Subtracting velocity has been key to Luis Severino’s improved changeup


Regardless of what happens the rest of the year, this has been an overwhelmingly positive season for the Yankees and their youth movement. Both Aaron Judge and Gary Sanchez emerged as bonafide middle of the order thumpers in their first full MLB seasons, and Luis Severino bounced back from a disappointing 2016 season to become a legitimate Cy Young candidate. He’s going to finish in the top five of the voting. Maybe top three.

What’s the difference between 2016 Severino and 2017 Severino? A lot of things, really. First and foremost, I think his command is so much better. Command is not something we can measure — walk rate is a control stat, not a command stat — it’s something we have to see. Based on my observations, Severino is doing a much better job locating on the edges of the plate, and just off the plate. He made a lot of middle-middle mistakes last year.

Second, confidence. Last season it appeared Severino got scared out of the strike zone at times, and a little fastball shy. And who could blame him? He was young and getting hit hard. This year he looks like a dominator on the mound. He’s oozing confidence and you can tell he knows he’s in command of the game. There is no doubt in my mind last year’s success out of the bullpen contributes to that. He took the bullpen mentality to the rotation.

And third, his changeup. Severino always had a quality changeup coming up through the minors. We saw it in 2015. Last year he didn’t have much success with the pitch, and as a result he basically stopped throwing it late in the year. Even when he moved back into the rotation. He’d throw one or two per start, maybe. Back in 2015 he threw 10-20 changeups per start. Last season he made eleven starts and threw double-digit changeups only five times.

This season Severino is again sitting the 10-20 changeup range, and he’s throwing more and more changeups as the season progresses. Heck, he threw 29 changeups in a game against the Red Sox back in July. That never would’ve happened last year. The thing that has caught my eye about Severino’s changeup, especially the last few weeks, is the velocity. Here is his changeup velocity by start with error bars showing the minimum and maximum velocities, via Brooks Baseball:


In 2015 and 2016, Severino’s changeup sat right around 90 mph and didn’t deviate too much from that velocity. His slowest changeup from 2015-16: 85.5 mph. His fastest changeup from 2015-16: 94 mph, and I’m pretty sure that’s a mistake. PitchFX must has classified a fastball with a little extra sink on it as a changeup. His second fastest changeup from 2015-16 was 91.7 mph, so yeah, that’s probably it.

For all intents and purposes, Severino’s changeup velocity range was 85 mph to 92 mph from 2015-16. This year, as you can see from the error bars in the graph, he’s thrown his changeup down around 80 mph in several starts, with a few in the 70s as well. Look at his last few starts in particular. Severino has been throwing his changeup anywhere from 80 mph to 90 mph. Heck, Severino threw a 79.0 mph changeup and a 90.1 mph changeup in the same at-bat Saturday. From Baseball Savant:

luis-severino-nomar-mazaraLast year I thought part of Severino’s problem was the lack of separation between his pitches. I’m pretty sure I wrote that somewhere, but I can’t seem to find it now. Whatever. (Update: Here it is!) Anyway, everything Severino threw was hard. It was all 90 mph and above, for the most part. Throwing hard is great, but big league hitters can time velocity if you give them enough time adjust, and when everything (fastball, slider, changeup) comes in around the same velocity, it’s that much easier to time.

This year Severino is achieving much greater velocity separation between his fastball and changeup. His average fastball is 97.7 mph. That’s ridiculous. It’s also the highest average fastball velocity among pitchers with enough innings to qualify for the ERA title by nearly one full mile an hour. (Chad Kuhl is second at 96.9 mph. No, really.) Catching up to 97.7 mph is hard enough. Now imagine doing it when the pitcher could bust out a low-80s (or even upper-70s) changeup. That’s how aces are made. We’re talking about a 15-20 mph separation between the fastball and changeup. That’s enormous. The league average is 9.1 mph.

