Archive for bullpen

Last night, the Chicago Cubs lost their seventh game of the season, falling to 6-7 in the process. In 2016, the Cubs didn’t lose their seventh game of the year until May 11, so this start represents a departure from last season’s 103-win, World Series champion.

That doesn’t mean there’s cause for concern, though. The team isn’t hitting all that well, but Javier Baez, Kris Bryant, Willson Contreras, and Addison Russell are all likely to improve upon their early-season lines. The offense, as a whole, ought to come around. The rotation is pitching well — Kyle Hendricks has had a rough start, but Brett Anderson has been a pleasant surprise — so nothing really to worry about there. The bullpen, though, might be worth a closer look.

The Cubs’ bullpen has put up a 4.10 ERA, which isn’t very good, and a 4.46 FIP, which is even worse. Through 41.2 innings, the team’s relievers have been slightly worse than replacement level as a group. They’ve blown four saves already, tied with the Marlins for the most in the National League — and for whatever shortcomings the save possesses as a metric, having a bullpen blow a lead four times in 13 games isn’t good.

On the whole, the bullpen has been bad. Has it been team-wide issue, though, or the product of a few poor performers? Let’s see.

New closer Wade Davis has been good. Koji Uehara has been fine. Carl Edwards, Jr. looks like he’s ready to step into a more prominent role. Mike Montgomery hasn’t been great, really, but he’s covered a lot of innings adequately. Hector Rondon seems like he’s probably back after a rough 2016 season, and Brian Duensing has only pitched two innings. If there’s blame to had it is coming from two guys: Justin Grimm and Pedro Strop. While Grimm’s start hasn’t been great, he’s also not expected to be more than the sixth- or seventh-best reliever on the team. Looking even closer, Grimm’s poor pitching hasn’t really even cost the Cubs. Here are the same pitchers by win-probability statistics.

Of Grimm’s six appearances, three have come in very low-leverage situations, one has been roughly neutral, and two have occurred in high-leverage situations. On April 10, Grimm came on in the seventh with the bases loaded and no outs with the Cubs holding a one-run lead. A pop fly and a double play later, the Cubs’ chances of winning moved from 39% to 79%, making up for Grimm’s few poor performances in low-leverage outings, as well as another appearance (April 14) during which he allowed two inherited runners plus a run of his own en route to turning a 2-1 lead into a 4-2 deficit.

Read the rest of this entry »

Last fall, Andrew Miller pitched 19.1 innings in 15 postseason games. Extrapolated out to a full 162-game season, that would equal 209 innings. That is basically an impossible load for the modern reliever. Consider, for context, that the only relief pitcher since World War II to top 170 innings is Mike Marshall, who did it twice, including 208.1 innings in 1974. The last major-league reliever to top 150 innings was Mark Eichhorn in 1986. Since the strike in 1994, Scott Sullivan’s 113.2 reliever innings in 1999 represents the majors’ highest mark. No reliever has crossed the 100-inning threshold in the last decade.

All indications suggest that, if a bullpen revolution is really to occur in baseball, it isn’t simply going to be a matter of a single pitcher recording a ton of innings. Exhibit A in an argument against the reality of an Andrew Miller Revolution is that Andrew Miller himself is only on pace for around 88 innings this season. What Miller did in the postseason last year is likely to remain a product of the postseason.

There might be a different sort of bullpen revolution occurring in Ohio, though. Only, this one isn’t happening in Cleveland. Rather, it’s unfolding about four hours southwest on I-71. If there is a bullpen revolution brewing, it’s happening in Cincinnati.

Consider: if a team were starting a bullpen from scratch and trying to create an ideal bullpen, they would likely abide by three basic principles.

1. Put the best pitchers in the most important situations.
2. Ignore the save statistic.
3. Use the best pitchers for multiple innings.

Let’s evaluate the Reds’ bullpen usage this season so far by each of these three principles.

