Murphy’s Law of Bullpen Construction and Performance

The Cardinal bullpen is not very good.

Scratch that; the Cardinal bullpen is bad.

Yes, I believe that’s the one. The Cardinal bullpen is bad. Now, there is a question to be asked when it comes to just how bad, but we don’t necessarily need to answer that today. When we have the whole of baseball history essentially laid out before us at any given moment, should we wish to peruse it, the question of how much a thing is what it is becomes a matter of interest and research. If one wished to know just how bad this year’s bullpen is compared to, say, the 2003 club, one could find that answer with enough care and energy.

Me? I am not particularly interested in just how bad. Say it is bad, and there let it lie, would be my advice on the matter. Rather, I am curious less about the quantity of bad, and more about the nature of bad. I want to know the how, if not the why. The why will probably always prove elusive, because why is anything anything on the baseball diamond? You pays your money and you takes your chances, and sometimes it’s Greg Holland all the way down.

But what I’m interested in is trying to quantify all the various ways in which the Cardinal bullpen has been bad this year, and just marvel at the sheer shit showedness of the whole endeavour. Many of us, myself included, thought coming into the season that the Redbird ‘pen would be a real strength of the club. No, they didn’t have that one elite arm to blow hitters away at the back of the ‘pen, but what they had was a depth of high-strikeout relievers that I couldn’t recall having seen from this franchise — and maybe not elsewhere, either, outside of New York — in a very, very long time. Maybe not ever. They had stockpiled an incredible number of arms capable of posting 25%+ K rates, and while maybe there was no single hammer, a Kenley Jansen or Craig Kimbrel, to put terror into the hearts of hitters, a choir of Ks from the sixth inning on of any given game was a suitably exciting, and intimidating, thought.

And that was before the club invested a huge chunk of change and gave up a draft pick to bring in perhaps the biggest fish in the closer market during the offseason, or at least one of the top, say, three. That was meant to be the finishing touch, a triumph of timing, a reward for waiting out the market and acting at a moment when there was an opportunity in play that had not been expected. Instead, we got...well, we got what we got. And what we got ain’t so good, folks.

So let us catalogue, for posterity and our own masochistic souls, all the ways in which the 2018 bullpen has gone wrong.

Oh, and before we dive into this, I think everyone should take a look at Jeff Sullivan’s very useful recent piece at FanGraphs on reliever reliability/predictability. Essentially, he compares year-over-year repeatability of performance between position players, starting pitchers, and relievers, specifically those performances which meet an above-average definition. And what he finds is that relievers are, in fact, somewhat more volatile season to season, but not by a huge margin. They are riskier investments than other sorts of players, but only just barely. Keep that in mind as we look at just how many things have gone wrong in the ‘pen for the Cards this year.

Greg Holland

2017 ERA: 3.61 —> 2018 ERA: 7.92

2017 K%: 29.8% —> 2018 K%: 16.7%

2017 BB%: 11.1% —> 2018 BB%: 16.7%

So let’s start right at the top with the big one, shall we? Greg Holland was, from 2012-2014, one of the very best relievers in baseball. As early as 2010 he was striking out gobs of hitters, but it was that ‘12-’14 window when he really locked things down as the Kansas City closer and became the force that drove a lot of that bullpen’s success. In 2015, he pitched most of the season with a partially torn UCL, then finally acquiesced to Tommy John surgery late in the year. He missed all of 2016, came back in 2017 with the Rockies, and was pretty good. He wasn’t great, but he was pretty good. Pitching in Coors Field, he posted an ERA in the mid-3s, struck out right around 30% of the batters he faced, and held the opposition to a batting average against under .200. Pretty good, right?

There was quite a lot of celebration in various corners of Cardinal fandom when the Holland signing became public knowledge. The reaction here at VEB was relatively even-keeled, if I remember correctly, but still mostly positive. There were other places, though, where the mood was positively jubilant. I remember several posters over at Gateway Redbirds essentially celebrating that, finally, the Cardinals were committed to winning. The comments section of Bernie Miklasz’s columns were filled with people saying much the same thing, that finally Bill DeWitt had opened DeWallet for something good, instead of just bilking the people out of their hard-earned money.

And then, of course, Greg Holland actually started pitching for the Cardinals, and the whole thing went to hell in a handbasket. He’s been one of the worst relievers in baseball this year, one of the worst Cardinal relievers of this century, and has pretty directly cost the Cards something like half a dozen games this year by my quick mental inventory. So, you know, great signing.