Severino’s success this season — there’s basically nothing that could happen the last two and a half weeks that would take him off Cy Young ballot at this point — is the result of many things. Improved command, more confidence, experience, natural growth, and an improved changeup. That pitch was a non-factor for him much of last season. This year it is a legitimate weapon, and it’s only his third pitch. The fastball-slider combination is damn near elite. Add in a changeup that is approaching Bugs Bunny status, and there’s every reason to believe this version of Severino is here to stay.

A check in on Masahiro Tanaka’s spin rates

(Adam Hunger/Getty)
(Adam Hunger/Getty)

Overall, this has been a trying season for Masahiro Tanaka. Given all the success he had in Japan and in his first three seasons with the Yankees, this has to be the most difficult season of his career. This is the first time he’s really struggled. I don’t mean for one or two starts. For an extended period or time. The All-Star break is only two weeks away, and he’s sitting on a 5.74 ERA (5.27 FIP) in 84.2 innings. Yuck.

Last time out Tanaka was excellent, striking out nine in eight shutout innings against the Rangers. He allowed three singles and two walks. That’s all. We’ve seen some flashes of brilliance from Tanaka this year, so we know it’s still in there. We just haven’t seen it consistently. Hopefully that start against Texas was a sign of things to come. Given how the season has played out, it’s way too early to say Tanaka has turned the corner.

Anecdotally, it seems Tanaka’s problems stem from his splitter and slider, his two go-to pitches. He’s not overpowering by any means. He succeeds by keeping hitters off balance with the splitter and slider. This year, for whatever reason, those two pitches haven’t behaved properly. Sometimes they do! And when they do, Tanaka has a start like he did against the Rangers. When they don’t, it’s a Home Run Derby.

For the most part, whenever Tanaka has allowed home runs this year, they’ve come on pitches that didn’t do what they were supposed to do. That usually how it works, right? Rather than dive out of the zone, those pitches stay up and get hammered. Here are the pitch locations of the 21 (!) home runs Tanaka has allowed in 2017, via Baseball Savant:


Five of those 21 home runs have come on splitters and four have come on sliders, and, as you can see in the plot, those pitches were left up. There’s no bad luck here. We haven’t seen someone go down and golf a diving splitter into the short porch or something like that. No, when Tanaka has been taken deep, it’s been a bomb on a pitch sitting middle-middle.

Since Tanaka has had trouble getting his splitter and slider to do what they’re supposed to do for much of the season, I figured it would be a good idea to look at the spin rate of each pitch. Spin rate is similar to velocity in that it’s not everything there is to pitching. It’s one tool in the shed. Spin rate could, possibly, shed some light on why the slider and splitter aren’t doing what they’re supposed to do. So let’s get to it, shall we?

The Splitter

Let’s get a few things out of the way first. One, spin rate is expressed in revolutions per minute even though it takes less than half-a-second for a pitch to reach the plate. Two, spin rate data only goes back to 2015, so that’s as far back as I went. I’d like to compare Tanaka’s pre- and post-partially torn elbow ligament spin rates, but alas. Can’t be done. And three, I looked at the spin rates on a month-by-month basis. Start-by-start is extreme overkill. Here are Tanaka’s splitter spin rates:


The spin rate on Tanaka’s splitter is down noticeably from last season, when he was one of the top pitchers in the American League. Spin rate is complicated though. More spin (and less spin) means different things for different pitches. High spin on a fastball correlates well to swings and misses while a low spin rate correlates well to ground balls, for example.

For a splitter, a low spin rate is actually better. A lower spin rate equals more tumbling action, and that leads to both more grounders and swings and misses. The higher the spin on a splitter, the more it acts like a true fastball. In theory, spin rate says Tanaka’s splitter this year should be getting more grounders and whiffs than last year because it has less spin, and:

  • 2016: 33.2% whiffs per swing and 65.1% grounders per ball in play
  • 2017: 41.5% whiffs per swing and 63.8% grounders per ball in play

Well look at that. Tanaka’s swing and miss rate on his splitter is up 8.3 percentage points from last year. That’s pretty significant. Going from 33.2% whiffs to 41.5% whiffs is huge. (The MLB average on splitters is 34.4%.) The ground ball rate is down 1.3 percentage points, which is relatively tiny. For all intents and purposes, the grounder rate has held steady since last year while the swing and miss rate has gone up quite a bit.

