Breaking Down Jason Grilli’s Struggles

The poor recent outings from Jason Grilli have only highlighted his shaky performance in 2014.  The Pirates closer hasn’t been good by either traditional or sabermetric statistics.  He’s 0-2 with a 4.08 ERA, while blowing 3 saves out of his 14 opportunities. In 17 2/3 innings, he has struck out 17 (8.66 K/9) and walked 10 (5.09 BB/9).  Grilli hasn’t posted that low a strikeout rate or that high a walk rate since coming to Pittsburgh.  Those numbers, combined with a sky high 1.53 HR/9, have resulted in a FIP of 5.24 and a xFIP of 4.73.

His work this year is still a “small sample size”, and there’s some reason to expect better performance going forward.  So far, 13.0% of  fly balls off of Grilli have left the park, which is above the league average and Grilli’s career rate.  Conversely, the 37-year-old’s BABIP is due to regress in the other direction, as it currently sits at .267.  Additionally, his batted ball profile looks like it is trending the wrong way, as he has a career low 27.1% ground ball rate, and a career high 47.9 fly ball rate.  He has never had great looking batted ball stats, but it didn’t matter nearly as much when he was striking out over 36% percent of hitters (as he did in 2012 and 2013).

The real problem for Grilli is his declining strikeouts and increasing walks.  Finding out the root cause of these problems is not as clear-cut, as his plate discipline stats don’t really hold any answers.  The right-hander isn’t getting quite as many swings and misses as the previous two seasons, but he is getting more first-pitch strikes. Overall, Grill looks to be a mixed bag this season. He is doing some thing slightly better, but other, more important things, slightly worse.

If you break it down by pitch type, Grilli has been much less effective with both his fastball and slider, pitches he throws over 90% of the time.  Using Fangraphs Pitch Values (which determine how man runs a pitch was below or above average per 100 pitches) both Grill’s fastball and slider are not only less valuable than last season, but are both below average.  His fastball went from being worth .86 runs per 100 pitches in 2013 to -.35 this season. His slider went from 1.55 to -.91 by the same metric.

Since he’s not doing anything different in terms of pitch selection or location, I decided to look at his fastball velocities. He hast lost about half a tick off his four-seamer from last year, dropping from 93.3 mph to 92.7 mph.  That’s not huge, but velocity loss can have different affects on different pitchers.  I took a look at some graphs of his velocity over the last two seasons. Here are Grilli’s fastball velocities from 2013, moving (counter-intuitively) chronologically right to left.

chart (2)

So you can see that he dropped off sharply after his late-season DL trip. In fact, he registered 25 pitches over 95 mph before the injury, and ZERO afterwards.  His average velocity before the injury was 93.6 mph, afterwards it dropped to 92.2 mph. So there is an even larger velocity decline if you are comparing this year to pre-injury 2013, when Grilli was at his best. After coming off the DL in September last season, Grilli struck out eight and walked three while allowing four runs in 7.2 innings, definitely a step below his dominant form.

So maybe he does need a faster fastball to get good results. His velocity chars in 2013 provide a little more evidence of this.  Again his most recent outings are on the left, with earlier results on the right.

chart (1)

Ignoring the mis-classified pitches in his second outing, Grilli was close to 2013 velocities up until the beginning of June. He hasn’t hit 95 mph since June 1, and needless to say, the results haven’t been good this month.

Maybe this is just a blip on the radar and his velocity will stabilize going forward.  Like I wrote before, the problems for Grilli are clearly the poor strikeout and walk numbers.  But it’s tough to point to a reason for those bad numbers.  He really isn’t doing much differently, with his velocity being a possible culprit.  It could be that he has declined slightly in a number of areas, with the result being he’s much less effective overall.

But I wouldn’t expect him to continue to pitch this poorly all season.  While he’s been below replacement level thus far in 2014, both Steamer and Zips think he’ll be better going forward. Over the rest of the season, they project him to have a 3.18 and 3.17 FIP respectively. I think his results will be closer to those projections than his current numbers.  But with relief pitching being so unpredictable, and Grilli dealing with injuries and aging, a sudden decline is a very real possibility.

Stats courtesy of Fangraphs, and current through 6/18/14.

Graphs courtesy of Baseball Savant.

Author: Max Fogle

Share This Post On
  • bucsws2014

    I was at today’s game. He was lucky to only give up the HR and basehit. Every batter that connected ripped into those pitches. The cracks were audible and loud.

    We all know Hurdle is going to be loyal to Grilli. But at what cost to the Bucs’ diminishing playoff hopes?

  • juniorkrz

    Is it time to start examining a potential trade for a Huston Street in SD, who appears to be available.

