Closer Handedness Splits
One of the ways a closer can lose his job is to not get hitters out from both sides of the plate. If a right-handed closer is only able to get RHH out consistently, teams will begin to stack the line-up with as many LHH as possible. Here is an in depth look at closer handedness splits.
I got the list of current closers from MLB Depth Charts a couple of days ago. Currently, several teams have shaky situations at closer, so this list will likely change before the beginning of Spring Training. I will try to give an update on it with the changes later.
To start off, here are the closers, their current team, throwing arm, FIP and TBF for LH and RH hitters. Additionally, I calculated the % of RHH the pitcher faced and the difference in their FIP. Only data from 2002 and later was used because the data at Fangraphs only goes back that far.
| vs LHH | vs RHH | |||||||
| Name | Team | Hand | FIP | TBF | FIP | TBF | % RHH | FIP Difference |
| Jim Johnson | Orioles | R | 3.47 | 531 | 3.82 | 565 | 51.6% | 0.35 |
| Andrew Bailey | Red Sox | R | 3.55 | 336 | 1.98 | 346 | 50.7% | 1.57 |
| Mariano Rivera | Yankees | R | 2.49 | 1318 | 2.61 | 1335 | 50.3% | 0.12 |
| Kyle Farnsworth | Rays | R | 4.27 | 1134 | 3.52 | 1448 | 56.1% | 0.75 |
| Sergio Santos | Blue Jays | R | 3.01 | 221 | 2.94 | 274 | 55.4% | 0.07 |
| Matt Thornton | White Sox | L | 2.71 | 809 | 3.86 | 1137 | 58.4% | 1.15 |
| Chris Perez | Indians | R | 4.30 | 404 | 3.91 | 520 | 56.3% | 0.39 |
| Jose Valverde | Tigers | R | 4.24 | 1070 | 2.87 | 1103 | 50.8% | 1.37 |
| Joakim Soria | Royals | R | 2.77 | 653 | 3.03 | 625 | 48.9% | 0.26 |
| Matt Capps | Twins | R | 4.18 | 745 | 3.66 | 956 | 56.2% | 0.52 |
| Jordan Walden | Angels | R | 2.08 | 171 | 3.51 | 147 | 46.2% | 1.43 |
| Brian Fuentes | Athletics | L | 2.82 | 730 | 4.12 | 1703 | 70.0% | 1.30 |
| Brandon League | Mariners | R | 4.58 | 682 | 3.41 | 759 | 52.7% | 1.17 |
| Joe Nathan | Rangers | R | 3.04 | 1029 | 2.40 | 1126 | 52.3% | 0.64 |
| Craig Kimbrel | Braves | R | 1.53 | 188 | 1.52 | 206 | 52.3% | 0.01 |
| Heath Bell | Marlins | R | 2.85 | 959 | 2.98 | 1012 | 51.3% | 0.13 |
| Frank Francisco | Mets | R | 3.67 | 684 | 3.53 | 738 | 51.9% | 0.14 |
| Jonathan Papelbon | Phillies | R | 2.51 | 913 | 2.71 | 816 | 47.2% | 0.20 |
| Drew Storen | Nationals | R | 2.63 | 222 | 3.77 | 313 | 58.5% | 1.14 |
| Carlos Marmol | Cubs | R | 4.03 | 841 | 3.54 | 1142 | 57.6% | 0.49 |
| Ryan Madson | Reds | R | 4.17 | 1235 | 3.28 | 1428 | 53.6% | 0.89 |
| Juan Abreu | Astros | R | 11.60 | 13 | 1.41 | 21 | 61.8% | 10.19 |
| John Axford | Brewers | R | 2.11 | 275 | 2.54 | 301 | 52.3% | 0.43 |
| Joel Hanrahan | Pirates | R | 4.00 | 667 | 3.06 | 809 | 54.8% | 0.94 |
| Jason Motte | Cardinals | R | 4.24 | 285 | 2.86 | 475 | 62.5% | 1.38 |
| J.J. Putz | Diamondbacks | R | 3.67 | 933 | 2.85 | 976 | 51.1% | 0.82 |
| Rafael Betancourt | Rockies | R | 3.83 | 974 | 2.48 | 1279 | 56.8% | 1.35 |
| Kenley Jansen | Dodgers | R | 2.00 | 144 | 1.59 | 183 | 56.0% | 0.41 |
| Huston Street | Padres | R | 3.64 | 769 | 2.69 | 979 | 56.0% | 0.95 |
| Brian Wilson | Giants | R | 3.17 | 647 | 3.00 | 718 | 52.6% | 0.17 |
Notes
- I would put no stock in Juan Abreu‘s results. Thirty-four hitters is not enough of a sample size. Sergio Santos, Kenley Jansen, Craig Kimbrel and Jordan Walden all have less than 500 TBF in their careers. Generally, it takes around 500 PA to get a good idea of a pitcher’s K% and BB%, which are major components of FIP. The trio’s results should be heavily regressed to the league mean. For the rest of the discussion, I will not be using these 5 in the calculations.