Read the rest of this entry »

You’ve probably heard about the Troy Tulowitzki trade. The one that turned the Toronto Blue Jays’ already mighty offense into a juggernaut incapable of losing. The trade for David Price made some news, too. A team with a few mid-rotation starters but lacking an ace at the top and depth at the bottom was reinvigorated with the acquisition of one of the best pitchers in Major League Baseball, simultaneously providing that much-needed ace and allowing the rest of the rotation to fill out the remaining spots nicely. Those were the major moves — the earth-shattering, capture-the-attention-of-two-nations moves. The Blue Jays made other moves, too, though, and getting LaTroy Hawkins in the Tulowitzki trade, adding Mark Lowe, moving Aaron Sanchez to the bullpen, and officially naming Roberto Osuna as closer has strengthened what was once a weakness. For months, the Blue Jays struggled to close out games, but the bullpen has been lights out in the second half and looks ready to compete in October.

The last time I looked at the Blue Jays’ pitching issues, it was late June, the team was fourth in the division but had a solid 50/50 shot at making the playoffs. In the six intervening weeks, the Blue Jays have moved from wild-card hopeful to near playoff lock with more than a 90% chance of making the playoffs and 57% chance of winning the division. In late June, the rotation had at least one hole, and the bullpen was still struggling. Brett Cecil was experiencing difficulties as a closer, and the team had recorded just one more Shutdown (40) than Meltdown (39) on the year. As the first half drew to a close, the team had as many saves (14) as blown saves (14). Beginning with the change in closer six weeks ago, however, the Blue Jays have transformed their bullpen.

Read the rest of this entry »

The number of innings a team’s bullpen throws over the course of the season has less to do with the performance of the bullpen than the performance of the starters. Teams with starters pitching deep into games rely less on relievers, leaving the bullpen well-rested and allowing the manager to leverage a team’s best relievers in more important situations. A great bullpen might cause a manager to pull his starter at the first sign of trouble, creating more innings for the bullpen, but for the most part, the starter will pitch as many innings as possible and the rest is left for the bullpen. Once the relievers are called upon, it is the manager’s job to divvy out appearances and prevent overuse. So far this season, the Boston and Tampa Bay rotations have put their bullpens in trouble and St. Louis also appear to be in danger of wearing out their core arms — points which I’ll address momentarily.

First, let’s consider performance. In unsurprising fashion, the Kansas City Royals’ bullpen has produced the lowest ERA among all major-league bullpens in 2015. Their relief corps was a featured strength as the team made it to the World Series last year. From 2012 to 2014, the Royals bullpen WAR of 17.7 is more than two wins greater than the second-place Atlanta Braves, and the bullpen is off to a great start in 2015 (even if their 3.35 FIP does not quite match their sterling 1.56 ERA). The graph below shows every bullpen’s ERA and FIP, sorted by the former.

Read the rest of this entry »

Last week, Colin Zarzycki reviewed the Milwaukee Brewer bullpen on RotoGraphs. The projected bullpen included the likes of Jim Henderson, Brandon Kintzler, Tom Gorzelanny, Michael Fiers, and Will Smith, among others. The 12 names on the depth chart combined for 12.009 years of MLB service time, with Gorzelanny accounting for half of that total. The depth chart contained some intriguing upside, but was certainly lacking in experience. Enter Francisco Rodriguez, again.

Read the rest of this entry »

The Mets’ farm system rates in the top third of baseball due to a plethora of pitching prospects. In addition to Zack Wheeler, Rafael Montero and Cory Mazzoni in the upper minors, the organization’s Florida State League affiliate has the minor leagues’ best rotation. It features Noah Syndergaard, Domingo Tapia, Luis Mateo (who is on the disabled list with an elbow injury), Michael Fulmer (DL – Meniscus) and Hansel Robles. The Savannah Sand Gnats are stocked with intriguing arms as well. Mike Newman recently discussed Gabriel Ynoa and Rainy Lara with his audience, and I mentioned the resurgence of Steven Matz several weeks ago. Read the rest of this entry »

The goal of an ideal bullpen is to relieve the starting pitcher and effectively shut the opposition down to the point that they stand no chance of coming back. We hear so often the idea that playoff spots are won and lost in the bullpen, and from reading numerous blogs, one might think that 26 of the 30 major league teams have “the worst bullpen in baseball.” In my eyes, the best type of reliever is one who limits his walks, gives up as few home runs as possible, and strikes out as many batters as he can. These are the controllable skills for a pitcher, or, in other words, game outcomes that have nothing to do with the defense. Which bullpen comes closest to this ideal reliever? Well, with the update to the team pages here at Fangraphs, we now have the capability to check.