Now, lest you think I’m taking the front office to task for signing Holland, I am, but I’m not excoriating anyone. Yes, I thought Holland was teetering on the edge of being not very good anymore last year based on his approach, and yes, I would rather have kept the draft pick and figured the ‘pen out without this expensive addition, but in no world did I envision Holland being the kind of abject disaster he has proven to be. I thought he might just not be very good anymore; the gulf between ‘not very good anymore’ and ‘Greg Holland 2018’ is a distance where we must measure sadness in leagues, rather than miles. I thought the Holland signing at the time was a needless outlay for a player that wasn’t a difference-maker based on his previous reputation; I didn’t expect a Ralph Nader book.

Brett Cecil

2017 FIP: 3.26 —> 2018 FIP: 5.35

2015 - 2018 K%: 32.7% —> 28.7% —> 23.8% —> 12.4%

2017 BB%: 5.8% —> 2018 BB%: 15.9%

Whew. Those are some numbers.

From 2013 to 2016, Brett Cecil was one of the best setup relievers in baseball. The Cardinals, who lost Zach Duke to Tommy John surgery following the 2016 season (which eliminated a big portion of the club’s reason for making the deal with the White Sox to bring him in, which was the fact he was under contract for 2017, and probably merits an entry of its own in a list like this not constrained to 2018), went into the free-agent market and signed Cecil to a four-year deal worth $30 million. They didn’t want to go the fourth year, from what we heard at the time, but much like the Dexter Fowler signing, they pushed beyond where they wanted to go to make sure they got the player.

In 2017, Cecil got off to an ugly start, one which probably guaranteed the fanbase would never really accept him, but his overall numbers were just fine. The strikeouts were down, which was concerning, but a lot of that had to do with some strange difficulties he seemed to have throwing strikes to left-handed hitters at times. Still, it was not a terrible season, even if maybe what we remember is the first three weeks of the season.

Then this year happened, Cecil got off to a late start in spring training, got a late start to the season, and then just never really got started. His strikeout rate this season is only a little over a third of what it was in 2015, and less than half what it was in ‘16. Hell, it’s not much more than half of 2017’s number. Oh, and he’s also lost two and a half miles per hour off his fastball since 2016.

Brett Cecil turned 32 years old less than a month ago, and he’s either hurt, or just...done.

Matt Bowman

2017 FIP: 3.65 —> 2018 FIP: 4.72

2017 BB%: 7.3% —> 2018 BB%: 10.1%

2017 HR/9: 0.61 —> 2018 HR/9: 1.57

2017 BABIP: .276 —> 2018 BABIP: .373

It’s funny, really, that Matt Bowman has actually managed to turn himself into a decent little strikeout pitcher this year, running a K rate of just under a quarter. Why is it funny? Because everything else has gone so horribly wrong this season, including multiple occurrences of blisters on his pitching hand, that the increased strikeouts have gone completely unnoticed.

Part of Bowman’s struggles this year can obviously be attributed to that very high batting average on balls in play, which probably showcases a decent amount of bad luck. However, that number combined with his elevated home run rate (along with just, you know, watching him pitch), makes me believe he’s simply made more mistakes over the plate this year than in the past. His groundball rate has fallen significantly since his debut in 2016, and this year over 20% of the fly balls he’s given up have gone over the fence. Matt Bowman is probably a little unlucky, yes, but he’s also just getting creamed on hard contact out there.

Tyler Lyons

2017 K%: 30.9% —> 2018 K%: 23.2%

2017 HR/9: 0.50 —> 2018 HR/9: 1.65

2017 BABIP: .295 —> 2018 BABIP: .420

Thank god Tyler is still handsome, because everything else about him has gone to shit this year.

Actually, that’s not completely fair; not everything has gone wrong with Tyler Lyons this year. Just...most things. His groundball rate has fallen off a cliff, his strikeouts are down by almost a quarter, and while there’s certainly a lot of bad luck baked into that BABIP, he’s also just given up a lot of very loud contact this season, in much the same way Matt Bowman has.

In 2017, Tyler Lyons was one of the best left-handed relievers in baseball. He looked poised to consolidate those gains this year and move into that Brad Hand tier of guys just below the Miller/Chapman echelon. Instead, he’s been on the disabled list a couple times, most notably with an elbow strain, and is currently running an ERA six runs higher than his 2017 line.

Dominic Leone

2017 IP: 70.1 —> 2018 IP: 13

Honestly, Dominic Leone’s performance in 2018 has been more or less fine. Elevated home run rate, certainly, but that kind of thing can happen in just over a dozen innings.

The problem, of course, is that the Cardinals once again brought in one of the best setup relievers in baseball in 2017 to bolster their bullpen, and he’s thrown just over a dozen innings. Last year, Leone was worth 1.5 fWAR, and this year he’s dealt with a nerve issue in his arm that has kept him out for something like three months now, and there’s still no concrete timeline for him to return.

Kind of feels like snake eyes on pretty much every one of these bets, doesn’t it?