- The two left handed closers, Brian Fuentes and Matt Thornton, struggled against RHH compared to LHH. Teams have noticed this difference with Fuentes. Seventy percent of the hitters he faces are RHH. Because there are many good RHH that a team can use, I would stay away from any LH closers. Like the 5 pitchers with a small sample size, I will not use these 2 when doing futher calculations.
- Of the other 23 closers left, 5 have a split of over 1.00 FIP
Andrew Bailey: 1.57
Jason Motte: 1.38
Jose Valverde: 1.37
Rafael Betancourt: 1.35
Brandon League: 1.17
Historically, these 5 will be faced more RHH (~55%) than LHH. They have some ability to get out of inning if a couple of LHH string together some hits hits because of the number of RHH they will eventually face.
The two pitchers that I see red flags with are Andrew Bailey and Brandon League. Bailey will not have the luxury of working through a string of rough outings without getting pulled as the closer. The Red Sox will not have the patience with him losing games. The Yankees lineup especially could cause him a ton of issues since it is LH heavy. Brandon League‘s issue is that he is not a good reliever to begin with and the split just compounds the problem.
- Seven pitchers have a handedness split of less than 0.4
Jim Johnson: -0.35
Joakim Soria: -0.26
Jonathan Papelbon: -0.2
Heath Bell: -0.13
Mariano Rivera: -0.12
Frank Francisco: 0.14
Brian Wilson:0.17
This group of pitchers contain some of the best relievers in the game. A good reliever with a small split seems to be a good bet for long career as a closer.
Conclusion
Truthfully, I expected to find a little more useful data on handedness splits to help fantasy owners stay way from certain pitchers. I found only a few useful pieces of data (stay away from LH closers and Andrew Bailey), but that was about it. I think handedness splits may be more of a issue with setup men moving into the closers role. The setup men may be facing only a certain handedness of hitter to maximize their effectiveness. As a closer though, they will have to face whoever is due up in the 9th inning and the split will be exposed.
11
When did Kenley Jansen become the Dodgers closer? I know the fantasy community would love it, but the indications are that Javy Guerra will start as closer (and that there isn’t going to be a competition in spring training).
I had seen a suggestion, also, that Vinnie Pestano might not be second in line in Cleveland, as he has a lot of difficulties with lefties. Would you agree with that?
I just took the closers from MLBdepthcharts:
http://www.mlbdepthcharts.com/2011/09/los-angeles-dodgers-2011-12-offseason.html
I figured they would have more knowledge then me on the subject.
2. Wow, I didn’t know his numbers were that bad. 4.1 K/BB vs RHH, 1.4 vs LHH. Definitely an issue.
I think everyone is looking for the next Kimbrel and based on Jensen’s absurb K/9, people think he will automatically get the opportunity.
I agree with Simon–I believe Mattingly has said that Guerra will get the first crack at the job, but we’ll see as the spring unfolds.
I found the lefty-closer splits pretty telling. Thornton had a roller coaster ride last season and Addison Reed is waiting in the wings.
It’ll be interesting to revisit this article later in the year.
