickets. Eventually, good teams can hit upon enough positive producers to get through a season.
This season might be no different. Teams certainly show no signs of shifting bullpen innings back to the rotation, though that might change in 2020, when we begin the era of three-batter minimums. But you have to wonder if there will be a limit to how many competent pitchers teams will be able to burn through.
You also have to wonder how fair all of this is to relief pitchers. Relievers are promoted and asked to max out in terms of effort, spin and velocity. Then they will almost invariably burn out before they reach a level of service time that translates into big money.
But that's how this contemporary model of pitching staff construction is fueled -- on an endless supply of young power arms. What happens if that supply is not truly endless, if the burn-out rate surpasses the replacement rate? How will this model hold up when there aren't enough quality arms to cover for starting pitchers collectively asked to throw around five innings on average?
If we reach that point, then the model is broken. All things considered, that might not be the worst thing for the game -- or for the pitchers.
[h=3]Extra innings[/h]
1. You've probably seen this number going around: Cody Bellinger's double on Thursday at Wrigley Field moved him within one of tying former teammate Chase Utley for the most total bases by any player by the end of April. Bellinger had 84 total bases after that two-bagger against the Cubs.
Teams used to open their seasons later than they have in recent years, so all the leaders on the early-season total-base chart are recent. It's a good list to be on, and it might be predictive of postseason honors. Here are the leaders, other than Bellinger, with their total bases at the end of April and where they finished in the MVP voting:
Chase Utley, 2008 (85 total bases, 14th in NL MVP)
Jermaine Dye, 2000 (83, no MVP votes)
Alex Rodriguez, 2007 (82, won AL MVP)
Larry Walker, 1997 (82, won NL MVP)
Alfonso Soriano, 2003 (81, 20th in AL MVP voting)
Ken Griffey Jr., 1997 (81, won AL MVP)
Also, if you haven't noticed, Bellinger leads the majors with a ridiculous .426 batting average. His average on balls in play is .397. Of course, these numbers won't last. They can't. But ...
According to baseball-reference.com, which has splits of this type mostly complete back to 1908, Bellinger would become the 14th hitter to post a .400 average through April over at least 100 plate appearances. (Again, because teams start earlier these days, the leaderboard is titled toward recent decades.) In fact, .426 ranks fifth on that list.
Best March/April batting averages
Minimum: 100 plate appearances
1. Larry Walker (.456, 1997)
2. Darin Erstad (.449, 2000)
3. Ivan Rodriguez (.442, 1998)
4. Hal Morris (.427, 1998)
5. Cody Bellinger (.426, 2019 through Thursday)
2. Generally speaking, strength of schedule isn't a huge factor at the team level by the end of the season. Teams competing for the same thing -- division titles, namely -- pretty much play the same schedules as their competitors. Wild-card races can be a little skewed, but it's not a huge thing. However, early in the season, the standings can be greatly affected by scheduling quirks.
Which brings me to the St. Louis Cardinals. The Redbirds currently rank second in the majors in run differential, behind the Tampa Bay Rays. However, St. Louis has compiled that result against the fourth-toughest schedule in baseball, according to baseball-reference.com. Early as it is, the Cardinals have already swept a series against Milwaukee (albeit after dropping two prior series to the Brewers), swept a four-game set against the Dodgers and taken two of three against the Mets.
It's too early to declare anything, but the Cardinals are shaping up to be one of the beasts in this year's National League playoff chase.
<article class="ad-300" style="box-sizing: border-box; clear: both; overflow: hidden; position: relative; z-index: 1000026; margin: 0px 0px 10px 10px; float: right; color: rgb(0, 0, 0); font-family: Georgia, "Times New Roman", Times, serif; font-size: 16px;">
</article>3. Recently, I was talking to a team official who told me I should dig in and do a feature on how big league baseball sets up its season schedule. The beef was that some teams haven't gotten to host Mike Trout as much as others. Obviously, this was a National League city, where Trout might appear only once in a blue moon. The spirit of the concern is well taken: It would be nice if fans in every stadium could get a regular dose of the game's greatest player.
Alas, that's not realistic, at least not without greatly expanding the breadth of the interleague schedule. If you do that, then you start to infringe on the integrity of the pennant races to a degree even greater than interleague play already does. Besides, there is almost certainly no bias when it comes to the Trout scheduling.
Through the early part of this season, here is how many road games Trout has played against National League teams:
15: Dodgers (natural rival)
7: Rockies
6: Brewers
5: Nationals
4: Cubs
3: Braves, Padres, Mets, Reds, Pirates, Giants, Marlins
2: Diamondbacks, Phillies
As the seasons pass, the numbers will even out, and cities such as Atlanta and San Diego will get another turn in the rotation. Trout recently missed a weekend series at Wrigley Field, which was a shame. But if there were bias at work here, these numbers wouldn't look like this. In other words, Trout would have played more than two career road games against his hometown Phillies.