  • R Edwards

    If the Pirates hadn’t so impatiently and impulsively traded Morris away for a draft pick, Morris could have been a viable closer candidate. Did he have his struggles this year? He did. But, given his history in the system and his obvious stuff and velocity, he was a strong candidate to turn that around – which he is doing now in Miami.

    If someone had to go in the bullpen, it should have been Grilli or Gomez.

    I don’t what you can do with Grilli now. If he blows one more home game, he will get the Oakland treatment of Jim Johnson. He needs to taken out of the closer role, until he proves he can get batters out consistently. Until then, I would alternate Melancon/Watson as closer – depending on the game situation. If Grilli is physically not right, put him on the DL and bring up Oliver or Mazzaro.

    • moose7195

      Yeah, the icing on the cake is that they used the pick they got for Morris on Conner Joe.

    • bucsws2014

      Morris had his chances here and failed miserably. Too many walks. Too few stranded runners.

      Give it up. If you like him, be happy he’s doing well in Miami. He wasn’t doing well here and gave no indication he would.

    • Andrew

      In what world would Bryan Morris be a closer candidate? Morris has a career 5.6% K% – BB%, that is one of the worst in baseball over the time he has been in the league. When Grilli was at his best he had 30% rate, if you think Bryan Morris has achieved some new talent level based on 34 batters faced in Miami, you are doing it wrong.

      • moose7195

        On the team with the most blown saves in the league, I’d say planet Earth would be that world. No one else has distinguished themselves. Morris doesn’t deserve the shot over guys like Melancon or Watson, but to eliminate him from the conversation entirely implies that we have a competent closer, which we don’t.

        • Andrew

          Bryan Morris has faced 427 batters in the major leagues, he doesn’t get strikeouts and lacks fastball command which leads to walks, he looks good is stretches when his BABIP is low, but BABIP is not a skill. I don’t understand why people think a reliever who has the 2nd worst K%- BB% since 2012, would be an effective pitcher to put into high leverage situations.

          I never see anyone argue for Jared Hughes or Vin Mazzaro to pitch in high leverage spots, just because a guy throws hard doesn’t mean he has good stuff.

          http://www.fangraphs.com/leaders.aspx?pos=all&stats=rel&lg=all&qual=y&type=1&season=2014&month=0&season1=2012&ind=0&team=0&rost=0&age=0&filter=&players=0&sort=9,a

          • moose7195

            I’m not denying any of this. I’m saying that you can’t count the guy completely incapable when no one is proving proficient in those high leverage situations. You don’t know how he’ll react to those high leverage situations, it changes some players. I fully believe the sabremetric belief that the 9th in just another inning, but I don’t think Hurdle buys it, and by extension his players don’t either. Maybe he’ll step up, who knows?

            And I clearly endorse Melancon, the pitcher with the lower velocity, before Morris. Please don’t put words in my mouth

            • Andrew

              Fair point, I preferred not to have defined role, if the
              middle of the order is up in 8th, I’d use my best pitcher then which would be
              Melancon then Watson.

          • bucsws2014

            I’ll argue for Hughes to pitch in high leverage situations. He’s got 100% strand rate this year. Thus my complaint is giving him his own inning to start instead of using him to put out everyone else’s fire :-)

    • Max Fogle

      I don’t think Morris would be a viable closer candidate if he was still here. He’s projected for between -.5 and 0 WAR the rest of the season, and I would much rather trust the projections based on years of mixed performance than on a couple good weeks.

      And at the time of the trade, the Pirates needed a roster spot. It was pretty incredible they got any value for a reliever who had been among the worst in the league over the past two seasons.

      • R Edwards

        Again, you are assessing him based on his one season in the league. Watson wasn’t so great in his first two years either. Go back and check his stats for his first two years in the league. If you were in charge of team, he would have been traded or released.

        Last year, Morris’ ERA was 3.46 and hitters BAA was .243. Far cry from being even remotely close to being one of the worse relievers in the league. Quit the hysteria and drama.

        If a reliever had to go, again it should have been Gomez, Grilli, or even Pimentel. We threw Morris away – a reliever who they traded for and developed and was only 27 – for a draft pick at a position of precious little need in the system.

        This bullpen was now surrendered 15-16 runs in the past 6 games, and that does not include the 2 given up by Snider. the team and the bullpen are both now weaker, because of that dumb trade.

        • Max Fogle

          I get your argument about his potential, but he clearly was among the worst relievers in the league. He posted a WAR of -1.1 last season. His FIP was 4.89 last season, and is at 4.58 as of today. His strikeout rate last season was 5.12 per nine innings last season. He was pretty bad. The fact that he’s 27 (which is past peak age for pitchers) is not a good thing all. And Connor Joe and the slot money we acquired on average will be worth more than Morris’s replacement level production.

