Edwin Diaz has changed

I’ve been thinking a lot about Edwin Diaz lately. I’ve been thinking a lot about Edwin Diaz because he has been incredible. As it’s just the beginning of May, I’ve tried my hardest to not overreact to any particular players’ performances, because the sample sizes are incredibly small. Because they are so very small, the statistics have not had time to become more stable. This will take time, but what we can do is speculate, and use several metrics to paint the fullest picture we possibly can at this point.

Several days ago, Diaz took the MLB lead in saves, and became the fifth fastest pitcher in MLB history to reach 200 strikeouts. In terms of relievers, Diaz is currently tied for 3rd in the MLB in WAR, and really, the only relievers who have truly been more lights out are Josh Hader and Adam Ottavino.

There have been a couple clear changes thus far with Diaz. The first is that Diaz is throwing his slider harder than ever.

Diaz mph

 

While his fastball has stayed around the same relative velocity, his slider has received somewhere around a 1.00 to 1.50 mph uptick in velocity. This has rendered hitters absolutely helpless, as his fastball and slider play well off of each other.

The second change is that he’s using his slider more, and thus his fastball less.

Diaz usage

 

This change in usage has optimized the values that Diaz’s fastball and slider provide him. Although it is widely considered a plus-plus pitch, Diaz’s heater has never before registered as such. It seems the effectiveness of his slider and his shift in usage may have helped his fastball play up.

When Diaz is locating his fastball on the edges of the strike zone, he’s near impossible to hit, and this year his fastball has been spectacular.

According to Pitch Info, his wFA/C (the number of runs above average batters were against his fastball) have been -0.29 and 0.45 in 2016 and 2017, respectively. This year, it’s 2.78. As for his wSL/C (slider value), it registered a 3.72 and 1.61 in 2016 and 2017, respectively. This year, it checks in at 2.34.

So far, his slider has been an absolute worm killer. The only batted balls in play that hitters have been able to manage against Diaz’s sliders have been grounders. Think about that: Out of Diaz’s 95 sliders this year, zero of those balls put in play have been line drives, fly balls, or pop ups. Further, 68.89% of sliders that hitters have offered at have been whiffs. That’s insane.

Along with more velocity, Diaz has kept his slider out of the zone more than ever.

Diaz out of zone

While pitching out of the zone didn’t make Diaz’s 2017 better than his 2016, hitters have been much worse at hitting balls pitched out of the zone this year in comparison to the years before.

Overall, these changes had led Diaz to see improvements almost everywhere. Hitters are:

  • Swinging at more pitches in the zone
  • Making far less contact on pitches outside the zone
  • Making less contact on pitches inside the zone
  • Swinging through a higher % of pitches

And yet, maybe the most interesting tidbit is that Diaz has increased the spin rate on his slider and fastball considerably.

Diaz spin rate

Since last year, the spin rate on Diaz’s slider and fastball have increased by about 1,000 revolutions per minute (rpm). This now puts him with the likes of Craig Kimbrel, Blake Treinen, Brandon Morrow, and Ken Giles.

Most likely due to his change in spin rate, the vertical movement of his pitches, particularly his slider, have both increased as well.

Diaz movement

Diaz’s newfound spin rate appears to be behind his added movement, and thus his early season successes.

And so, it looks as if Edwin Diaz has made adjustments. When pitchers make adjustments, hitters make adjustments back. Baseball, really, is a game of back-and-forth adjustments. Maybe Diaz loses his feel for his pitches. Maybe Diaz starts throwing his fastball too much. If hitters somehow manage to stop swinging at his pitches, maybe Diaz starts walking more hitters. Really, he already is walking a lot of hitters. Too many hitters.

It’s early, and so these early season results could be a whole lot of nothing. This is just a blip in the season, and maybe Sugar is simply throwing a few more sliders than usual. As recently as April 26th against the Indians, Diaz threw 73% fastballs. Maybe he didn’t have a feel for his slider, but it is certainly easy to envision a 24-year-old fireballer overindulging in a pitch of his that can surpass 100 mph.

It’s been just 14 games and 14.1 innings. Diaz’s BABIP is .105, and he’s stranded 91.7% of hitters on base. Clearly, this isn’t completely sustainable, and if you’ve watched some of his outings, it’s easy to imagine them ending up much worse than they did. One thing to keep an eye on is if he continues to walk a lot of hitters, which Diaz is prone to doing at times. That said, if he continues utilizing this mix of his fastball and slider, it would not be surprising to see him finish the year as a top-5 closer.

 

Is Marco Gonzales for real?

For quite some time, the Mariners’ starting rotation has been bad. So very bad that Yovani Gallardo pitched 22 games for the Mariners last season. Just years ago, Felix Hernandez was regarded as the best starting pitcher in the American League, and Hisashi Iwakuma was considered by some to be one of the most underrated #2 starting pitchers in the MLB. The former threw a perfect game, while the latter tossed a no-hitter. Nowadays, Felix is more bad than good and Iwakuma’s shoulder might as well have fallen off.

Last year, Jerry Dipoto took a lot of heat for trading one of the Mariners’ top prospects in Tyler O’Neill for a soft-tossing lefty coming off of Tommy John surgery in Marco Gonzales. Interestingly, both were taken in the 2013 MLB draft. O’Neill was taken with the 85th overall pick (3rd round), while Gonzales was taken 19th overall (1st round). Ever the optimist, I thought there was some logic in trading a dice roll in Tyler O’Neill for something (theoretically) steadier in Marco Gonzales.