For starters, which relief corps posted the best strikeout rates? The Cubs (8.68), Yankees (8.66), and Dodgers (8.62) vastly stood out from the rest in this category. The impressive Rays bullpen came in fourth place at 8.03, narrowly ahead of the Reds at 7.97. At the bottom of the league were the Pirates (6.45), Orioles (6.80), Tigers (6.83), Rangers (6.85), and Cardinals (6.87). The Cardinals may have stuck around for a while in the playoff picture, but their season was not as statistically sound as the results would indicate. And I don’t think anyone will argue that the Pirates, Orioles, and Rangers have very poor pitching staffs, or that the Tigers were an all-around disappointment this season.

When we move to walks, we see that the Dodgers (3.22) once again find themselves in the top five, at the number one spot, but their “colleagues” are different. The DBacks come in second at 3.24, followed by the Twins at 3.26, White Sox at 3.31, and Indians at 3.33, meaning that the bullpens that struck out plenty of batters were not too tremendous at limiting walks. At the bottom of the heap we once again find the Orioles, Rangers, Pirates, and Tigers, with the Mariners now mixed in. See any trend emerging?

Moving to K/BB ratio, we get a mixture of the K/9 and BB/9 leaders, as the Dodgers (2.68) lead the Cubs (2.57) and Yankees (2.46), with the White Sox (2.38) and Diamondbacks (2.32) rounding out the top five. Guess who is at the bottom? Yes, the Orioles, Tigers, Pirates, Rangers, and new-found friend the Mariners. The Orioles bullpen had an ugly 1.40 K/BB. The Mariners at least registered a K/9 above 7.0, they just had issues limiting walks.

How about home runs? The Phillies, whose bullpen has been lauded for the entire season, finally break into the top five, finishing first with a 0.69 HR/9. Led by Brad Lidge and his unsustainable 3.9% HR/FB, the Phils narrowly edged the Blue Jays (0.70), the familiar Dodgers (0.73), and the Mariners and Athletics (0.78 each). When we put it all together, in the form of FIP, the ERA equivalent of controllable skills, the Dodgers led by a wide margin over all others at 3.47.

The Yankees, Phillies, and Blue Jays then ranged from 3.82-3.85. Controllable skills are not the only measure of success, but they are very important, and nobody realistically came close to the Dodgers in 2008. The Blue Jays did have a big ERA and FIP differential, as they posted the lowest bullpen WHIP at 1.25 and highest bullpen LOB rate at 79.8%. Their FIP of 3.85 suggests their 2.94 ERA was a bit lower than expected, but elite relievers are known for being able to consistently post high strand rates, low BABIPs, and other luck-based indicators that tend to normalize for starters. Based on these results, it seems the best controllable skills in the NL belong to the Dodgers bullpen, while the Yankees and Blue Jays topped the junior circuit.

For those unaware, David revamped the team pages today to provide some incredibly useful information for readers and researchers. Now, by clicking the link at the top of the page, you can sort the same statistics found on the league leaderboards for individual teams, as well as sort the overall team totals in several different categories. With the dominant bullpens of the Phillies and Rays, on display in Game One of the World Series last night, in mind, I decided to take a look at the pitch repertoires and frequencies for the bullpens of each team to see how they stacked up.

The Phillies bullpen threw an aggregate 52.5% fastballs, the lowest amount of any bullpen. Their relief corps of Brad Lidge, Ryan Madson, JC Romero, Scott Eyre, Clay Condrey, and Chad Durbin, rely moreso on offspeed pitches than all others. On the flipside, the Texas Rangers throw 69.5% fastballs, by far the most. As Dave noted earlier, perhaps a lower percentage of fastballs could help their 5.51 bullpen ERA moving forward.