Luke Gregerson

2016 K%: 29.1% —>2017 K%: 26.6% —> 2018 K%: 21.6%

2017 IP: 61 —> 2018 IP: 11.2

Stop me if you’ve heard this one before, but Luke Gregerson was one of the best setup relievers in baseball from basically 2009 to 2016. Now, he actually served as the Astros’ closer for a period of time in there, but for the most part he served as an elite setup man for the Padres, then the A’s, and then Houston, over the course of eight years.

In 2017, Gregerson pitched roughly the same as he had before, but suddenly had a home run issue, where he never really had before. He still struck out a ton of hitters, and still got a ton of ground balls, but nearly a quarter of the fly balls he allowed in ‘17 left the park, whereas before it had never been over 14%. So, some bad luck on a few fly balls sneaking over the wall aside, he still looked like an excellent setup reliever for the most part. He had also been remarkably durable over the course of his career, having made at least 59 appearances per season every year from 2009-2017.

And so, of course, he’s barely seen the field this season, with a litany of injuries having kept him on the shelf. When he has pitched he’s been shaky, but it’s awfully hard to get going when you’ve been on the DL so much in a given year.

So what do we have here? Well, we have a group of six relievers who, in 2017, produced 5.4 fWAR collectively for their teams. And that’s even with Gregerson grading out as exactly replacement level due to his homeritis. He was worth 1.2 fWAR in both 2015 and ‘16. If we go with the Baseball-Reference version of WAR, which is based on runs allowed rather than FIP and which I personally prefer, we get a remarkably similar 5.5 bWAR for the group of six. And again, that’s with a replacement-level performance from Gregerson.

How have those six relievers fared for the Cardinals in 2018? Well, by FanGraphs’ WAR they’ve been worth -0.8 wins. So a swing of over six wins in value by that model.

By the runs allowed model, which again I prefer due to it being more reflective of what happened, and even with the extra noise introduced captures more signal in my opinion, things are even worse. Much worse, in fact.

The 2017 versions of all those relievers were worth 5.5 wins above replacement in terms of actual runs allowed. In 2018, that group has “contributed” -3.5 wins to the cause. That’s hard to even get your head around, isn’t it? By runs allowed, those six relievers averaged nearly a win apiece last season, and this year have been nine wins worse. Obviously, we can’t just take those nine wins and tack them on to the Cardinals’ record; WAR doesn’t translate to wins and losses that cleanly. But if you’re looking for where all the value this team was supposed to derive from investing heavily in bullpen arms went, I can point you to exactly where to look.

Now, to be fair, it hasn’t all been dark clouds for the Cards in the bullpen this year. Bud Norris has been outstanding, Jordan Hicks has emerged as an extraordinarily exciting arm, if not an entirely finished product, and Sam Tuivailala, Mike Mayers, and John Brebbia all look like perfectly serviceable sixth- or seventh-inning type relievers most days. Norris has obviously been the biggest boon to the club, even if Hicks has been the bigger story.

Even with those positives, though, it’s impossible to ignore the fact that the Cardinals went all-in this past offseason in trying to build a bullpen so deep with quality arms that they could overwhelm the opposition for nine or even twelve outs on any given night, and instead ended up with a group of six relievers who have seen their collective performance decline to an utterly shocking degree this year.

Perhaps one could find fault with targeting so many previously excellent relievers, rather than pouring resources into future bullpen arms, but it isn’t as if the Cards signed a bunch of late-30s reclamation projects. Gregerson best fits that bill, and so perhaps his health concerns at age 34 shouldn’t come as a huge surprise. Still, that’s 34, not 37 or 38, and he’s easily the oldest of the group we’re talking about here. (For the record, Norris is 33 and has been awesome, while none of the others of that group of six are over 32.)

Honestly, I don’t know what you do with this information going forward, and what it means for future Cardinal teams or Cardinal bullpens or anything else. Unless you want the message to be never sign relievers, nor trade for relievers, I’m not sure what this tells us. I’m sure there will be people in the comments who will find some way of turning this around to reflect the cheapness and apathy toward winning of ownership and the front office, because that seems to be the thing to do right now for a certain segment of the fan base. But I have a very tough time squaring that circle in my mind, no matter how frustrated I might be with this team.

The 2018 Cardinals are just above .500, and falling out of contention, with 63 games left in the season. We can find plenty of reasons for that state of affairs, from the big offensive investment of the offseason (Marcell Ozuna), being a bust of Tino Martinezian proportions, to the now-former manager consistently deploying assets in a less-than-optimal way, to Dexter Fowler going to bed at 31 and waking up one day later at 38. But of all the things that have gone wrong for this team, and let’s face it: they are legion, the one that most immediately, and directly, explains why things have gone so poorly is that group of six relievers who went from +5.5 wins to -3.5 wins out of nowhere.

And I really don’t know how you avoid that. Sometimes, shit just happens.