clock menu more-arrow no yes mobile

Filed under:

White Sox decision review: Relievers

Despite a couple blown saves too many, David Robertson looked like the reliever the Sox expected, a claim Zach Duke can't make

Jason Miller/Getty Images

In a perfect world, a team wouldn't have to spend heavily on its bullpen. It would be able to curate relievers from its farm system, supplement it with cast-offs and trades, keep them until saves make their cost prohibitive, and then it's next man up.

But sometimes you get what you pay for, and the White Sox certainly did in 2014. They tried a bare-bones approach around a groundballing, hard-throwing staff, but the injuries to their two most tenured relievers -- Nate Jones and Matt Lindstrom -- caused the rest of the bullpen to collapse.

The Sox did mine a couple of useful performances out of the mess -- Zach Putnam and Jake Petricka, namely -- but they still lacked bona fide high-leverage guys, especially with Jones still on the mend from Tommy John surgery. If they had immediate designs on challenging for a wild card spot, they would have to pay. And because they spent so little on the bullpen the year before, they had room for one potentially regrettable contract.

Rick Hahn went ahead and took on two of them. They signed Zach Duke for three years and $15 million, then surprised everybody by landing David Robertson, the best reliever in free agency, for four years and $46 million.

While Robertson topped the class, there were a few ways Hahn could've invested in free-agent relievers. Below are eight relievers who ended up signing considerable contracts after extensively exploring free agency. I could go further down the line, but after a point, the signings are all too low-risk to register. The White Sox ended up getting a high-reward performance out of one of those with Matt Albers, but if he were the prize bullpen acquisition last season, Hahn would've had a hard time explaining it.

Robertson 4/$46M 60 63.1 34/41 7 13 86 3.41 2.52 0.5/1.9
Miller 4/$36M 60 61.2 36/38 5 20 100 2.04 2.16 2.2/2.0
Gregerson 3/$18.5M 64 61.0 31/37 5 10 59 3.10 2.86 0.8/1.2
Duke 3/$15M 71 60.2 n/a 9 32 66 3.41 4.62 0.5/-0.3
Rodriguez 2/$13M 60 57.0 38/40 6 11 62 2.21 2.91 1.7/1.0
Neshek 2/$12.5M 66 54.2 n/a 8 12 51 3.62 3.94 0.3/0.3
Grilli 2/$8M 36 33.2 24/26 2 10 45 2.94 2.12 0.5/1.2
Janssen 1$5M 48 40.0 n/a 5 8 27 4.95 4.08 -0.1/0.0

The thing that jumped out to me -- Duke is the only one with a negative WAR by FanGraphs' estimation ... and yet his WAR is the same as Robertson when it comes to's system.

Unlike the position player side of the equation, there's not much murky about the difference here. Runs allowed factors into bWAR, while fielding-independent numbers go into fWAR. So there's Duke and Robertson posting identical ERAs over nearly identical innings totals, and they end up with the same bWAR. And there's Robertson with a 6.6-to-1 strikeout-to-walk ratio while Duke barely clears 2-to-1 with more homers allowed, and that's the difference in FIP.

Robertson inadvertently serves as an example of the value of both systems. He allowed just a .196 average and a .241 OBP. Take those rates and combine them with that aforementioned excellent strikeout-to-walk ratio, and he should've had the ingredients for a great season.

Except ... it wasn't a great season, just a respectable one. The runs and seven blown saves happened somehow. There are the seven homers, which matched his total from the previous season. He stranded only 65.7 percent of runners, well below his career rate of 79 percent, which makes you look at his numbers with runners on:

  • Bases empty: .153/.188/.252
  • Runners on: .292/.354/.414

Part of that may be luck. But Larry also pointed out that Robertson, who continued his career trend of significant reverse splits, faced more righties than lefties for the first time since 2010, and by fair amount:

  • vs. RHB: .210/.238/.413, 6 HR over 143 PA
  • vs. LHB: .175/.245/.216, 1 HR over 107 PA

In each of the last four seasons, Robertson has faced more lefties, even though he established those reverse splits long ago. So why the change this year? Perhaps managers are wising up, but it may be more a byproduct of the White Sox having four left-handed starters. Opposing lineups start the game loaded with righties, and it's probably a lot easier to leave a righty against a pitcher with reverse splits rather than make a conscious decision to remove a left-handed hitter who would theoretically has the stronger side in almost every other such matchup.

Robertson experienced his share of misfortune, too -- Baseball Prospectus puts his Deserved Run Average at 2.78, which is closer to his FIP than his ERA.  There's also no evidence that his stuff is lacking, as he's getting more swinging strikes than ever before. If anything, it's possible that he's throwing too many strikes (his 5.2 percent walk rate was a career low).

Even if you weigh the results more than the process, it'd be difficult to call it a bad decision, because the White Sox needed strikeouts, and he delivered them. That's why the dueling WARs tell Robertson's story well. By bWAR, you get a guy who had a few too many frustrating outings for his liking with legitimate underlying theories for them, but fWAR suggests that he could have a similar kind of season and come out way ahead next year, which is probably true.

Factoring in both peripherals and actual results, the only guy who comes out notably ahead of Robertson is Andrew Miller. It sure is easy to admire what the Yankees did. They gave Robertson a qualifying offer, then replaced him with Miller, who didn't get one from the Orioles. The result? A better season and an extra draft pick.

/low whistle.

If the White Sox could do it over again, they might want to sign Miller, even if they had to survive a slight bidding war with the Yankees to do so. That would've allowed them to keep their draft pick, and maybe it it would've dissuaded them from signing Zach Duke, who had a disappointing season for easy-to-diagnose reasons. Basically, his ability to locate took a big step back.

They could've also played the waiting game for Francisco Rodriguez, a Boras client who didn't sign until February, even if it would've flirted with stalling the effort to rebuild the bullpen.

Both are just too hindsight-based, though, because Robertson was far and away the most credible relief pitcher on the market. Miller and Duke were each coming off their first good season as relievers, with Duke compensating for lesser stuff with more distinct mechanical changes to make him the Old Forester to Miller's Woodford Reserve. (Duke ended up being the Early Times, but it's not out of the possibility that a left-handed reliever could bounce back.)

Moreover, the Sox needed both a ninth-inning guy, and a go-to lefty before the ninth inning. Robertson offered security and the Sox ponied up for it. Maybe the Sox could've just signed two lefties, but giving Miller the toughest tasks was a moderate gamble, and one that's easier to take on when the backup plan for the highest-leverage work is Dellin Betances, and not, I dunno, Putnam.

A year into Robertson's four-year commitment, the Sox shouldn't have any regrets. And even if Robertson had Duke's season, it'd be hard to blast them for gunning for the best available talent for the position. I wouldn't want to dissuade them from doing so in the future, because as our offseason plans are showing, it'd certainly help if they went big again this winter.