The Padres bullpen threw the slowest fastball, on average, at 89.9 mph, with the White Sox leading the league at 92.5 mph. The White Sox led very narrowly over the Red Sox and Diamondbacks, both at 92.4 mph, and the Tigers at 92.3 mph. Matt Thornton is the only White Sox reliever in the top ten in the junior circuit, though Bobby Jenks does join him in the top thirty. The Red Sox, however, have Jonathan Papelbon, Manny Delcarmen, Craig Hansen, and David Aardsma in the top thirty.

The Cubs threw sliders most often, at 33.1%, while the Orioles threw 11.6% curveballs, the highest frequency of any team. Moving to changeups, the Cardinals ‘pen threw them just 3.1% of the time, while the Royals threw at a percentage over five times higher, at 16%. And, lastly, the Angels and Royals led in splitter frequency, both chiming in under 5%.

No team threw 10%+ for four different pitchers, with the Rays and Rangers coming closest. The Rays threw 66.4% fastballs, 14.8% sliders, 8.6% curveballs, and 9.1% changeups. The Rangers threw 69.5% fastballs, 11.9% sliders, 7.5% curveballs, and 8.0% changeups. Both teams rely moreso on the fastball than anything else, but they did come closest to throwing four pitches ten percent or more of the time. This is just one of the neat things we are now able to do with the updated team pages.

Following a recent sweep by the Cubs at Wrigley Field, the White Sox last night got revenge by completing a sweep of their fellow Chicagoans on their own South Side turf. The 5-1 win gave the White Sox a 46-35 record and kept them in first place by 1.5 games over the Minnesota Twins. One of the key ingredients to their success this year has been the bullpen. Joe Morgan discussed how valuable they have been this year and the numbers back it up.

The White Sox bullpen currently has an MLB-best 2.61 ERA and has done so in 213.2 innings, a number lower than everyone except the Anaheim Los Angeles Californian Angels of Los Anaheim. The relief corps has not been called upon an exorbitant amount of time thanks in large part to the White Sox starting staff and their league leading 50 quality starts. That means in 62% of their games so far the bullpen hasn’t even gotten to rear its head until the seventh inning or later.

Overall, the White Sox have an MLB best 3.39 team ERA, and have walked just 227 batters—2.8 per game—which ranks second only to the Twins and their 198 free passes. Here are the five major components of the White Sox bullpen and their numbers:

Bobby Jenks: 32 G, 32.1 IP, 7 BB, 19 K, 1.95 ERA, 1.11 WHIP
Scott Linebrink: 36 G, 33.0 IP, 6 BB, 29 K, 1.36 ERA, 0.82 WHIP
Matt Thornton: 33 G, 31.1 IP, 9 BB, 40 K, 2.30 ERA, 0.80 WHIP
Boone Logan: 33 G, 28.1 IP, 6 BB, 29 K, 2.22 ERA, 1.09 WHIP
Octavio Dotel: 37 G, 36.2 IP, 19 BB, 53 K, 3.19 ERA, 1.36 WHIP

The games played numbers speak volumes toward the lack of overusage or abuse. In fact, you won’t find any White Sox relievers in the league leaders here on Fangraphs until you venture to the second page of players. Looking at all relievers with at least 20 IP—in order to get rid of those who wouldn’t qualify for relief rankings—the ChiSox have four in the top 35 in WHIP as well as three with K/BB ratios above 4.40; the other two chime in at a very respectable 2.79 and 2.71.

One aspect of bullpens I enjoy investigating involves how they were put together. With this team:

1. Bobby Jenks: selected off Waivers from Angels on 12/17/04
2. Scott Linebrink: signed 4 yr/19 mil deal this offseason
3. Matt Thornton: traded from Seattle for Joe Borchard on 3/20/06
4. Boone Logan: drafted 20th rd in 2002
5. Octavio Dotel: signed 2 yr/11 mil deal this offseason

Querying and exporting all relievers with 20+ innings allowed me to find some current averages. The average LOB rate for such relievers is 76.80%. Of the five White Sox discussed here, all but Dotel’s 71.7% come in at 78.5% or higher; Linebrink even sits at 92.1%.