More recently, Gonzales has been making the trade look a lot more justifiable than it had been just a month ago. After pitching 4.2 innings with no earned runs against the Astros, Gonzales blanked the White Sox yesterday through six innings while striking out eight.

Gonzales was truly bad last year, or at best uninspiring. There’s a lot to unpack here, because it’s not entirely clear why Gonzales has been so much better through his last couple outings. According to the man himself, he feels that he was only 60-70% last year. Now he feels like he’s 90-95%. Turns out, it helps to be healthy.

In the aforementioned article, Gonzales talks about how his curveball was his most affected pitch due to complications from his injury.

Marco SLG against

Slugging percentage vs. Gonzales’ pitches in 2017 and 2018

The numbers are in agreement with Marco’s reasoning. In 2017, hitters slugged .588 off of his curve. This year, they’re slugging just .364 off of it. FanGraphs concur with this sentiment, as his wCU/C (the value of his curveball) has improved from -2.76 to -0.31. It remains to be a good pitch, but it’s a big deal that it is no longer bad.

Gonzales has completely changed the way he’s utilizing his pitches. While he leaned heavily on his four-seam fastball last year, his changeup is now (just barely) his most used pitch and he’s throwing them all at more similar rates than he had previously.

Marco pitch mix

Gonzales is mixing all of his pitches in

But! Looking at it from year-to-year oversimplifies things. Especially because there hasn’t been a clear pattern that Gonzales has used his pitches this year.

Marco pitch mix by game

Gonzales’ game-to-game pitch mix

As you can see, 2018 has been incredibly messy. In terms of pitch mix, his first three games were somewhat similar, and those games ended with a total of 11 earned runs in just 12 innings. Against the Astros, Gonzales leaned much more heavily on his cutter and curveball, while ditching his changeup. In his last game against the White Sox, he almost completely abandoned his four-seam fastball and used the combination of his sinker and changeup almost exactly two-thirds of the time.

Gonzales is experimenting with what works for him, which makes sense, because his sinker and cutter are both new pitches. Technically, the sinker is something he has used sparsely before, but he only used it in one game in 2017. Through this experimentation, Gonzales will, theoretically, be able to decipher which pitches are his best and how he can properly sequence them.

Coming out of college, Gonzales’ changeup was arguably the best changeup in college ball, and became one of the best in the minor leagues. There’s no doubt that it will be heavily featured in his repertoire. His sinker and cutter have been his first and second best pitches in terms of slugging percentage (.278 and .294, respectively). This is notable given they’ve both been introduced this year.

In 2017, Gonzales threw his four-seamer 47.93% of the time, and that usage is down to 21.84% this year. Although it appears he is starting to go away from it, it’s peculiar that Gonzales continues to use his four-seam fastball at all. The interesting thing is that it is, without question, his worst pitch. Hitters have slugged .578 and .650 off of it in 2017 and 2018, respectively, and his wFA/C (the value of his fastball) is listed at -3.28. That’s incredibly bad, and to make matters worse, the average velocity on his four-seamer has dropped from 92.11 to 90.86. (Although to be fair, his velocity is down in general.)

But Gonzales hasn’t only changed his pitches. He’s also discussed moving to a more natural arm slot. This has worked in the past for many others, including Mariners ace James Paxton. Once Paxton lowered his arm slot, his fastball became faster, and he became better. Curiously, Marco has seen his all of his pitches get slightly slower.

Below you can see that, on most of his pitches, the horizontal release point has moved by almost exactly half a foot across the board.

Marco horizontal

Horizontal release point (’17 and ’18)

As for his vertical release point, it seems to have dropped slightly for his pitches. The error bars are still pretty wide at this point in the season, so it isn’t apparent exactly how much he’s adjusted his release point vertically.

Marco vertical

Vertical release point (’17 and ’18)

Given all of these changes, there’s nothing to suggest Gonzales should be nearly as bad as his 4.37 ERA. The difference between his xwOBA (expected weight on-base overage; .335) and wOBA (.334) is nearly non-existent. His BABIP (batting average on balls in play; .406) and LOB% (left on base percentage; 62.9%) are also inflated. His FIP (2.60) and xFIP (2.41) suggest that he’s pitched a lot better than his ugly ERA shows. More precisely, his FIP and xFIP convey Gonzales (so far) as a ~2.50 ERA pitcher, not a 4.37 ERA pitcher.

Marco Gonzales has, in the matter of two starts, catapulted himself into relevancy and given Mariners fans some badly needed hope in terms of starting pitching. He also, for the time being, has made himself interesting. Whether his success is due to an arm slot change, new pitches, or injury recovery isn’t entirely clear. But just as with everything else in life, it’s likely a mixture of factors coming together to cause substantive changes in his performance.

Gonzales could come out and be miserable in his next start and we could go back to writing him off. This article would be all for naught! Boo. Surely, many pitchers have had elite two-game stretches before. Most pitchers, however, have not endured the amount of changes as Marco Gonzales has. For now, Gonzales is showing why he was once drafted 19th overall. It’s far too early to tell if Marco Gonzales is for real, but it seems reasonable to think that he’ll be serviceable. For the Mariners, that just might be enough.

Mitch Haniger is good at baseball

The Mariners haven’t been particularly good, but they also haven’t been particularly bad. The White Sox, though, have been particularly bad and are currently 5-15. The Mariners got absolutely steamrolled by the White Sox on Monday to the tune of 4-10, and Tuesday eeked out a 1-0 win. The Mariners, for the most part, failed to do anything noteworthy.