Okay, so what the hell does that mean? I’m not sure, exactly. But! This is actually good news, right? Or maybe it would be better to say this is not bad news. I’d be worried if Tanaka’s splitter spin rate jumped a bunch this year. That would indicate far more “straight” splitters, or hangers. On a macro-level, the splitter is doing exactly what it’s supposed to do. On a micro-level, some individual splitters are not, and those are the ones getting hit a long way. Those are the mistake pitches.

The Slider


Kinda interesting the league average slider spin rate keeps climbing, isn’t it? High slider spin rate correlates very well to swing and misses, though there’s basically no correlation with ground ball rate. A high spin slider gets as many grounders as a low spin slider. Weird.

Anyway, it sure looks like teams have really started emphasizing slider spin the last few seasons, and Tanaka’s slider spin rate has climbed along with the league average. It’s gotten better and better with each passing month so far this season. Not coincidentally:

  • April: 26.8% whiffs per swing
  • May: 44.9% whiffs per swing
  • June: 45.9% whiffs per swing

As Tanaka’s slider spin rate has gone up, hitters have come up empty with more of their swings against the pitch. The MLB average swing and miss rate on sliders is 34.9% this year, and Tanaka has been well above that in April and May. It’s probably not a coincidence then that Tanaka’s four highest strikeout totals this season have come within his last six starts.

Tanaka’s slider spin rate this season is good news. It’s getting better and the pitch is missing more bats. Similar to the splitter and a high spin rate, I’d be worried if Tanaka’s spin rate on his slider was way down. Overall, it’s been great. It’s those one or two (or three or four) mistake pitches per start that have cost him dearly. Limiting those is the key going forward, which is something we probably already knew, huh? Yeah.

* * *

All things considered, the spin rates on Tanaka’s splitter and slider are right where they should be so far this season. That’s good! That’s at least an indication he’s not broken for good. I never really through that was the case though. If it were, Tanaka wouldn’t throw these random great starts every once in a while. This leads me to believe his problems are mechanical, which is what he’s said since Spring Training. Funny how that works.

With Tanaka, he can never be a normal pitcher and just struggle. Every time he has a bad game or a bad stretch of games or hell, even throws a bad pitch, it’s because of the elbow. Always the elbow. That’s lazy though. We’re better than that. If Tanaka’s elbow were acting up, he wouldn’t be able to spin the baseball the way he normally does, and right now the overall spin rates on his slider and splitter show no red flags.

Jordan Montgomery’s Adjustment

(Elsa/Getty Images North America)
(Elsa/Getty Images North America)

The Yankees season has largely been a story of adjustments. Or, perhaps, the greatest questions regarding the roster have revolved around adjustments: how would the league adjust to Gary Sanchez? Could Aaron Judge adjust to the majors? Could Luis Severino re-adjust to being a starting pitcher? How would Dellin Betances adjust to his career as an astronaut? And so on. For the most part, these questions have yielded positive answers, small sample sizes be damned (and dissipating at a rapid pace, to boot).

Heading into Tuesday night, we wondered how Jordan Montgomery would adjust to facing the Royals for the second time in six days. It was the first time that a major league lineup would see Montgomery twice, and it had an added layer of seeing how he would fare follow the worst start of his young career (5 IP, 4 H, 5 R, 3 BB, 4 K). The Royals are a bad offensive team – the worst in baseball on the season – but they have been heating up, and Montgomery is still a rookie. It may well have been the biggest test this side of his debut this season.

