How I learned to stop worrying and love the Reds bullpen

How I learned to stop worrying and love the Reds bullpen

The 2019 Reds bullpen has produced a great deal of anxiety among Reds fans. From time to time during the long season there has been occasional cause to worry. But overall, the Reds relief corps, particularly the part that remains on the roster and likely to return in 2020, is a bright source of optimism.

If that’s impossible for you to believe, examine your standards. You can’t measure the 2019 bullpen by your memory of Rob Dibble or Aroldis Chapman circa 2012. Static criteria don’t cut it when the game has changed so rapidly, including a nightly whiteout of juiced baseballs rocketing out of every park.

To accurately evaluate the 2019 Reds bullpen, it’s important to consider the current context.

Context: More Bullpen Innings and Diminishing Returns

Across the league, front offices and managers have absorbed and accepted data about pitchers facing batters a third time. They have adopted concerted strategies to shorten starts. As an inevitable byproduct, bullpens are being asked to cover more innings. This chart shows the recent trend in percentage of innings pitched by relievers since 2014.

In 2019, bullpens take on 41 percent of innings pitched, a historic high. The Nasty Boys’ bullpen covered 32% of the 1990 Reds innings.

That trend is clear but less of a straight line when you look at individual teams. Bullpen load is partly a function of the quality of the starting rotation. Year-to-year variation from smaller samples sizes produces jagged lines. This chart shows the Reds percentage of bullpen use, back to 2011.

The great starting pitching staff of 2012 covered 70% of the team’s innings. That low usage rate lasted through the 2014 season, when the Reds rotation consisted of peak Johnny Cueto, Mat Latos, Homer Bailey and Mike Leake. After that, as the Reds rotation cratered, innings pitched by Reds relievers soared to 42%.

The 2019 Reds bullpen has thrown 38% of the team’s innings despite a starting rotation that’s better than the 2012 and 2014 versions relative to the league. A strong starting five has allowed manager David Bell to be judicious in his use of the bullpen. But Bell has shown a quick hook when starters begin to fade. Bell has spoken of the Third Time Through penalty several times and he manages with it in mind.

More innings covered by the bullpen means a decline in overall quality of relief pitching. It’s a classic example of the economic principle of diminishing marginal return. The more of a resource you use, the less productive the additional units are. Whether that means involving additional pitchers who aren’t as good, or overusing the ones you have, the result is lower productivity.

To demonstrate, consider this: Randy Myers, Rob Dibble and Norm Charlton covered 55% of the Reds bullpen innings in 1990, and that’s despite Charlton’s two months as a starter. In contrast, Michael Lorenzen, Amir Garrett and Raisel Iglesias have taken just 38% of the bullpen’s load in 2019.

The Law of Diminishing Returns applies to major league bullpens and it is happening before our eyes in 2019.

Context: All Those Home Runs

Sometime later this week, MLB’s league-wide home run record of 6,105 hit in 2017 will be surpassed. Projections point to the record being obliterated, with guesses as high as 6,800 homers or more before 2019 is done. As recently as 1997, there hadn’t been season with 5,000 homers.

Bullpens haven’t been spared from the deluge. Whether it’s the baseball’s seams, the launch angle revolution or something else, relievers aren’t exempt. Remember, they’re pitching even more than before.

Here’s a chart of home runs per game per team each year back to 1991. Note the spike in 2019.

Even though pitchers in 2019 are throwing harder and striking out more batters, they’re also give up hundreds of more home runs.

The number of runs per game has grown to a level surpassed in recent times only by the Selig Steroid Era. A standard for an average pitcher ERA just five years ago of 4.00 is way out of date today. Average runs allowed per game is almost a full run higher now.

Historically, metrics (ERA, FIP, xFIP) for starters overall have been worse than for relievers. If you analyze that through the lens of diminishing marginal utility, managers should have used their bullpens more and starters less. We’ve seen managers adopting that strategy in the past few years.

In fact, the brunt of the impact of all the runs is being born mostly by relievers. Yes, starting pitchers are performing worse than they were six years ago. But higher offensive output is partly offset for starters by being assigned fewer innings.

But as of 2019, bullpens have overtaken starters. You can see in the next chart the overall upward trend plus the jump in 2019. Graphs of ERA or FIP show similar patterns. Bullpens have taken on more innings in 2019, so they’ve absorbed more of the impact of the accelerated run scoring environment.

That’s the context. More of the innings load carried by bullpens, which means worsening numbers. That trend is occurring at the same time that offense is exploding. The two factors have a compounding effect on overall bullpen performance stats. Context changes, so must our standard of evaluation.