Except, there was one thing. Mitch Haniger hit a single.  There wasn’t anything especially notable about it, other than it scored the game’s lone run. In fact, Haniger has been driving in a lot of runs lately. If you’ll remember, Haniger was the proud owner of a 1.054 OPS through March and April last year, until a Grade 2 oblique strain sidelined him. What he was doing pre-injury probably wasn’t sustainable, but Hanny never seemed fully healthy until the last month of the year.

This year, it seems like Hanny may be able to keep it up. From 4/3/17 to 4/23/17, he slashed .325/.435/.597 with a .434 wOBA and 180 wRC+. From 3/29/18 to today, Hanny is slashing .321/.389/.692 with a .449 wOBA and 189 wRC+ on the season. Clearly he can’t keep up this pace, but his peripherals suggest that he’s earned everything he’s produced thus far.

Last year’s hot start was buoyed by an inflated .396 BABIP, suggesting a little bit of good luck. This year, he owns a .327 BABIP, which is just a touch above his career .319 BABIP. Haniger also has an xBABIP (expected BABIP) of .290, so there may be a few more balls that have fallen in for hits than there should have been, however, xBABIP isn’t perfect.

As I’ve mentioned before, xwOBA uses Statcast’s exit velocity and launch angle data to measure what a player’s wOBA should be. Mitch Haniger owns the 8th highest xwOBA in the MLB at .465. After today’s game, Haniger’s wOBA is .461, meaning he isn’t necessarily playing over his head right now. He’s just hitting the ball really hard. So hard that he’s in a virtual tie with Mookie Betts and Eric Thames for 13th highest average exit velocity in the MLB at 92.8 mph. Holy cow.

There are a couple of things going in. First and foremost, Hanny is healthy. Oblique injuries are known to be a huge pain in the neck, and it’s clear that his swing was impacted by his injuries last year.

Second, Haniger is lifting the ball more than ever. In 2017, Hanny’s average launch angle was 10.6 degrees. That’s… not very high. This year, Hanny’s average launch angle is 19.4 degrees, good for 20th in the MLB. For reference, the average launch angle in the MLB in 2017 was 11.1. That’s a substantial difference, and significantly higher than notorious fly ball hitters like Trevor Story, Justin Smoak, and Eric Thames. In other words, Haniger might be buying into the so-called fly ball revolution. However, with just 90 plate appearances, it could also be too small of a sample to draw meaningful conclusions from.

Haniger FB BIP

As you can see, I am not crazy

Again, it’s not very far into the year, but the amount of fly balls per ball in play have increased significantly this year with both hard and breaking pitches.

There is one place in which Haniger is sure to regress. Haniger’s 29.6% HR/FB is about 12% higher than his career average, and it shouldn’t be much higher than 20%. This means that Hanny’s fly balls are inevitably going to start being caught by outfielders instead of going over the fence, but it would be ludicrous to expect Haniger to continue to hit at a clip 89% above league average.

Another form of forthcoming regression is in strength of opposition. To date, Haniger’s oppRPA+ (opponent’s run per plate appearance) is 92, or 8% below league average. This means that Haniger has faced competition that is a slightly below average. Going forward, Haniger may face stronger pitchers than he has up until this point.

As of right now, Mitch Haniger is 8th in the MLB with a 1.3 WAR. Just as he’s done before, Hanny is knocking the cover off the ball. What Haniger has not done before is hit the ball in the air at this rate. It remains to be seen how pitchers will adjust to him. What we know for sure is that Hanny has been one of the best players in the league this year. He’s hitting the ball at a higher angle than Aaron Judge, as hard as Eric Thames, and he’s been more productive than Bryce Harper. Sure I’m cherry picking hot, sexy household names, but maybe Mitch Haniger is on his way to becoming one.

Kyle Seager hasn’t changed, but maybe that’s the problem

Last year was a different year for Kyle Seager. Although it was ultimately a solid year for Seager, it was also his lowest WAR output (3.6 fWAR) since his rookie season. Despite this, Seager played a lot more like someone referred to as Corey’s Brother than Mr. Consistent.

His 2017 triple slash (.249/.323/.450) does not deviate very far from his yearly numbers in the past. However, his advanced stats better illustrate his regression in 2017. His wOBA (.326) and 106 wRC+ were both career lows (excluding his rookie season).

xwOBA (expected weighted on-base average) is a fun statistic to play with, because it uses Statcast’s exit velocity and launch angle metrics to assign batted balls a hit probability. These probabilities are then calculated and used to convert into xwOBA, which is directly comparable to wOBA, and so players’ skills can be compared with defense quite literally taken out of the equation.

According to xwOBA, Seager’s 2017 wOBA was expected to be .352. This is a far cry from his actual wOBA of .326. By subtracting Seager’s wOBA from his xwOBA, this shows that Seager’s wOBA was 0.025 less than expected compared to similarly batted balls.

This can be interpreted in a couple different ways. One of the most common viewpoints may be to say that Kyle Seager was one the wrong side of luck. A lot.

Rather, I think there has been a change in the way that teams have treated Kyle Seager. For years, teams have utilized variations of the ‘Ted Williams’ shift against Seager. This includes stacking the right side of the infield in a manner in which the third basemen moves to closer to second base, the shortstop plays to the right field side of second base, and the second basemen plays in shallow right field. This is commonly used against extreme pull lefty hitters, and Seager is no exception.