By now you know that Montgomery responded with a gem of a performance, pitching to the following line: 6.2 IP, 2 H, 1 R, 0 BB, 6 K. The lone blemish on that line was a solo shot by Lorenzo Cain in the 7th inning, the result of a 93 MPH that caught too much of the plate. It was nevertheless his best start to-date, and he outpitched Royals ace Danny Duffy. All of this raises a question, though – what changed in the last week?

The short answer is pitch selection and pitch location. Let’s look into Montgomery’s offerings on May 18:

(Brooks Baseball)
(Brooks Baseball)

Montgomery threw 83 pitches the first time he faced the Royals, and just over half of those (42) were some variety of fastball. He picked up just seven whiffs on the day, largely due to the fact that he threw just 11 sliders. As per PITCHf/x, his slider is worth 2.55 runs per 100 thrown and has a 22.1% swinging strike rate, which makes it his best pitch by a fairly comfortable margin. With that in mind, take a look at Tuesday night’s start:

(Brooks Baseball)
(Brooks Baseball)

This time around, 41 of his 98 pitches were fastballs, and he threw more than twice as many sliders (which led to twice as many swings and misses). Montgomery threw fifteen more pitches this time around, and essentially all of them were sliders. It was a completely different mix of pitches, and it helped to keep the Royals off-balance; and the results were excellent.

It wasn’t just a matter of throwing more sliders, though. Montgomery was also far more successful in keeping the ball around the edges, as well as in the bottom-third of the strike-zone.



In the first outing, Montgomery was, to oversimplify, throwing the ball down the middle or outside of the zone. And, given that most the pitches he threw were fastballs or change-ups, it’s no surprise that he was hit, and hit hard.


Montgomery threw a few too many pitches near the heart of the plate both times around, but he was clearly living on the edges far more often on Tuesday night. He was also pounding right-handed hitters down-and-in (and lefties down-and-away), and it worked quite well. The majority of his pitches move, and he has shown the ability to locate most of them well-enough, so the latter plot is exactly what you’d expect to see when Montgomery is on his game.

The usual “it’s only one game” caveat applies here, yet it is encouraging to see Montgomery make such a significant adjustment from one game to the next. He went with what has worked best for him this season, and held the Royals to 1 run in 6.2 IP. On most nights, that would be a winning effort – but I digress. One of the most often cited pluses on Montgomery’s scouting report was his pitchability, and that was on full display for at least one night.

Extra rest for Masahiro Tanaka shouldn’t be a priority the remainder of 2015

Can he play second between starts? (Brian Blanco/Getty)
Can he play second between starts? (Brian Blanco/Getty)

Following last night’s games, the Yankees are three games back of the Blue Jays in the AL East and 2.5 games up on the Astros for the first wildcard spot. They’re four games up on the Twins for a wildcard spot in general. The Yankees have only 18 games to play, so their postseason odds sit at a healthy 96.7% according to FanGraphs, but they haven’t clinched anything and they certainly aren’t out of the division race. Not with three more games to play against Toronto.

For much of the season the Yankees prioritized extra rest for their starters and it was understandable. Every single one of them had some kind of injury or workload concern aside from Nathan Eovaldi. Adam Warren was starting for the first time in three years and both Michael Pineda (shoulder) and CC Sabathia (knee) had lingering health concerns. So did Masahiro Tanaka, who missed just about the entire second half last year with a partially torn elbow ligament.

The Yankees were so diligent with Tanaka this year that just one of his first 15 starts came on normal rest. He had at least one extra day for the other 14. Tanaka has made 22 starts this year overall and only four have come on normal rest. No pitcher in baseball has more starts with extra rest in 2015. The combination of off-days and spot sixth starters have allowed the Yankees to give their ace extra rest way more often than not this year, which ostensibly has allowed his elbow to hold up and remain in one piece.

But now, with the season over in less than three weeks, the priority has shifted. Keeping Tanaka rested and healthy was the focus earlier in the season. Now? Now the focus has to be on winning and getting to the postseason, either as the division winner or as a wildcard team. And to do that the Yankees will need Tanaka on the mound as much as possible. He’s their only reliable starter at this point.

