In 2012, Major League Baseball saw an explosion in volatility from the closer role, as over half of the league saw someone unexpected garner most of the teams’ total saves by the end of the year. With the continued focus on specialization in the game expanding the use of the bullpen more effectively, and even being sighted as a probable cause of the trend in increasing strikeout rates in the league overall; was there something more going on than just random bad luck with injury for closers in 2012?
Injuries always play a large part in the actual, or perceived, volatility in players’ roles. It’s not hard to name a number of pre-season established closers getting sidelined early in 2012: Brian Wilson, Joakim Soria, Ryan Madson, and, of course, Mariano Rivera. However, there still was a lot of non-injury related turnover in the closer role as well last year.
Three things have led me to wonder if there isn’t a subtle shift in the type of closer being employed by some teams in the league away from the prototypical fire throwing closer most exemplified today by Craig Kimbrel and his 50.2 strikeout percentage rate. The first is Jim Johnson and his 15.2% K rate last year. It’s easy to pick last year’s saves leader for the league and declare him the harbinger of change, but then, this off-season, we saw another interesting thing happen when the Los Angeles Dodgers decided to sign Brandon League for the next three years to be their closer. That’s Brandon League, former closer of the Seattle Mariners, and provider of a 17.9% K rate last year. That in and of itself isn’t as interesting, considering League has the, likely more perceived, value of experience successfully closing for a team at the major league level. However, the Dodgers already have a guy that fits the prototypical closer bill pretty well in Kenley Jansen and his near 40% K rate.
Finally, there was the recent news that suggested an unusual solution to the Tigers' deteriorating plan of letting would-be rookie, Bruce Rondon, and his reported triple digit velocity arm handle the contender’s closing duties. Specifically, the idea of Rick Porcello filling the closer role didn’t seem like that much of a crazy idea to manager Jim Leyland. Last year Porcello had the less than intimidating strikeout rate of 13.7%. While this move was then downplayed the next day, it’s enough to at least wonder if we’re not witnessing a change in the thought process around the league regarding what the acceptable model of a closer can look like.
Jason Collette over at Baseball Prospectus has already done some great work over the last couple of years highlighting the key categories one should look for in determining desirable targets for closing duties. His thought process is to find both the up and coming relievers, as well as the ones who should be most stable in regards to fantasy selection. He determined that there are five key things to focus on when looking for the best closer: 1.) Strikeout Rate, the more the better, 2.) Walk Rate, less is more, 3.) Home Run Rate, the deus ex machina for the opposition at the end of their rope, 4.) Ground Ball Rate, if the ball is on the ground then it can’t fly over the fence, (see #3), and maybe the most neglected aspect of a quality closer is number 5.) the difference in platoon splits.
Platoon splits are important, because no one wants to bring in another reliever to save their closer due to a bad platoon match-up. This is the primary reason why you shouldn’t be drafting Octavio Dotel with the view that he’s going to wind up with the majority of saves out of the Tigers' bullpen this year. Using these categories as a template, let’s see if the weighting of these skills is tipping at all away from the high K% type towards one of the other skill sets.
This chart below is a list derived from Baseball Prospectus’ 2012 pre-season depth chart of most desirable fantasy closers. I added each closer’s projected saves for 2012, and the actual number achieved in the next column. In cases where someone else finished with a significant number of saves unexpectedly, I added them below the anticipated primary closer with an indent. For example, Rafael Soriano is indented below Mariano Rivera. This shows just how much turnover there was throughout the season last year.
I also included everyone’s 2012 stats for their strikeout rate, walk rate, home run per fly ball rate, ground ball rate, and the platoon difference between lefty/righty OPS. These are relief pitchers and we’re using just one season of stats here, so a massive caveat applies in the form of small sample size. For example, a number of the larger platoon splits on this list actually were of the reverse nature when the pitcher’s career mark doesn’t show such a common characteristic. Still, I wanted to use the real data, because that’s how events are interpreted in real time during the season. If Soriano had served up a few more gopher balls to lefties that barely cleared the wall, then I doubt you would have heard a lot of talk about park adjusting his numbers. Instead, you would have just seen David Robertson getting another chance.
A few things jump right out without even looking at the names associated with the numbers: Double digit rates of walks looks like a problem, double digit home run rates on fly balls also looks like an issue. You get both and maybe you can see why the Pittsburgh Pirates decided to trade Joel Hanrahan for a haul and a half in return, and then let Jason Grilli take over.