I think it’s a little misleading to conclude that Bailey should be outright avoided. He’s excellent against righties, in the top five on the list, so the fact that he has an average-ish FIP against lefties makes his split look like more of a red flag than it is. Is he an elite closer? No. Should we “stay away from [him]?” Maybe at his current ADP (which would have been helpful to see) but I’d sure rather have him than a lot of guys on this list, including some on your “safe” list like Frank Francisco and Jim Johnson.
Maybe you’re just saying that he’s more likely to lose his job than mediocre but steady guys like Francisco and Johnson but that has a whole lot more to do with the manager and the organization than handedness splits. If Francona were still in Boston then Bailey’s leash would be about a mile long but who knows with Valentine. That would have been a fantastic addition to this piece if you’re going to strongly recommend against Bailey, a detailed look at Valentine’s career in terms of bullpen usage and defined roles.
I totally agree, Aaron. Why avoid Bailey but not J.J. Putz? Bailey is better against LHB and RHB, it’s just that he’s *so* much better against righties that he has a large platoon split.
Aaron, I think this (and most of these types of advice articles) means to avoid Bailey relative to his current ADP/MDP or auction price. As the closer in Boston he’s looked at as a 2nd tier closer due to injury risk. However if he has problems against LHHs, and specifically the Yankee LHH, that’s more risk than most people are expecting. I’d still rather have him than Franky Francisco though. Of course there are dental procedures that I’d rather have than Franky. :)
I’m not sure we should worry about platoon splits very much at all in terms of closer performance (versus how good the closer is overall). If we look at just guys who were closers all of last year, it looks like they face RHH about 46% – 58% of the time, not a huge spread. Steven Goldman pointed out a few days ago that in the AL (the DH league) teams only use on average about 0.75 pinch hitters per game. Furthermore, to the limited extent that teams adjust their personnel to take advantage of closers’ platoon tendencies, is it that much greater than what they do when guys are pitching in a set-up role in the 8th inning?
I do know that it is harder for a LHP pitcher to become a closer, but that’s more due to managerial preference and the difficulty in finding another quality LHP to handle matchups earlier in the game. For example, see Gardenhire’s comments about Glen Perkins.
Untrue. Your causation is flawed. If teams started using LH closers with massive splits, opponents would start pinch hitting much more often. There is no static rule saying how often teams choose to pinch hit. They do so based on a calc. of how much edge it gains them, minus the penalty of losing the orig. player. Facing RH closers with small expected splits minimizes this edge and makes pinch hitting generally unattractive in most situations.
Is League really that bad? His ERA might have been a little lower than expected thanks to a low HR/FB, but at the same time his strand rate was pretty terrible as well. His K rates are low, but they seem artificially low based on his swinging strike rate. To me, he’s a solid closer with a good chance of being cheap come draft day.
I agree. I think the problem is that there’s just no “there” here. It totally seems like splits might be useful when evaluating closers but in reality I don’t think it makes a lick of difference.
“Truthfully, I expected to find a little more useful data on handedness splits to help fantasy owners stay way from certain pitchers.”
That’s really the bottom line, here. Stretching to make it seem like we found even a few guys to avoid is the problem.
This is good work. People often get too excited about a LH closer prospects without taking a close look at his splits (incl. in the minors to give proper sample size).
My only suggestion would be to show IP in your table above, as R/L splits take a long time to become predictive.
Relatedly, is there any chance FG could provide minor league splits? They seem to get minor league stats already form MLB, so I imagine getting/publishing split data would be feasible (?)
One final point: for low sample sizes (say, less than 300 IP), it may make sense to look at xFIP instead of FIP. The assumption being that there have not been enough innings for the HR/FB rate to stabilize and become predictive. This is obviously worth further study, but I do believe it makes sense in principle. For the Bailey example, his xFIP difference is only 0.85 (vs. 1.50 for FIP). He has given up 3 HR vs. RHB and 8 HR vs. LHB. These numbers are low enough that I would be much more comfortable using xFIP than FIP for future projections.
Craig Kimbrel: .01
awesome