As MLB.com notes, Seager is one of the most shifted against players in the MLB. Anthony Rizzo led the league with 370 shifts faced, while Seager ranked second with 369. In other words, Seager is in a virtual tie for the most shifted against player in the league. Consequently, he is also one of the most affected by the shift.

This trend doesn’t appear to be slowing anytime soon. As teams have utilized the shift against Seager more and more, he has gotten worse at hitting against it by year.

Seager vs. shift

With statistics such as these (i.e., defensive shifts), they are often imperfect, and there are likely some errors in categorization of shifts and outcomes. Regardless, what this table tells us is Seager has been shifted against more every year, and it seems to be effective against him. This year, the differential appears to getting even more extreme, as Seager has 2 PAs with no shift and 14 PAs with a shift. The sample is ridiculously small, though, so it’s early to say with confidence.

The league has been adjusting to Kyle Seager. Kyle Seager does not seem to be adjusting back. Every so often, we will see Seager bunt to try and beat the shift, but this has not sufficiently dissuaded teams from using the shift any less. Seager may need to effectively bunt against it more often, but teams could simply be willing to give up bunt singles in exchange for more effective protection against the doubles that Seager is so inclined to hit into right field.

Defensive positioning isn’t the only way in which Seager has been dealt with differently. Pitchers made slight tweaks to how they attacked him in 2017 with great success.

Over the years, teams have learned to stay farther and farther away from Kyle Seager’s hands because he loves to pull the ball. Pitchers have always known to pitch away from him, but in 2017 the lower right quadrant was targeted more than ever, and he started getting pitched off of the edge of the plate like never before. Seager responded with a career high in swinging strike percentage and a career low in contact percentage.

While the location has changed, the types of pitches have changed as well. In 2016, Seager’s pitch breakdown was 59.25% hard pitches, 27.20% breaking pitches, and 13.55% offspeed pitches. In 2017, it changed to 55.14% hard pitches, 29.55% breaking pitches, and 15.20% offspeed pitches. These changes may seem minute, but when considering the large amount of pitches hitters see per year, it is more consequential than it appears.

Given this, it is unclear what adjustments Seager can make.  It doesn’t appear that Seager will benefit from the fly ball revolution, as his 2017 fly ball percentage (51.6%) was significantly higher than his career average (43.7%). At 20.2 degrees, his launch angle is already the 11th highest in the MLB. You could say he needs to learn to go the other way, but at 30 years old, that seems unlikely. If you are a believer in lineup protection, then maybe Seager would benefit from hitting ahead of Cruz or Canó.

Taken together, there seems to be a lot going on. Teams have positioned their defenses more extremely, pitchers have made adjustments, and Seager’s offensive production has suffered. There’s are reasons why his BABIP has plummeted to .262 (i.e., defensive shifting), although I reckon that a small portion of that is due to bad luck.

Kyle Seager hits the ball plenty hard. It may be time to venture into changing how he’s hitting the ball both horizontally (i.e., towards opposite field) and vertically (i.e., at a lower angle). Even if Seager can’t recreate his 2014 or 2016 offensive production, the criticisms Seager has received in the infancy of the 2018 season are irresponsible. With a few adjustments, we could see more consistency out of the player deemed Mr. Consistency.

Mariners acquire Dee Gordon, ammo for Ohtani

Yesterday I wrote this article and mistakenly posted it incorrectly. The Shohei Ohtani portion of the article is now very, very sad. Enjoy!

In the midst of a seven-team pursuit for Shohei Ohtani, Jerry Dipoto did something only Jerry Dipoto would do in acquiring Dee Gordon and $1M in international bonus space. In the process, the Mariners also freed Robert Dugger, Christopher Torres, and their #2 prospect Nick Neidert from their perpetually middling franchise.

To make it that much easier:

Mariners receive:

  • Two-time All-Star Dee Gordon
  • $1M international slot money to pursue Ohtani

Marlins receive:

  • Nick Neidert
  • Robert Dugger
  • Christopher Torres
  • The relief of Dee Gordon’s contract

The Mariners accomplished two things with this deal, and I’m not sure which part is more sexy. In the acquisition of Dee Gordon, the Mariners are getting a player who can play second base and shortstop.  What the Mariners hope is that he will be able to adequately play center field, which is where they intend on playing him for now.

On its own, acquiring international slot money is very much an unsexy thing to receive in a trade. But! What if I told you that the Mariners now have a slight lead in international slot money. (Or… really more of a tie with Texas.) This means the M’s can offer Shohei Ohtani more money than any other team. If you ask me, that’s as sexy as it gets. If you know some of his backstory, money clearly isn’t a pressing matter to Ohtani, but I’ve never heard of a player who doesn’t care about money.

The framework of this trade is almost identical to that of the Mike Leake trade. In the Leake deal, the M’s took on most of Leake’s contract and gave up SS Rayder Ascanio in exchange for $0.75 million international bonus space. The formula seems to be taking on contracts slightly underwater, giving up middling prospects, and getting a major league ready player and Ohtani ammo (international slot money) in return. Hell, Gordon is 29 years old; the same age as Mike Leake at the time he was traded.

Before the days of Leonys Martin and Jarrod Dyson, the Mariners were thinking of ways they could get creative to solve their black hole in center field. Back then, some suggested that Brad Miller could move off of shortstop to try and hack it. He ended up playing 146.0 innings in center for the Mariners in 2015, and the results were… not good.

