Winning Pieces: How Texas can maximize their bullpen in 2016

There are few things in baseball more amazing than the evolution of bullpens. From not existing when the game began to the reason Kansas City won their title last season, the progress is stunning. Despite the leaps forward, there’s old thinking infesting the concepts of relief pitcher use. Teams cling to the set closer role, creating a rigid structure that lends itself to overuse of a team’s winning pieces. The Texas bullpen is no different.

To talk best about right now, we need to examine how we got here. Let’s turn back time a year to 2015. Manager Jeff Banister names Neftali Feliz his closer out of Spring Training, yet by May Shawn Tolleson was getting the save chances and Feliz was an afterthought (he’d be DFA’d in early July).

Fast forward to 2016. Tolleson walks a mile in Feliz’s shoes, struggling early before being replaced by current closer Sam Dyson. Tolleson isn’t even with the big club anymore after being optioned to Round Rock. There’s even reason to be concerned about Dyson, whose 2016 isn’t as dominant as his 2015.

Teams in general, Texas in specific, are asking the wrong question. They’re asking “Who should my closer be?” when they should be asking “What is the best way to use my relievers in a specific situation?” The save statistic is becoming like the pitcher win; an arbitrary stat that tells us little. The save itself is not important. It’s what the save represents, winning the game, that matters.

So what we’ll do here is lay out a few common situations, and give a rundown of which Rangers reliever should pitch in that situation. Obviously, not every pitcher will be available every day, so I’ll rank them in priority order from best suited to worst suited. For the purpose of this exercise, we don’t care about saves in any way. They will be a byproduct of success, a cherry on top of the winning baseball sundae.

(Note: With all due respect to Tony Barnette and Alex Claudio, I view them more as multiple inning guys who won’t see late inning work consistently unless other pieces aren’t available. So we’ll be focusing on the five main parts).

1. Nobody on, outs regardless(starting an inning most common situation)

  1. Jake Diekman
  2. Keone Kela
  3. Matt Bush
  4. Sam Dyson
  5. Jeremy Jeffress
Sam Dyson.242.308.358
Jake Diekman.157.231.217
Jeremy Jeffress.298.347.372
Keone Kela.140.260.279
Matt Bush.210.281.247

The ideal situation where an inning is fresh, Texas has a lead and needs to keep the other team from scoring. Diekman is the best in this situation, keeping hitters to a .157/.231/.217 line. Kela allows a lower average, .114, but his OBP and SLG are higher. Jeffress grades out as the worst, in this case, posting a .298/.347/.372 slash in these situations. This isn’t the last category where Jeffress performs worse than expected.

2. Pitching with men on base, any outs

  1. Jake Diekman
  2. Matt Bush
  3. Sam Dyson
  4. Jeremy Jeffress
  5. Keone Kela
Sam Dyson.237.307.315
Jake Diekman.146.266.278
Jeremy Jeffress.234.300.330
Keone Kela.263.316.667
Matt Bush.242.266.410

Once again Diekman takes home the crown for shutting down an in progress rally. A filthy .146/.266/.278 slash for opposing hitters makes him the go-to option. The problem is, past Diekman, the options get worse. Bush is the next best in this situation, but even he’s allowing a .242/.266/.410 line. The one guy that you should feel most uncomfortable in this situation is Kela. He’s bad to the tune of .263/.316/.667.

3. Pitching with runners in scoring position, any outs

  1. Matt Bush
  2. Jake Diekman
  3. Sam Dyson
  4. Jeremy Jeffress
  5. Keone Kela
Sam Dyson.241.302.310
Jake Diekman.191.346.400
Jeremy Jeffress.208.302.326
Keone Kela.429.5001.286
Matt Bush.114.139.206

A new champion approaches, as Bush slots in as the best option in these scenarios. With a sparkling .114/.139/.206 slash against, he’s lockdown when it comes to this situation. It’s worth nothing that he’s got only 10.1 IP in that situation, but that small sample size is sterling. Diekman, who’s been outstanding so far in this endeavor, has only 5.2 IP in this situation with a .191/.346/.400 line. Dyson and Jeffress are somewhat interchangeable here, and while Kela is a train wreck he also has a tiny sample size (1.1 IP).

4. High leverage situations

  1. Jake Diekman
  2. Jeremy Jeffress
  3. Matt Bush
  4. Sam Dyson
  5. Keone Kela
Sam Dyson.215.315.346
Jake Diekman.094.216.188
Jeremy Jeffress.164.259.181
Keone Kela.200.333.400
Matt Bush.179.233.321

I’m using Fangraph’s definition of high leverage for this, which they explain here in long form. Feel free to read it all and dissect, but it’s a lot of math and higher level concepts. I won’t bore you with it. Looking at the results, once again Diekman is tops. This season he’s posting a .094/.216/.188 line, establishing himself as the go-to guy if you’re in a bind. Jeffress gets a big time bump here up to two, over Bush and Dyson. It’s unexpected, given he’s scored so low already in the previous categories. Numbers don’t lie, though; in 21.2 IP he’s rocking a .164/.259/.181 slash. The refreshing thing here is none of the Ranger pitchers are terrible in this situation. Diekman and Jeffress are the clear best two, but if neither are available there’s little difference between the other three.