The stats say Tanaka is actually more effective on normal rest: he has a 2.88 ERA and 5.85 K/BB ratio in 12 big league starts on normal rest and a 3.30 ERA and 5.21 K/BB ratio in 23 starts with an extra day of rest. I don’t think the Yankees were concerned so much about Tanaka’s effectiveness on regular rest though. I think the concern was his health, first and foremost. That should change late in the season for the sake of getting to the postseason.

Tanaka started Sunday afternoon, and keeping him on a normal five-day schedule sets him up to start these days down the stretch:

Friday, Sept. 18th at Mets
Wednesday, Sept. 23rd at Blue Jays
Monday, Sept. 28th vs. Red Sox
Saturday, Oct. 3rd at Orioles

The AL wildcard game is scheduled for Tuesday, October 6th, which means Tanaka would not be lined up to pitch that day if he remains on a five-day schedule through the end of the regular season. That’s something that can be re-evaluated in the future as the Yankees get a clearer picture of their postseason situation, however.

The more immediate priority is that Blue Jays series next week. The Yankees have to start Tanaka in that series if they want to have any chance to win the division. He’s their best hope for a win. There are only two ways to start Tanaka in that series too. They could simply push his next start back a few days and have him skip the Mets series all together, or they could start him on normal rest against both the Mets and Blue Jays. That lines him up for the final game of the series in Toronto.

Given their place in the standings, pushing Tanaka’s next start back and having him start against the Blue Jays but not the Mets seems ridiculous. The Yankees need to pitch Tanaka more to get to the postseason. Not less. The “normal rest against the Mets and Blue Jays” plan makes too much sense. I’m sure the team doesn’t love the idea of Tanaka making three straight starts on normal rest — he started Sunday’s game on regular rest — but what are they supposed to do? Winning is the priority now.

That series in Toronto could easily decide the AL East. The Yankees could be buried in the standings or atop the division after those three games. Once the team gets through those three games, they can reassess their rotation plan, and decide whether they need to continue pushing Tanaka on regular rest to get to October, or take their foot off the gas and line him up for the wildcard game. The race should be clearer a week from now. For now, things are too tight to worry about getting Tanaka extra rest. Making these next two starts against the Mets and Blue Jays on regular rest seems like a no-brainer.

Taking stock of Ivan Nova’s post-surgery performance

(Maddie Meyer/Getty)
(Maddie Meyer/Getty)

It’s often said pitchers coming back from Tommy John surgery are vulnerable to erratic performances as they regain arm strength and command of their pitches. When you combine that thinking with the notion that Ivan Nova’s career has largely been defined by periods of brilliance mixed in with extended stretches of mediocrity, it was probably inevitable that Nova was not going to be a model of reliability when he rejoined the pitching staff in late June.

Nova flashed signs of being that near-dominant pitcher in his season debut when he fired 6 2/3 scoreless innings against the Phillies, but two starts later was perhaps at his worst this season when he gave up four runs and had just one strikeout versus the Rays. And most recently against the Indians last week he put together another uninspiring start, allowing three runs before being pulled after five innings of work.

To his credit, Nova acknowledged that he’s had his ups and downs this season. “You’re going to have days like this,” Nova told the Associated Press after his dud on August 20. “Not going to feel perfect every time you go out there.”

Despite battling through bouts of inconsistency and posting a 3.72 ERA that is just barely above league-average (and a below-average 4.12 FIP), there are still a bunch of encouraging signs from Nova’s first 10 starts. Digging deeper into his numbers, there is a hint of optimism that he can be a viable starter for the Yankees as they battle for a playoff spot and the division crown in the final six weeks.


The first thing you typically look for in a pitcher trying to come back from Tommy John surgery is changes in velocity and throwing mechanics. Nova passes that test with flying colors, as his velocity is on par with previous seasons and the release points on his pitches are unchanged. He’s averaging 93.3 mph on his four-seamer and sinker, nearly the same as his rookie season (93.4) in 2011 and his last healthy season (93.9) in 2013.