Here we have the top ten closers sorted by K rate, which I would call the traditional closer model. All pretty solid names. Axford had some issues last year, but he still has his job heading into 2013. Ernesto Frieri pitched fantastically upon his trade to the Angels, but they brought in Ryan Madson this off season to take over the closer role whenever he’s healthy enough. Frieri's double trouble on walks and home run rates might be the culprit there, not to mention he has a high platoon split as well.
Sort by ground ball rate, what I think may be the new mold of closer gaining traction, and interestingly enough, we have the top two saves leaders from 2012 followed by the Rays' surprise closer of 2011 when Kyle Farnsworth produced a ground ball rate of 50.6%.
While this chart certainly doesn’t prove a tidal change, there are still many interesting trends. Even though Casilla eventually lost the closer gig, maybe the fact that he was originally handed the job isn’t as surprising when you see his ground ball tendencies in comparison to the field. The Cincinnati Reds' somewhat surprising decision to hand Jonathan Broxton a three year deal to take over the closing reigns might not be so farfetched if he can continue to keep the ball down close to his 2012 rate. The chart suggests that the Nationals' decision to pick up Rafael Soriano instead of just going with Drew Storen is questionable. Other than Storen’s split differential, it’s hard to put too many dents into his candidacy beyond last year’s post season. Now we'll see if there appears to be any change in the acceptance of this type of closer versus the past. Here are the top 5 relievers by ground ball rate in 2007 that had at least 10 saves.
That’s a very different profile of worm burners versus last year. In fact, outside of Bob Wickman’s 16%, the other four members of the list were all in the top tier of relievers in terms of the strike out rate. So if there’s anything to this supposition, then we should see one more attribute in teams’ bullpen construction, and that’s the traditional prototypical closer pitching as a setup man instead. Last year’s surprise dream pen, the Baltimore Orioles, is highlighted below.
Outside of Strop’s high walk rate, all of the other figures are pretty solid from the group. However, I’m guessing that five years ago, it would have been Darren O’Day closing games for the Orioles instead of Johnson. If you’re still questioning whether or not there’s a conscious thought process that can be gleamed from these numbers, then I’ll point you to the fact that, out of that group, the player who received the most save opportunities other than Johnson was Pedro Strop, despite his high walk rate. There could be many seemingly random reasons to explain the opportunity and success for each closer that fits the new model. However, I find it pretty fascinating that, in an era of significantly rising strikeout rates, we also seem to have a trend of more pitch to contact ground ball relievers getting the opportunity and finding success in the closer role. Every manager or team isn’t going to follow the philosophy that Baltimore appears to have taken last year to heart, but assuming some validity to this thesis poses some interesting potential strategies when it comes to picking the least volatile closers.
The relevance to fantasy baseball should be readily apparent: avoid closers that project to have double digit home run to fly ball and walk rates, as this can translate into higher likelihood of losing one’s job. Target closers that project to have high ground ball rates, as well as ones that should have high strikeout rates. There may be a growing acceptance of two types of applicable closer in the league, and the new model type is probably going to come significantly cheaper to acquire than the traditional flame thrower. Remember their splits, since no manager wants to bring in another reliever, if possible, to avoid a ninth inning platoon mismatch. Here is a list based on Steamer projections for 2013, of the top 5 relievers projected to get saves this year, ranked by ground ball rates.
Jim Johnson and Fernando Rodney are established, and are not likely to come cheap after last year’s performances. League, however, may come at a discount due to his struggles in Seattle before being traded, and Bobby Parnell may be able to keep his role with a solid start before Frank Francisco comes back from injury. Considering Francisco’s tendencies with both home run and walk rate, Parnell should be considered a prime candidate to win that job outright.
Mark Melancon could be an interesting late round $1 pick. Jason Grilli looks to have the skills to be a successful closer with a 27.5% projected K rate and a 2.90 projected ERA for this year, but his walk rate is a bit high, almost touching the problematic 10% level. Melancon is one of the four players coming back in the Hanrahan deal with Boston. The last time he pitched in the NL Central he recorded 20 saves with the Astros in 2011. His home run rate was a bit high that year at 11.1%, but his ground ball rate was an excellent 56.7%. Pitching in Pittsburgh should help reduce that home run rate a bit, which could make Melancon a good fit for our new model closer type if he gets the chance.
Suddenly, Rick Porcello closing doesn’t seem so crazy at all: Steamer projected 51.8% ground ball rate, 6.2% walk rate.