Arizona Locks Up Unique and Useful Brad Ziegler with Team-Friendly Deal

Nearly two weeks ago, the Diamondbacks signed Brad Ziegler to a two-year deal worth $10.5M in guaranteed payments, avoiding arbitration for the team’s last year of control.  The yearly salaries — $4.5M in 2014, $5M in 2015, and even $5.5M for the 2016 option year — are extremely reasonable rates for the club.  Make no mistake: Ziegler is a very good and very unusual player, made more valuable because he is unusual in a way that is hard to replicate.

A quick review: in our Offseason Plan, Jeff and I urged the team to try to sign Ziegler to a 3-year, $13M extension.  The actual deal is just about as good.  Although Ziegler could end up making $15M over the life of the deal, the extra salary is essentially the cost of the option.  If Ziegler is hurt or terrible, the D-backs pay only $10.5M ($2.5M less than $13M) — and if he’s worth at least $4.5M to the team for the 2016 season, the D-backs pay $15M ($2M more than $13M).  In other words, I think we were really close on that deal.  The thing is, though, that I did not appreciate just how much of a positive impact Ziegler can have on the outcome of games.  After a week of research on a related project, I’m ready to declare: this extension is really, really good.

I wrote in November about why it’s substantially smarter for a team like the D-backs to make peace with using mostly young pitchers in their bullpen.  I’m not backtracking from that — even in that piece I carved out a “Ziegler exception” for two reasons (predictability and the undervalued nature of his skills) that are explored below.

Brad Ziegler is Really Good

Among 135 qualified (50+ innings) relievers in 2013, Brad Ziegler finished 23rd in ERA — really good.  Unfortunately, ERA is particularly ill-suited for short relievers who frequently come in mid-inning, because when a pitcher enters with runners on base, the pitcher profits from the fact that it is easier to get outs in those situations (more places to make plays, and double plays are possibilities), even though the pitcher will never have to pay the piper for that benefit (if they score, the inherited runners will count towards the ERA of a predecessor).

So we could turn to Fielding-Independent Pitching, a metric that runs away from results in favor of metrics that are more stable and reliable, like ground ball % and strikeout %.  Ziegler tied for 68th place in FIP with a 3.40.  That’s not bad, but it just so happens that FIP does not do Ziegler justice — FIP does not account for the fact that Ziegler is profoundly efficient in many of the situations in which he is generally deployed.

Win Probability Added sheds some light — WPA takes a the win expectancy of the player’s team before a player event, then the win expectancy after, and calculates the difference.  It’s a counting stat, so the more damage you do, the higher WPA goes.  Of those same 135 relievers, Ziegler came in 5th in 2013 with a 3.67 WPA.  That’s really damned good, and it tells us a lot about how important Ziegler was in 2013.

But like RBI, WPA is ideally suited for recording good things that happened (such as making Paul Goldschmidt’s MVP case), rather than for whether a player was good at making them happen.  To have a high WPA for a season, a player needs to be put in high leverage situations.  Fortunately, this is baseball, and we have a way to handle that — the Leverage Index that measures the importance of the spots in which the player was asked to play.

Enter WPA/LI, which purports to measure the rate at which a player made the most of his opportunities.  It is with WPA/LI that we can answer the question of whether Brad Ziegler was particularly good at his job in 2013, rather than whether Ziegler was particularly good as a pitcher.  Of the 135 on the list, Ziegler comes in at 7th overall with a 1.64 WPA/LI.  Forget FIP — WPA/LI tells us that Ziegler was really good last season.  Here’s the list of other top WPA/LI relievers — it’s pretty impressive.

By now, you may be wondering about the effect that Ziegler’s stint as closer had on his season.  Although he went without a save between August 7th and September 5th, Ziegler was essentially the closer from the beginning of July until the end of the season.  Using “2nd half” statistics as a proxy (“2nd half” is smaller than the first half), it does look like that helped — 1.97 of Ziegler’s WPA came from the second half.  But Ziegler accumulated just 40% of his WPA/LI during that time frame (0.64 WPA/LI for “2nd half”), in part because as closer, he was routinely deployed in high-leverage situations.

So even before Ziegler took over as closer, he was having an important impact on games.  His 2012 numbers also bear this out — despite zero saves, Ziegler came in 13th among 136 relievers in WPA/LI in 2012 (1.34 WPA/LI).  Spending no time as closer did mean a lower WPA total — he tied for 37th among the same group with a 1.31 WPA.  But that discrepancy is exactly the point.  Ziegler is particularly helpful when he’s doing the job his profile is perfectly suited for: the situations in which ground balls are particularly helpful.