          Watson did struggle in his first partial season in the majors in 2011(4.66 FIP), although he had good strikeout numbers (8.1 K/9), posting a -.5 WAR. He immediately reversed things in 2012 and has put up good numbers ever since. I guess he would be the best case scenario for Morris, although the fact that Watson has put up just 1.4 career WAR is not super impressive.

          Now Gomez isn’t a whole lot better and it would have sense to trade him if the Pirates could have got similar value as Bryan Morris, but I don’t think that was the case. The Pirates like Gomez because he can spot start/long relief as well. Pimentel is really similar to Morris, but much younger. And in hindsight, I think its much easier to say the Pirates should have traded Grilli. But it’s pretty likely that Grilli will be better than Morris over the rest of the season.

          • R Edwards

            I have heard all of the same reasons for why they traded Morris – to add flexibility, etc. I remember the team making similar statements after trading Aramis Ramirez. I am not equating the two as players or the trade circumstances, just pointing out the “company line” isn’t always right or the truth.

            For my Pirates sake, I hope Grilli bounces back – but I suspect he is done. His age is against him and lets face it – he’s not the same pitcher since last year’s injury.

            Baseball is played by humans, not computers, so stats do not always tell the whole story. Only a fool cannot see Morris has everything needed to be a great reliever – possibly a closer. For a reliever, he was well before his prime – most top relievers and closers are well into their 30s. The Pirates dumped him, hurt this team in the process, for a draft pick that will not contribute to the parent club for at least 2-4 years – if at all. Pimentel could have filled Gomez’ spot starter role, and Gomez let go – even if they had to DFA him. Its not like Connor Joe is going to help thme this year.

            I’ll check back in a month and see how this looks then.

            • Max Fogle

              I definitely understand why Morris looked liked he could be a good reliever, but as you said we’ll have to wait and see. I don’t have any sort of agenda in defending the Pirates decisions, and I can see why Conner Joe isn’t exciting. I’m not very good at predicting the human side of baseball, so I’m stuck looking at stats.

              I just look at what the data says first, and then look for possible reasons why a certain player might be better or worse. In Morris’s case, I don’t see enough to say “He’s going to defy all of the stats and become a stud.” If from a scouting perspective you think his stuff is really good, then you might adjust the numbers a little. But I think the burden of proof is on the person saying he’s going to be an exception to the rule. For me, he has 1 above average pitch (the cutter thing) and the potential for an above average fastball if he can locate it better. And 27-years-old is really past peak (7 out of the 10 top relievers by WAR last year were under 27, and all the aging curves say the same thing).

              And I partially agree on Grilli. I think he’ll pitch better than this going forward just because it’s hard to make a case that he won’t. But I agree that it shouldn’t be as the Pirates closer. I can easily see your point of view, I just feel you are weighting recent results a little too heavily.

            • Y2JGQ2

              Grilli has looked good and bad this year. Good when his control is good, bad when his control hasn’t been there or his velocity has been down. Realistically, he only needs one or the other to be effective, and likely he will be fine going forward. Lets all stop with the drama (including you tim) and let this one go. If he doesn’t see improvement within a few weeks lets make a change, until then, let him and searage and benedict try to figure it out

        • bucsws2014

          In 2011, Watson’s first season, he gave up runs in 11 of 45 appearances (24.4%). He allowed inherited runners to score in 8 appearances and had a 79.2% strand rate, a 12.8% HR/FB and a 4.20 xFIP.

          In 2014, Morris’s third season, he gave up runs in 9 of 21 appearances (42.9%) and allowed inherited runners to score in 6 appearances. His HR/FB rate was over 30% while with the Pirates and his xFIP was over 4 at the time of the deal.

          Watson got better. Morris got worse. Case closed.

  • https://www.facebook.com/clementewall21 N_Cap

    Max fogle played for fc Harrison city solid defenseman

  • Max Fogle

    Basically, the reasoning for regressing, is that a pitcher’s BABIP in Year 1 is not very predictive of their BABIP in Year 2. There are obviously some exceptions, so for specific pitchers you consider their career BABIP as well. HR/FB is the same sort of deal.

    In reality, we regress all statistics towards the mean when we make regressions. The two you mentioned just have a of random variation, and don’t really say much about a pitcher’s true talent. They are okay for describing what happened, just not “what will happen”. A cool thought experiment is to play around which statistics are most predictive of each other with neat tool on fangraphs:
    http://www.fangraphs.com/blogs/tool-basically-every-pitching-stat-correlation/