Bullpen Worrying is League-Wide

Reds fans watch the Reds bullpen every night. If they paid attention to other teams relief staffs as closely, they’d find similar reason for bullpen anxiety in other cities. Looking for a major league team comfortable with its bullpen? The pickings are slim. Grant Paulsen, who covers the Washington Nationals for The Athletic, noted this week that the Wild Card-bound Nats “are not the only team marred by bullpen ineptitude. Almost every club in baseball can relate. It’s an epidemic, spreading throughout the game.”

How about a sample of the misery of others. Lauren Theisen, wrote about the Cubs in Deadspin last month: “Eighth and ninth innings like the ones seen last night might become a disturbing regularity for Cubs as they continue to try and put some space between themselves and their division rivals, because this bullpen is a haphazardly glued-together collage of injured talents, mediocre journeymen, anonymous filler, and unproven kids. Most prominent among these shaky arms is big-name signing Craig Kimbrel, who came to the team in June and proceeded to pitch, for obvious reasons, like a guy who didn’t have the benefit of Spring Training this year.”

Reds fans who feel bad about our bullpen may not realize it’s a league-wide condition.

The One-Way Ratchet and Negativity Bias

There’s a natural unevenness or bias in the way we experience bullpen success and failure. Research on the brain indicates it reacts with greater force to negative news. It’s called Negativity Bias. We’ve inherited genes that predispose us to pay more attention to negative rather than positive information.

When the relievers do their jobs, it barely registers. We move on to what’s next. But when a Reds reliever gives up a run we dwell on the pain. The worry ratchet turns up, not down.

For example, when Raisel Iglesias gave up a home run against Miami on August 29, the “oh no, not again!” cries from that segment of Reds fans reached a crescendo. Forgotten was the previous month (July 27-August 27) when Iglesias had pitched 13.1 innings, struck out 18, walked only 1, put up an ERA of 2.70 and an xFIP of 2.98.

Iglesias had been terrific — for a month! In September, Iglesias has faced 12 batters, struck out 6, allowed no hits and one walk to a base runner who was erased on a double play.

The average fan thinks Raisel Iglesias has had a terrible year. [This is in part because of a stupid, lazy narrative about “his” losses, which are really team losses, but that’s another post.] Here is a partial list of pitchers with more blown saves than Iglesias: Kenley Jansen, Josh Hader, Roberto Osuna, Sean Doolittle and Edwin Diaz (there are 15 in all). The 16 relievers with the same number of blown saves as Iglesias (that would be 5) include Aroldis Chapman, Ryan Pressly, Brad Hand and Liam Hendricks.

Performance Numbers: Reds Bullpen as a Whole

The Reds bullpen has improved vastly in 2019 compared to previous years. For evidence of that claim look at a number of statistics. This is a graph of Reds bullpen xFIP- from 2010 to 2019. xFIP- is a composite number on a 100-point scale, adjusted for park, season, run scoring environment, defense and luck. It measures pitchers on what they control: strikeouts, walks and ground ball rates.

CHART: MLB average = 100, below is better, above is worse

Well, look at that.

The 2019 Reds bullpen xFIP- is 94, that means 6 percent better than average. That’s an enormous improvement over 2014-2018. In the past decade, only the stellar 2012 bullpen of the NL Central division champions (led by Aroldis Chapman, Tony Cingrani, Sean Marshall, Jonathan Broxton and Sam LeCure) was better than the 2019 Reds bullpen. Read that last sentence again if necessary.

The 2019 Reds bullpen is second in the NL in strikeouts, fourth in FIP, second in xFIP, fourth in SIERA and sixth in fWAR.

And yes, those numbers include the malfeasance of Zach Duke, Wandy Peralta, David Hernandez, Jared Hughes and Sal Romano.

Performance Numbers: Individual Reds Relievers

Michael Lorenzen, Raisel Iglesias, Amir Garrett, Robert Stephenson and Lucas Sims have all been better than league average. In some cases, they’ve been quite a bit better. Kevin Gausman has been spectacular in his new relief role, albeit over a small sample size. He’ll be back in the bullpen if the Reds front office and Gausman can agree on fair payment,

League average xFIP- is 100, with numbers less than 100 being better than average. Every point lower is a percent better. Amir Garrett, mountain of walks and all, has been 17 percent better. Robert Stephenson 14 percent better. Raisel Iglesias, with all those home runs allowed, has still been 14 percent better than average.

The second metric in the table, xwOBA, measures actual quality of contact independent of fielding, luck and ballpark; plus walks and strikeouts. League average is .316. Numbers lower than that are better that average. Every one of those pitchers has been much better than average.

Conclusion

Our understanding of what it takes to win baseball games is evolving faster than Michael Lorenzen runs around center field. Sure, you’ll still hear a lousy take or two from the stone gargoyles, the folks guarding the microfiche at the public library and defending creaky old baseball bromides. They have a strange love for the cliché that you can’t have too much pitching.

Well, in a world where baseball teams have budgets, meaning this world, yes you can have too much pitching.