A few more assorted notes here regarding individual pitchers:

  • Jeffress is awesome with two outs and runners in scoring position, not surprising considering his high leverage prowess. He owns a .168/.288.232 line in 111 PAs over his career.
  • Despite the inclination to use him a lot, Kela on back to back days is dangerous. On no rest, Kela’s line is .310/.383/.476. It shifts to .207/.262/.322 when given a day off between appearances.
  • Meanwhile, Dyson has the best results when used on back to back days. Hitters are notching a .212/.277/.252 slash when Dyson doesn’t take a day off between outings. A valuable tool for any late-inning reliever.
  • While the Houston Astros continue to circle the drain, they’re glad to see Diekman leave H-Town. Diekman’s owned the Astros to the tune of an anemic .114/.204/.136 line. In eight appearances against the Astros this year, he’s kept Houston hitless while striking out 12 and walking two.
  • Bush’s sample size makes assessing him complicated. That said, one category he does dominate in is BB-Ref’s “Late and close” situation. Bush so far this year in 101 PAs, under those conditions limits hitters, to a .202/.257/.330 slash. It’s a great indicator of his success going forward, even when the other strong numbers are still in their infancy.

All this demonstrates that Texas is full to the brim with quality relievers who excel at different things. It’s a misuse of their talents to constrain them to roles that don’t take into account game context. For the Rangers bullpen, their best option is to cast off the roles and unlock the full potential of their relievers. If they do, the results will be positive.

A lot of wins, another AL pennant, and this time around a championship.

All stats courtesy of both Baseball Reference and Fangraphs.

Samuel Hale on EmailSamuel Hale on InstagramSamuel Hale on Twitter
Samuel Hale
When Samuel isn't displeasing you with his opinions about the Texas Rangers, he's trying to corral young broadcasters at UTA Radio. If you buy him pizza and high class chocolate milk, he'll probably be your best friend. Probably. He got to see Texas clinch a World Series berth in person, and sports cried when Pudge Rodriguez went into the Rangers Hall of Fame. He enjoys the Oxford comma and over tweeting.


  • Just an FYI:
    Any idea who, after Tolleson, has the most career professional saves? It is actually the player you stated that we would not look at in this article, Tony Barnette. Now, granted, his saves were in the NPB of Japan but it does appear more and more like we need to accept that players that play over there have some talent. The thing with Barnette is he was actually released back in 2010 by the D-Backs and not a single team wanted him then or for 6 YEARS afterward while he plied his trade in Japan. He was volatile there to say the least but eventually learned to control himself and now Texas has reaped the rewards. They certainly got a bargain for 2 years and 3.5 million, a deal that probably 95% of people scratched their head at in March.

    • So I feel like you’re not happy with my decision to not consider Barnette. I can’t deny that Texas did a great job of plucking him out of Japan, and I’m glad he’s here. He’s a valuable piece in this bullpen. That said, your question undercuts your point in relation to the article. I don’t care about how many saves a player has. I care how good they are. Saves are a byproduct of wins, which we tend to focus on as somewhat important. I don’t care how many saves one player has, I care about how many games the teams won. That’s what matters.

      Also I’m not sure exactly why you’re using this as a vehicle to express your displeasure about how we handle players from Japan. I never once said that his time/success in Japan was less valuable. My statement was that he is more valuable to the team as a multiple inning reliever, which the five players I profiled aren’t good at. Not everyone can be the stars. Barnette and Claudio offer a valuable service to the team. Taking how I approached them in the article as disrespect is all you. Just FYI.

  • Really enjoyed this piece Sam. Great work.

  • Good stuff, Sam! I love the stat breakdowns that lie below the surface discussed by talk radio and MLB Network.

    I’ve been trying to preach this idea to anyone who would listen for the last few years, going back to when Wash was overusing Nathan in non-save situations despite Nathan allowing a base runner almost every game. A bullpen should be about putting the best guy in for the specific situation, not about Bush always pitches the 7th, Diekman the 8th and Dyson the 9th.

    One of the most interesting stats you presented was that Dyson is more effective on the 2nd day in a row. This supports the theory that sinker-ball pitchers are better when slightly fatigued, because the sinker has more movement when they don’t throw it as hard. I’m not really sure how that can be logistically put into play on a regular bases, but it’s something I’m going to watch for the rest of the season.

    They all have individual strengths and weaknesses and as far as I can tell none of those situations are tied to what inning it is, beyond being late in the game. Jeffress has some high leverage ability, but he’s also been really bad against left-handed hitters this season, and situational use of the bullpen is key. Not just 8th inning guy pitches the 8th, or left-handed guy pitches to left-handed batters. I’m sure the Rangers have a stat department who tracks details even finer than the stuff you included here, but sometimes I wonder why they don’t put that information to use in the bullpen. They trust it enough to move the entire infield over to one side and leave a gaping hole on the other side. They trust it enough to make a line-up or batting order move despite it causing everyone else in DFW to scratch their heads.

    Everyone has been thinking this year that it seems like Diekman and Bush have been more effective relievers than Dyson (despite him only blowing 2 saves this season so far, he always makes me nervous.) These numbers you presented support that line of thinking and I’m wondering why Banister, despite saying that this bullpen has no defined roles, hasn’t experimented more with his bullpen, using a by-committee approach.

Leave a Reply