Brad Ziegler is an Amazing Ground Ball Machine

Casual baseball fans know that Ziegler is good at inducing ground balls.  But I was really taken aback at how good Ziegler is in that regard.

In 2012, Ziegler ranked 1st among relievers with a 75.5% GB%.  Second place was 67.3%.  Only 9 of 136 pitchers were above 60%.  The median was 45.7% (actually, the median was J.J. Putz — but his best skills are other things).  No one came close to Ziegler.

In 2013, Ziegler ranked 1st among relievers with a 70.4% GB%.  Second place was 68.4%.  Only 7 of 135 pitchers were above 60%.  The median was 44.1%.  No one came close to Ziegler.

For what it’s worth, the GB% leaders for those years among qualified starters were Justin Masterson (58.0%, 2013) and Trevor Cahill (61.2%, 2012).  Not even close to Ziegler, and it’s not even a sample size issue (there’s more potential for variance with smaller samples) — Ziegler has pitched 393.2 innings, and he’s got a 66.1% GB% for his career.

Here is a comprehensive list of qualified relievers that had a GB% over 70% between 2013 and 2002 (FanGraphs does not have GB% before that): Jonny Venters (72.5%, 2011), Cla Meredith (72.0%, 2007), and Jason Grimsley (70.4%, 2004).

To have a high groundball rate, it helps to throw from a low arm angle, and it helps to have a good sinker, and to use that sinker often.  Ziegler is all of those things.  Even among great groundballers, Ziegler stands alone — I think we all remember the GB prowess of Chad Bradford, but for the portion of his career for which FanGraphs has GB%, Bradford’s career percentage was 63.7%.  He topped out at 66.5% in 2008.

Brad Ziegler is Valuable — More Valuable than His Contract

There are dozens of baseball “events,” including singles, triples, and triple plays.  When the ball goes into play, there are no guarantees — it could find a hole, it could find the cheap seats, or it could get misplayed.  That’s part of why, from a pitching perspective, there is no more helpful event than a strikeout.

Ok, I lied just then.  There are actually several — run value tables convert events into averages based on the difference in run expectancies between the base-out state before the event, and the base-out state afterwards.  So according to a run table, there are a few pitching events that are more helpful than strikeouts — ones that involve runners already on base (with runners on base, run expectancy is already higher).  That includes pickoffs, and it includes fielder’s choices.  It also includes double plays.

The run value of a Grounded Into Double Play is -0.85.  In other words, the average difference between a the number of runs a team could expect to score, on average, and the number of runs a team could expect to score in that inning after event, is .85 of a run.  The run value of a strikeout is -0.30.

One way to get a lot of GDPs, as a pitcher: have a high groundball rate.  Another: have an unusually high proportion of situations in which GDPs are possible (man on first, less than two outs).  Ziegler gets double plays.  His total in 2013 (8, tied for 23rd) was not so electric, but that’s partly because he spent half the season as closer, and in general, during that time period he started his own innings (did not inherit runners on first).  In 2012, when he was used in his optimal role all season, Ziegler led all relievers with 21 GDP.  The next-best reliever had 12.

Again, one way to get a lot of GDPs is to get put in GDP-likely situations a lot.  But for crying out loud, Ziegler is really damned good at actually getting the GDPs in that situation, and the underlying numbers (his outstanding GB%) strongly indicate that it hasn’t happened by accident.

There’s real value there — Ziegler gets ground balls at a much, much higher clip than the best strikeout pitchers get strikeouts, but replace 21 GDP (21 x -0.85 = -17.85) with 21 strikeouts (21 x -0.30 = -6.30), and it looks like Ziegler may have saved more than a full win’s worth of runs beyond what an elite strikeout pitcher could have done, just in situations particularly favorable to Ziegler.

In reality, the value of having Ziegler available for those Ziegler-specific situations (runners on 1, 1 and 2, or 1 and 3, less than 2 outs) is much more complicated than that.  I’m not done with my research, so I can’t offer anything more at this time.  But if Ziegler is a really good reliever in situations when he starts his own innings (and I think his closer stint shows that to be true), he’s worth $5M a season.  Regardless of whether it’s a full win of run-saving through Ziegler’s additional helpfulness in key situations, or whether it’s less (or more) than a win, that’s all gravy for the D-backs — and it could mean that they’re getting the benefits of a pitcher that should be worth $10M or more.