Brooksbaseball-Chart (3)

His signature curveball has also been really sharp, with a top-20 whiff rate (37 percent) and top-10 marks in both batting average against (.143) and slugging (.196). Maikel Franco is one of the top rookies in the NL this season, but he had no chance on this two-strike hook from Nova back on June 24:


Pitchers returning from Tommy John surgery often struggle with their control but that hasn’t been the case with Nova. His walk rate of 7.5 percent this season is identical to what he did from 2011-13 (7.7 percent), and although he’s throwing pitches out of the strike zone at a career-high rate of 58 percent, he’s also locating pitches in the heart of the zone at the lowest rate of his career (18 percent). It seems like he is still trying to get comfortable pitching on the edges of the zone, but he’s done a good job of avoiding mistakes and grooved pitches right down the middle.

Another good omen for Nova is that he’s back to being a ground ball machine, with a ground ball rate of 52.5 percent that almost matches his 2013 mark. His hard-contact and soft-contact rates are also his best since 2011, and he’s generating popups at a rate that is nearly double his previous career best.

nova contact

Despite those positive trends, one concern is that Nova’s strikeout rate is below his peak 2012-13 levels, and he seemingly hasn’t yet regained the feel for his four-seam fastball this season. Opponents are hitting .309 and slugging .546 in at-bats ending in his heater, and have whiffed on just seven percent of their swings against it.

Perhaps realizing its ineffectiveness, Nova has ditched his four-seam fastball recently in favor of the much more effective sinker that ranks fifth among starting pitchers in ground ball rate (68 percent).

Brooksbaseball-Chart (2)

The fact Nova has been able to make these adjustments mid-season is an excellent indicator that he’s evolving as a pitcher and getting closer to reaching his potential.

Another sign of his maturity is the way that he’s been able to get out of jams and pitch under pressure this season. Batters have a .151/.230/.236 line against him with runners in scoring position and he’s stranded nearly three-quarters of his baserunners so far.

Although Nova is far from a finished product and is still clearly trying to find his pitching rhythm post-surgery, he’s shown a lot of promise in his first 10 starts this season. He’s keeping the ball on the ground with his sinker, mixing in a nasty curveball when ahead in the count, and pitching with confidence and poise from the stretch.

There’s still one hurdle, however, that Nova has yet to overcome: the inconsistency that has defined not just this season, but his entire career. Sure, he can’t shed that label in single game. But a strong performance tonight against the Astros would not only be an encouraging sign of progress in Nova’s return from Tommy John surgery, but also a key step forward in his long-term development from a talented yet unpredictable pitcher into a reliable top-of-the-rotation starter.

Digging into Andrew Miller’s post-DL issues

(Patrick Smith/Getty Images)

Andrew Miller tossed a clean, 1-2-3 inning and got the win in Monday night’s walk-off victory over the Twins — probably his best effort since coming off the disabled list — yet his performance over the past month has been anything but perfect. Sure, one solid outing is encouraging, but looking at the whole body of his post-DL work reveals a few troublesome trends for the lefty.

When Miller went on the disabled list June 11 with a forearm injury, he was in the midst of an incredible season as the Yankees’ closer, invoking comparisons to Mariano Rivera at his peak. Unhittable, dominant, video game-like … All those words described Miller, and they weren’t used as hyperboles.

He was 17-for-17 in save chances and had allowed just three runs and eight hits in 26 1/3 innings. Miller didn’t give up his first run until the 40th game of the season on May 19 and it was another two weeks before he gave up multiple hits in an outing. He struck out more than 40 percent of the batters he faced in those first two months, and made countless hitters look silly chasing a nasty slider in the dirt or swinging through a 96-mph fastball up in the zone.