Dee Gordon will now attempt to succeed where Brad Miller failed. Some have proposed the Gordon serve as a Ben Zobrist-esque superutility player, but Dipoto seems adamant that Gordon will only play center field. When Robinson Cano’s contract inevitably becomes an albatross, Gordon would hypothetically be a candidate to replace him at second base.

Today’s trade does not come without potential drawbacks. In Nick Neidert, the Mariners lost their top pitching prospect. John Sickels projects Neidert as somewhere around a number three starter. That’s not game-changing, but the back of the Mariners’ rotation isn’t exactly game-changing either. Unless you’re the other team. Our pitchers will serve up runs all day if you’re the other team, and that by definition changes the game.

Christopher Torres is an 18-year-old shortstop that wasn’t expected to join the M’s until 2020 or 2021. He is said to have the tools to stay at shortstop while also possibly being able to hit for contact and some power. Robert Dugger has been a starter and reliever in the minors and has worked a 3.22 ERA in 156.2 innings pitched.

The big deal here is not losing Neidert, and it certainly isn’t losing Torres or Dugger. It’s in getting Dee Gordon. In 2016, Gordon was popped with an 80-game suspension for performance enhancing drugs. Despite taking performance enhancing drugs, Gordon’s highest home run total for a season is 4.

It’s obvious that Gordon’s value comes predominantly from his speed on the base paths. In his last four seasons, he has averaged or been on pace for at least 60 stolen bases over 162 games. In order to swipe bases, though, Gordon will have to get on base. Going into 2018, Steamer projects that Gordon will hit .286/.325/.368 with a .301 wOBA and 85 wRC+ (i.e., 15% below league average). For comparison, Jarrod Dyson is currently projected to slash .261/.328/.369 with a .304 wOBA and 85 wRC+.  Gordon, you could argue, isn’t far from Dyson as a hitter.

Except there are clear differences between the two. The most important of which is their splits by handedness. To compare them, we’ll use wRC+. (For reference, 100 is league, 110 is 10% above league average, and 90 is 10% below league average.) Versus lefties over their careers, Dee Gordon hits a respectable 97 wRC+ while Dyson hits a comparable 93 wRC+. Versus righties, Dee Gordon hits a subpar 82 wRC+, but Dyson hits an abysmal 55 wRC+. That’s even worse than Seth Smith’s wRC+ versus lefties, and this difference will allow Gordon to play in more games and face more pitchers than Dyson would have been able to.

If the Mariners win the Shohei Ohtani sweepstakes, this trade looks incredible. If they lose, well, it doesn’t look nearly as good. If you think about it, though, it is curious that the Mariners have continued to add international money recently. Ohtani is passing up millions and millions of dollars to come to the United States now. So theoretically, he can’t have that much desire for a couple million extra dollars, at least one would think. If we’re making the assumption that Ohtani is to sign here, then this signing becomes more important. To accommodate Ohtani as a DH, Nelson Cruz is going to have to play in the field where he’s below average at best. With an athletic, defensive first outfield, Gordon should fit right in.

Mariners take a Leake

Associated Press (AP)

The Mariners currently rank 28th in the MLB in starting pitcher WAR, and so they have attempted to bolster their rotation by acquiring Mike Leake, cash, and international cash in return for SS Rayder Ascanio.

Leake is owed a $16M APY for the next three years. With the Cardinals kicking in cash, Mike Leake is essentially on a three-year, $36M deal with a $5M buyout in 2021. Leake waived his no-trade clause in order to come to Seattle.

With a 2017 WAR of +1.9 to date, Leake joins the Mariners as their second most productive starting pitcher, and it isn’t particularly close. Next in line is Ariel Miranda at +0.5 WAR.

After posting a 3.12 ERA in the first half, Leake has floundered, conceding a 6.90 ERA to batters since the All-Star break. One reason for this regression seems to be that Leake’s velocity on his sinker has been declining. Both Brooks Baseball and FanGraphs have his sinker’s velocity as the lowest it has been since 2012, and nearly an entire tick down from just last year. That’s problematic considering Leake throws his sinker almost 50% of the time.

Regardless, a stellar first half and pitiful second half combine to equal a fairly normal year for Leake. He now has seven consecutive seasons of at least 150 innings, and most of them have fallen under a 4.00 ERA. It’s reasonable to expect somewhere around 185 innings and a 4.00 ERA per season from Leake for the remainder of his contract. With Seattle’s current rotation, it’s hard to complain about that.

The going rate for middling, veteran starters is currently somewhere around $10M-$15M per year. Jeremy Hellickson, he of 9.9 WAR from 2010-2017, is on the books for $17M in 2017. Currently, one WAR is priced at around $8.5-9M, and so Leake will need to be worth around +4 WAR for the next three years. In the past seven seasons, Leake has averaged +1.89 WAR per season, so he would be hard pressed to not be worth the money.

The dip in velocity is concerning, but there are some reasons to be hopeful. First, Leake is getting more swings out of the zone than he has in many years. His swing percentage on pitches outside of the zone is up a full percentage point from his career average. Second, his outside-zone (O-Swing) contact and inside-zone (Z-Swing) contact percentages are both down around 2% from his career averages. Lastly, his swinging strike percentage (SwStr%) is the highest it’s been in his career.