For optimal results, teams allocate their finite resources to areas of greatest need. For some clubs at certain times, that’s pitching. For the 2020 Reds — clear as can be — it’s the offense. In 2019, the Reds rank #12 of 15 in the National League in runs scored. This isn’t Cold War rocket science.

That’s why it’s important to place the performance of the 2019 Reds bullpen in context. Given the heightened run scoring environment and greater innings demands being placed on relievers, the Reds 2019 bullpen comes out looking good. The Reds have five (Lorenzen, Iglesias, Garrett, Stephenson, Sims) or six (add Gausman) relievers who have had strong seasons relative to the league. For 2020, the bullpen isn’t a problem that needs significant off-season attention.

In fact, the success of Reds relievers along with the return of five solid-to-excellent starting pitchers means the club should focus its tens of millions of available dollars on getting the hitting.

Go ahead and worry about the Reds bullpen if you want. That’s any fan’s right. But based on its performance, the Reds bullpen is more deserving of your love than angst.

Steve Mancuso

Steve Mancuso is a lifelong Reds fan who grew up during the Big Red Machine era. He’s been writing about the Reds for more than ten years. Steve’s fondest memories about the Reds include attending a couple 1975 World Series games, being at Homer Bailey’s second no-hitter and going nuts for Jay Bruce at Clinchmas. Steve was also at all three games of the 2012 NLDS, but it’s too soon to talk about that.

10 Responses

  1. Armo21 says:

    Another excellent post about the bullpen that is data driven…I got in a debate 3 weeks ago about Iglesias when he had 4 blown saves and 9 losses. Guy could not believe he had less blown saves than the players you mentioned.

    Clearly the one/two run losses is reflection of lack of hitting when you see there are 3 teams that have scored less runs than the Reds.

    Thanks Steve.

    • Clay Marshall says:

      “Clearly the one/two run losses is reflection of lack of hitting when you see there are 3 teams that have scored less runs than the Reds.”

      I think this is what I still have a hard time grasping. Call it the eye test, the smell test, whatever, but if the Reds bullpen is so good, then why aren’t they winning more games? In theory, if the Reds are perpetually playing teams close and they have an above-average bullpen, you’d think they’d win more of those close games than they’d lose. As we know, the opposite is true, and last night’s loss was more fuel for the fire. That said, is the Reds offense really that bad — and forgive me if I’ve missed a deep dive on this, but it certainly seems worse later in games than it is in early innings — and/or are other teams’ bullpens outperforming their peripherals when they face the Reds? It certainly seems that they lose more late-inning heartbreakers like last night’s game than they win.

      And Steve, congrats on finding your zen. I’ve long made peace with this season not counting, but bullpen meltdowns never fail to rile me up. My first thought this morning was the lede I wish I’d written last night: “David Bell wanted a LOOGY. Instead, he got BS.”

      • Steve Mancuso says:

        Losses are a combination of offense, pitching and defense. The alternative to losing close games isn’t necessarily winning close games. It’s also losing by a lot. But the Reds excellent pitching this year has held opponents down. It’s much harder to blow a 2-run lead than a 1-run lead and even harder with a 3-run lead. If the Reds offense scored more, the bullpen’s win-loss bottom line would look better. When you enter tied games late an your offense seldom scores, that’s going to mean losses, not wins. Wins are a team stat and shouldn’t be assigned to pitchers of any flavor.

        Where to assign blame? The offense is #12 out of 15 NL teams in run production. The pitching is #2 of 15.

    • Steve Mancuso says:

      The Negativity Bias is real and strong when it comes to bullpens. People remember the bad games more than the good ones.

      • kondwani1M says:

        5 relievers isn’t going to cut it Steve, this club needs to sign or trade for at least 2 proven relievers this winter. To go along with added offense.

      • Steve Mancuso says:

        Gausman makes six. They can promote players from within the organization for the last two slots, like Cody Reed. The relievers who have been most reliable this year have all been from within the organization other than Sims who was in the Duvall trade. I’m not against signing a veteran (although skeptical), but it’s a low priority compared to other needs.

  2. Great stuff, as always. I’d love to see them add a veteran lefty to the pen, but the offseason needs to about rebuilding this wretched offense.

    • Steve Mancuso says:

      Agreed about needing another lefty or two. I’d rather it be Cody Reed than a veteran, but both wouldn’t be bad. I’m just generally skeptical of veteran relievers. Shopping based on previous season stats feels like a dog chasing its tale.

  3. R Smith says:

    Really good article . Thanks Steve. Just wish Cody Reed had been healthy this year. He and Garrett could be a potent young duo from the left side. Interesting to see in 2020 the impact of the 26 man roster and 3 batter rule on bullpen construction.

    • Steve Mancuso says:

      Agree that Reed’s absence was a real downer. This could have been his year to really make an impact. They tried to get by with Zach Duke and Wandy Peralta and Reed would have been much better than either of those.