After a month on the shelf, Miller returned to action July 8 against the A’s but hardly looked like the nearly untouchable reliever we saw before the injury. He gave up a single and a home run to the second and third batters he faced in the ninth inning, before finally getting Ben Zobrist to ground out with the tying run on second base to seal the 5-4 win.

It soon became clear that first shaky outing was not just a result of him being rusty after an extended stint on the DL as his struggles continued the rest of the month and into August. In 16 games since coming back, he’s allowed eight runs and 14 hits in 15 2/3 innings.

Miller overall2

When asked about his recent struggles, Miller acknowledged he was scuffling but offered little explanation for the dip in his performance.

“I am certainly having some tough innings lately,” Miller told the New York Post. “All of it is not explainable … It’s going to happen. If you expect otherwise you are naïve.”

Miller is correct, it’s nearly impossible for any player not to experience a few bumps in the road over the course of a 162-game season. But the stark contrast in his performance before and after going on the disabled list is hard to ignore, regardless of the small sample size. Let’s take a deeper dive into the numbers to see if we can figure out what might be the cause of Miller’s sharp decline.


The good news is that his pitches look pretty similar since coming off the DL compared to his first two months. His fastball is rising a tiny bit less now but the velocity is virtually unchanged, he’s still getting good depth with his slider, and his release points on both pitches are nearly the same as before the injury.

Another positive sign is that despite giving up more runs and hits since coming off the DL, batters aren’t crushing his pitches and he’s actually doing a better job of inducing weak contact recently. His hard-hit rate has decreased (from 33 percent to 27 percent) and his soft-hit rate has increased (from 17 percent to 30 percent) in the past month, while the exit velocity on those batted balls has also fallen slightly (from 86 mph to 84 mph).

Although the lower hard-contact rate is promising, it hides a more troubling trend of Miller allowing a lot more contact overall since his injury. Batters have been much more aggressive in swinging at pitches both inside and outside the strike zone, and at the same time, have also been more successful in putting those pitches into play.

Miller contact2

He’s still getting good results with his slider — .158 batting average allowed and 48.5 percent whiff rate — but his four-seam fastball has been much less effective and pretty awful by his standards.

Miller fastball2

What used to be a really uncomfortable at-bat for most hitters — trying to get wood on Miller’s unhittable fastball-slider combo — has become a much easier matchup now. He’s been able to consistently bury his slider below the knees, but he’s leaving more fastballs in the heart of the zone during the past month (see the big red blob in the middle, that’s not good).

Miller pre

Perhaps lacking confidence in the pitch, he’s relying on the heater less and less in recent outings, and that trend was never more evident than against the Blue Jays last weekend.

He threw just eight fastballs — two for strikes — and the one hit he allowed came off a four-seamer located dead-center in the middle of the plate. Of the 20 sliders he threw, he got five whiffs and both his strikeouts, without yielding a single ball in play off the pitch.


That epic 12-pitch strikeout of Troy Tulowitzki to end the game last Friday night showed that Miller still possesses one of the nastiest sliders in the game, and hasn’t lost any of his competitiveness on the mound or his ability to execute in clutch situations.

One month of appearances — fewer than 16 innings pitched — is admittedly a small sample size to make any definitive judgments on his future performance, and there is every reason to believe that Miller can turn it around and pitch better down the stretch (see Monday’s solid outing). But if he can’t locate his fastball and is unable to lower his bloated contact rate, he’s going to find himself in trouble more often than not. And we’re probably going to see more white-knuckle saves (along with some meltdowns) and fewer of those dominant pre-injury outings over the final two months of the season.

Does Tanaka have a home run problem?


Masahiro Tanaka’s numbers from his last start against the Orioles on July 23 looked solid enough to declare the outing a success — he pitched into the eighth inning, got seven strikeouts, didn’t walk a batter and allowed just three runs on five hits. Yet that sentence leaves out one important number: Tanaka also gave up three home runs.