As for the Rayder Ascanio, he’s a 21-year-old, 155-pound shortstop who has spent most of his time in 2017 in A+ ball. As a hitter, he is extremely light-hitting, but is an absolute wizard with his glove. While not an issue for many big league shortstops, it’s concerning that Ascanio is hitting a mere .656 OPS in A+ (21% below league average). He currently has an uphill battle to be a capable major league starter. John Sickels graded him a ‘C’ in February.

In short, the Cardinals used this as an opportunity to free up salary cap space. Seattle has made it even more clear that they aren’t tapping out anytime soon. Paired with the Segura extension, picking up a $12M annual contract through 2020 shows they are trying to win now. Fortunately they were able to do so with a 29-year-old proven veteran pitcher. Look for Mike Leake to get his first start as a Mariner sometime this weekend against Oakland.

The Mariners’ bullpen might be good enough

If the Mariners are to have one weakness, it’s their starting rotation. The Mariners, in fact, have several weaknesses, but their starting pitching happens to be the most glaring one. Currently, they stand 1.5 games back in the American League Wild Card race. While the Astros have the rest of the AL West in their rearview mirror, the rest of the American League is in disarray. Just three American League teams are no longer in the playoff picture.

The Mariners know they sorely lack competent starting pitchers, which is why they went out and acquired starting pitchers Marco Gonzales and Andrew Albers. Neither Gonzales nor Albers are first-class starters, and so they acquired 1B Yonder Alonso and RP David Phelps to shore up other areas of the roster.

Alonso adds even more punch to an offense that ranks #6 in the MLB in wRC+ (Weighted Runs Created Plus). More relevant is that David Phelps was acquired to bolster the bullpen. Phelps instantly is penned in as one of the M’s best relievers. With the Mariners’ current rotation, the bullpen will be relied upon heavily to backpack them to a Wild Card spot. This is already evident as their 51.1 IP in August is the third highest in the MLB. If the Mariners were to theoretically make it into the playoffs, it is of even more importance that the bullpen has arms that can be confided in in high leverage situations.

The question, then, is if the Mariners have those relievers. Last year, Edwin Diaz was that reliever. This year, while he’s still flashed elite stuff, he’s also been outright atrocious at times.

Nick Vincent currently owns a 6.93 K/9. Not terrible, but certainly below average. Vincent has been striking out less hitters than ever before, but he has also walked fewer, given up fewer home runs, and limited hitters to a 2.01 ERA (2.67 FIP, 4.49 xFIP).

In January, Jeff Sullivan wrote an article about how Nick Vincent had the second most unhittable fastball in baseball in the PITCHf/x era, right behind none other than Aroldis Chapman. Seriously. Ahead of Craig Kimbrel, Dellin Betances, Kenley Jansen, and Andrew Miller. Everyone but Aroldis Chapman. In 2016, he actually had the most unhittable fastball.

Many of these pitchers have fastballs that are either blazing fast or have wicked movement. Amazingly, Vincent’s heater falls into neither of those categories. It averages 90 mph and tops out at 92 mph. As for movement, Brooks Baseball has the following:

His fourseam fastball generates a high number of swings & misses compared to other pitchers’ fourseamers, results in more flyballs compared to other pitchers’ fourseamers, has slightly below average velo and has slightly less natural movement than typical.

In other words, Vincent should have a very ordinary fastball. But he doesn’t! His fastball is elite because he locates it accurately and consistently, has a deceptive delivery, and lives up in the zone with his four-seam fastball. Living up in the zone generally means lots of whiffs and lots of long fly balls. In 2016, those fly balls gave Vincent a 1.64 HR/9, but in 2017 he owns a 0.37 HR/9. One thing we know is out of all types of batted balls, home runs are the noisiest. What that means is home run rates often fluctuate, and what we have seen between the two years is a lot of fluctuation. Nick Vincent was by and large, very unlucky in 2016. This year, he seems to have had a lot more luck on his side.

One of my favorite metrics is a newer one made available by Baseball Savant, xwOBA (Expected Weighted On-Base Average). What it does is take Statcast data and assign hit probabilities to all batted balls. In this way, you can more accurately gauge a player’s true performance and compare it to their actual performance. In this way, you can see how lucky (or unlucky) they have been.

Nick Vincent currently has a .250 wOBA. (Think of wOBA on the same scale as on-base percentage.) His xwOBA is .287.

Hypothesis confirmed! Nick Vincent has benefited from good luck (and similarly good fielding) Using this metric, I can even attempt to say just how lucky he’s been. By subtracting his wOBA from his xwOBA, Vincent has performed 0.037 better than expected. Only six relief pitchers have overperformed more than Vincent. Still, Vincent is still in good company. By xwOBA, Vincent is performing around the level of the likes of Alex Colome, Carl Edwards Jr., and Seung Hwan Oh.

For the sake of (your) time, I will say that David Phelps has been just slightly worse than Vincent has been (which is still good!).

The standout pitcher from the Mariners’ bullpen this year will surprise you. It shouldn’t surprise you, but it will! When you look bad in a couple outings – and I mean bad – people will take a narrative and run with it, and apparently that has been done this year with Edwin Diaz.

Sugar ranks #26 in xwOBA at .258, while his wOBA stands at .282. This means, unlike Vincent, Diaz has not only not but lucky, but he’s been unlucky!

He’s walking just about everybody and he’s given up too many dingers, but since the arbitrarily chosen date of May 19th, Diaz has an FIP- of 74 (100 is league average) while his FIP- on the year is 95. Several of the advanced metrics available expect Diaz to be worse than his current 3.40 ERA is, but there is reason to believe the opposite: that Diaz hasn’t gotten a fair shake due to some bad luck.