In terms of WPA, the average for relievers with 20+ innings is 0.375. So, if we were to build a bullpen consisting of five “average” relievers (with 20+ IP), their WPA would amount to 1.88. Additionally, the average WPA/LI for this parameter is 0.282; five average pitchers with regards to this WPA/LI would combine for 1.41 context-neutral wins. The White Sox five currently have a 4.08 WPA and 3.48 WPA/LI, meaning they are 2.2 WPA wins above an average bullpen and 2.1 context-neutral wins above an average bullpen.

They have allowed the fourth least amount of total bases and have the second lowest OPS against at .687—only Oakland’s .661 ranks higher. Their bullpen may not sustain performance like this all season, but South Side fans should hope they do, considering their relievers have contributed the most win probability success to the team.

The DL bound Joe Borowski has apparently had soreness in his right triceps that the Indians staff decided to let him pitch through for weeks. I guess that helps explain his 18.00 ERA and super slow fastball.

Despite leading the AL with 45 saves last year, he also led it in blown saves with 8 and his ERA of 5.04 was by far the highest of any closer last year. He did have pretty decent peripherals though, with a K/9 of nearly 8 and a BB/9 of 2.33. His season was mainly marred by an extremely high .348 BABIP and a pretty sub par LOB% of 68.4%. Typically both of these will regress toward the league averages in following years.

The main problem with Borowski is that he’s a bit of an extreme fly ball pitcher and will be prone to giving up home runs. Last year he allowed 9, which was just 4 home runs shy of Al Reyes‘ reliever leading 13.

So the question is, if Borowski is out for an extended period of time, who will get his job?

Rafael Betancourt is the obvious choice. He had a gaudy K/BB of nearly 9 last year, and a 1.47 ERA in 79+ innings of work. He also entered games in high leverage, non-save situations last season with a gmLI of 1.78. Borowski entered games in slightly higher leverage situations with a gmLI of 1.94. Even though he has an ERA over 5 this year, his strikeouts and walks are almost identical to last year, so there’s probably not much to worry about.

Rafael Perez had nearly as great a season as Betancourt did in 2007. He struck out over a batter an inning and had a BB/9 of just 2.23, which led him to a career low 1.78 ERA. This year his ERA is over 8 and he’s allowed a few more walks then he should. But, he’s only pitched 4 innings this year and allowed his only 4 earned runs in one of his appearances so definitely take those numbers with a grain of salt.

Masa Kobayashi has had a solid season so far with a with a 2.08 ERA and 3 strikeouts and 1 walk in 4+ innings. He amassed over 200 saves in Japan, and technically has more career saves than anyone on the Indians. Kaz Sasaki and Takashi Saito had zero problems stepping right into the closer roles when they first arrived in the majors.

Finally, there’s Jensen Lewis who last year had a 34/10 strikeout to walk ratio in 29 innings pitched with a 2.15 ERA. This year, it appears to be more of that same excellence, though his K/9 is down to 8 from 10, but small sample size and all that….

Here’s how they’ve been utilized this year so far, according to gmLI:

Joe Borowski – 1.86
Jensen Lewis – 1.61
Rafael Perez – 1.48
Rafael Betancourt – 1.27
Masa Kobayashi – 0.23

Borowski should have the highest gmLI as the closer and does. Oddly enough, Jensen and Perez have entered the game in higher leverage situations than Betancourt has, despite his reputation as a fire putter-outer. Kobayashi is clearly not used in confidence, but there are only so many high leverage situations and he just hasn’t had the opportunity with such an overall excellent bullpen.

Judging from tonight’s game, Betancourt was being “saved” to close out the game and was not utilized in a tie game in the 8th. Nor was he used in the 7th with the Indians up 2-1 and the Red Sox threatening with a man on 1st and 2nd.

So for the moment, it looks like Betancourt is going to get to close out games, but the Indians have a pretty stellar bullpen and I think if needed any of these guys could handle the closer job.

When Borowski comes back and if he’s healthy, chances are he’ll be given a shot to reclaim his role as closer. And while he’s not as bad as he’s often made out to be, he really is one of the worst options in what appears to be a very stacked Indians bullpen.