If this was a mere blip in his season gamelogs, there might not be much to discuss here, and we could dismiss the trio of homers as a random event that happens to even the best pitchers in the sport. But those home runs were the 13th, 14th and 15th that Tanaka has allowed this season, matching his total from all of 2014 in 55 fewer innings pitched.

Not only is Tanaka more homer-prone this season compared to last year, but he’s also giving up longballs at one of the worst rates in the league. His 1.65 homers allowed per nine innings is the fifth-highest mark among pitchers with at least 80 innings pitched. Tanaka is one of 29 pitchers that have allowed at least 15 homers this season; his 325 batters faced are the fewest among that group.

Tanaka was frustrated after giving up those three homers against Baltimore, especially the two in the eighth inning that sent him to the showers.

“I felt pretty good all the way through the day, and you want to kind of go out of the game strong, but I gave up those two home runs,” he told reporters after the game through a translator. “So, not particularly happy about that, but we’ll adjust and go about it next time out.”

Joe Girardi acknowledged the issue, but wasn’t too worried because Tanaka was able to limit the damage.

“It’s not what you want to see,” Girardi said in his postgame press conference. “But you can survive giving up solo home runs. You can survive and be very successful.”

So do we make of the situation, do we need to press the panic button and start worrying about Tanaka’s sky-high home run rate? On to the analysis!


The big number on Tanaka’s page that sticks out like a sore thumb is his 2015 home run to fly ball rate of 17.9 percent. That’s the third-highest rate in the majors (min. 80 IP) and well above the league average rate of roughly 10 percent.

Typically, pitchers with inflated homer-to-fly ball rates should see some regression to the mean over the course of the season, and so we’d expect fewer of his fly balls to go over the fence in August and September. Of course, that’s not always true — if you’re continually throwing meatballs and giving up lots of long hits and hard contact, you’re still probably going to allow a lot of homers, regardless of what the statheads say. But that doesn’t seem to be the case with Tanaka, who legitimately might be suffering from some bad luck so far this year.

The right-hander has one of the highest soft-contact rates in the league (21 percent), and his hard-contact rate (29 percent) is right around the league average and even lower than last year (35 percent). Tanaka’s average fly ball distance is up a few feet from last year (287 vs. 284), but it still doesn’t crack the top-50 among pitchers this season. And Tanaka is not throwing more pitches right down the middle of the plate, either. His percentage of pitches located in the “heart” of the strike zone is the virtually the same this year (20.6%) as last year (20.0%). Comparing his heat maps from last year and this year, he still seems to pounding the bottom of the zone and keeping his pitches out of the hitters’ sweetspot.

tanaka 2014

Another factor to consider is that he’s had to deal with one of the toughest schedules in the majors this season — the combined slugging percentage of the batters he’s faced (.412) ranks seventh-highest among pitchers (min. 80 IP). Looking at the list of the 13 players that have taken Tanaka deep, you can see he’s been victimized by some pretty good power hitters. Eight of them have hit at least 15 home runs this season and two rank in the top-6 in the majors (Bryce Harper, J.D. Martinez). The only lightweights on the list are probably Jose Altuve and Derek Dietrich.

Another encouraging number is that he’s not giving up a ton of homers when ahead in the count and in those situations when he should be putting away batters. Just four of his 15 home runs allowed have come in pitcher-friendly counts, and he’s given up just one two-strike homer all season.

However, he is finding trouble when he falls behind early and into too many fastball counts. He’s allowed eight homers after throwing a first-pitch ball, but just five after getting to 0-1 in the count. Nine of his 15 homers have come off either his sinker or four-seamer, a pitch that has been a hot topic this season and one that he continues to struggle with even after making some adjustments recently.


The bad news is that the box score stats say that Tanaka has been homer-prone this season, and you can’t erase the 15 longballs he’s already allowed.

The good news is that a deeper analysis into his core numbers shows that there probably isn’t much to worry about right now, and there’s every indication that his luck should even out during the final two months. So what do you think, is the glass half-empty or half-full?