Still, this is not the Edwin Diaz of last year, and it certainly seems like something is off. Diaz was probably not going to repeat his insane 2016, but there haven’t been many times this year where he has mirrored that sheer dominance.

One thing is clear: his command is not completely there. From what I can gather via Brooks Baseball, both his vertical and horizontal release points are not in line with where they were last year, and so he’s not locating. Most notably, he’s locating his slider lower than ever.

Diaz is living in the zone far, far less. When you have a slider like he does it can work, except this year both the swing percentage and whiff percentage have gone down on his slider. Diaz has got to get his fastball going or he is going to continue to walk around his rate of 4.56 BB/9. Why swing in the dirt when you can just look for his fastball?

Given this, in the bottom of the ninth of a must-win game: give me Diaz. Vincent and Phelps make for good options, but Diaz by far has the most firepower and was one of the best relievers in baseball last year. He still has until October to return to form.

The Mariners are better and worse than before

In 2016, the Mariners had the 10th best run differential in the MLB and missed the playoffs. Similarly, the Mariners finished 10th in Pythagorean Win-Loss record (i.e, expected win-loss record). So in theory, the Mariners performed as the 10th best team in the MLB. Expected to be 87-75, the Mariners were 86-76. A ten win improvement from 2015, the Mariners should have been happy, but the Mariners were sad. 2016 marked the 15th straight year that the Mariners were sad. The reason the Mariners have been sad, is because not making the playoffs is sad. How sad?

Trump tweet

I made this fake tweet that could be a real tweet

The Mariners owned a record neither better than the Orioles nor the Blue Jays, who both happened to be the AL Wild Card teams. And that is how you miss the playoffs! Losing to a terrible Athletics team in a massively important game is also a really good way to miss the playoffs.

This year, the Mariners have a completely new look, and so they will win in very different ways.

season projections

The 2016 Mariners blew AL West teams out of the water in run differential (a good predictor of win percentage), but still finished nine games behind the Astros. This year, their offense is projected to score less runs than the Astros and Rangers, and their pitching is projected to be slightly worse than the Astros and Angels. The Astros, clearly, are going to be extremely good because of their potent offense, likely the best in the AL, and underrated pitching.

After much roster turnover since the Jack Zduriencik days, the Mariners finally have something close to the philosophy that Jerry Dipoto has envisioned. Get on base, play good defense, pitch well, and control the zone (on offense and pitching). The biggest difference with our new Mariners is the ability to play stellar outfield defense, and Segura should be an upgrade for Ketel Marte. Mostly on offense, but also defense as well.

The offense is less potent, yet still potent. That’s the trade-off you take when increasing a team’s defense. Hopefully that defense will help stave a very volatile, uninspiring starting rotation that could be pretty bad. Before Drew Smyly was put on the DL, things didn’t feel great. Now that Drew Smyly is on the DL, things especially don’t feel great. While the depth of pitchers overall is good, the dropoff is pretty significant from Paxton and Smyly to Felix and Kuma, and even further to Yovani Gallardo. Dipoto purely wants volume of innings from Gallardo, which is understandable. Every pitcher in the rotation has dealt with injuries in the past, and no one seems like a sure thing to stay healthy. That is where the value of 180-200 innings from Gallardo comes in. He’s a Kevin Millwood-esque #5 who is surely not sexy but valuable in his own mediocre way. I personally prefer Chris Heston to both Ariel Miranda and Gallardo but that’s just me.

Theoretically the lineup has no black holes, but Zunino and Martin have the potential to be worse than mediocre. I don’t think it’s likely, but lucky for them they both are premier defenders at their respective positions if they do wet the bed.

Projection system overperformers

Projection systems are not perfect. In fact, they’re very imperfect. What’s great is they objectively take statistics, and they make estimates based on previous performances and the natural aging curve. Players generally get better as they age, and after age 30 they get worse. It’s pretty simple.

On a yearly basis, players overperform and underperform projections. In this way, projections even out. Players like Nelson Cruz hit until they’re 40, and players like Rich Hill become studs seemingly out of nowhere. On the flip side, Mike Trout and Bryce Harper-types become superstars not long after they are 20 years old.

As such, the Mariners have a couple players who are hopeful candidates for stronger performances than projection systems like ZiPS and Steamer account for.

Jean Segura

Over the offseason, Jerry Dipoto unloaded an underwhelming Tajuan Walker for a rejuvenated Jean Segura. It was a sell-low, buy-high situation, but it is a trade that I fully co-sign.

Last season, Segura was one of the best shortstops in the league. He played solid enough defense, and showcased a .319/.368/.499 triple slash (.371 wOBA, 126 wRC+) with 33 stolen bases and 20 dingers. It was a far better season than his 2014 and 2015 seasons, and an improvement from 2013. There are reasons to think that 2016 was an unsustainable year for Segura.

Case 1: Segura’s BABIP (Batting Average of Balls in Play) was .353. What that means is for every ball that Segura put into play (strikeouts, walks, sacrifices, and home runs do not count), his batting average was .353. BABIPs can fluctuate due to bad luck, or consistently hard or weakly hit balls. Generally, the average player’s BABIP will hover around .300. However, speedier players often have higher BABIPs because they can beat out batted balls in the infield. Segura is a speedy player.

Segura’s career BABIP is .314, which seems normal. That said, his future BABIP is likely to fall somewhere in between his .314 career BABIP and 2016 .353 BABIP.

Segura was solid in 2013. Segura was one of the league’s worst hitters in 2014 and 2015. In 2016, he was a very good hitter. What the projections have decided is that if Segura has been extremely bad and extremely good, he’s probably somewhere in between. This is reasonable enough logic.

Case 2: In 2014, Jean Segura’s nine-month old son passed away due to illness. 2014 was also the year that Jean Segura’s performance on the baseball field took a nosedive. If you can imagine, one of the worst things that can happen to a person is losing a child. An even worse thing is losing a child that has yet to reach the age of one.

Projections don’t have feelings. Jean Segura has feelings! ZiPS and Steamer unfortunately don’t know that Jean Segura lost his son, but we do. This is one way that we can try and beat projections at their own game.

There’s a hole in this logic. That hole is that leading up to Segura losing his son, he owned a .232/.266/.315 triple slash in 316 plate appearances. Upon returning, he actually improved his line to .271/.328/.348.

For the Mariners, the hope is that Segura plays more like 2016 than 2014 and 2015.

What Steamer says: .273/.316/.395, 12 HRs, 24 SBs
What ZiPS says: .269/.308/.393, 13 HRs, 30 SBs
What I say: .285/.345/.440, 16 HRs, 28 SBs

Mitch Haniger

Haniger has 34 major league games to pull from, and he wasn’t good. In spite of this, the sample was extremely small, he ran into bad luck (.256 BABIP), and a year ago he overhauled his swing. His triple-A performance was nothing short of dominant and some actually view Haniger as the headliner to the Segura-Walker trade.

Because of the small sample of 2016 appearances and his minor league pedigree, I’ll take the over on his 2017 projections.

What Steamer says: .249/.309/.412, 18 HRs, 7 SBs
What ZiPS says: .237/.302/.409, 18 HRs, 7 SBs

What I say: .260/.325/.440, 20 HRs, 8 SBs

Mike Zunino

Zunino was one of the worst hitters ever in 2015. In 2016 he mashed in Triple-A, and he looked improved against major league competition. He’s always going to strikeout around 25-30%, but it looks like he’s going to walk more and continue to play great defense.

His 2016 performance is probably something like his 2017.

What Steamer says: .220/.289/.412, 16 HRs, 1 SB
What ZiPS says: .216/.288/.418, 22 HRs, 0 SB
What I say: .225/.315/.430, 20 HRs, 0 SB

Leonys Martin

Our poor Leonys spent much of 2016 dealing with a bum hamstring. Not only was he battling his hamstring issue, but because of his integral role in the outfield’s defense, he wasn’t able to play as hard as his hamstrings would allow. Working with Edgar Martinez, Robinson Cano, and Nelson Cruz has helped his swing, and it’s of my opinion that his hamstring hampered his offense, defense, and baserunning.

Martin is probably the one I think beats his projections the most (save for maybe Jean Segura).

What Steamer says: .242/.298/.363, 10 HRs, 17 SBs
What ZiPS says: .246/.298/.371, 11 HRs. 22 SBs

What I say: .245/.315/.410, 19 HRs, 30 SBs

As for pitching, I don’t have as strong of opinions about those projections. Felix will likely be a tad better than 2016, but surely not someone deserving of being a #1 pitcher in a rotation. Gallardo might suck less. Zych, Cishek, Diaz, Scribner, and Vincent are going to ballers in the bullpen.

Happy Almost Mariners Baseball Day!

Optimizing the Mariners’ opening day lineup Pt. 1

In 2016, I wrote an article about optimizing the Mariners’ lineup a couple months into the season. To make it easier (so as not to force you to open another article) here’s a paraphrased version of what I said lineup optimization means:

In economics, there is something called the rational choice theory. What the rational choice theory assumes is than an individual has preferences among choices. These preferences are assumed to be made by surveying all possible options, taking in all information, and to make the most logical decision that is of greatest benefit to them.

To put out the best lineup possible, one should use all available information. In The Book: Playing the Percentages in Baseball, Tom Tango (as well as Mitchel Lichtman and Andrew Dolphin) attempt to quantify this using empirical data. In the book, there is an entire chapter devoted to sabermetric lineup construction.

Following is a graphic with the conventional, old-school method of building a lineup, as well as an optimized, sabermetric version of how to construct a lineup. The standard version is more philosophy and trait-driven while the sabermetric version revolves around statistics (duh).

BTBS optimization

Admittedly, over the course of a baseball season using an optimized lineup only adds a couple wins. That’s relatively insignificant, and one could argue it’s trivial. But if you’re the Mariners who look like favorites to finish as the second Wild Card team, it can be the difference between playoffs and, well, not playoffs. With an injury to Drew Smyly before the season has started, little things like lineup optimization become increasingly important.

This lineup will likely look very different from the opening day and regular lineup put out by Scott Servais, and that’s okay!

Because year-to-year numbers are so different due to injuries, sequencing, mechanics, and so on, we will focus on players’ career numbers with 2016 performance and 2017 projections as context. On we go!

I came up with the following orders of lineups versus left-handed and right-handed pitchers:

Versus LHP:

LHP Optimization

Versus RHP:

RHP Optimization

(“Stat type” denotes if I used career numbers or my personal projections.)

You will notice that the optimized lineup doesn’t follow the sabermetric setup to a tee. Why? As an example, managers like to have speedy players who can get on base some at the bottom of the lineup, because it’s as if having two leadoff hitters. Also to better match certain philosophies and because some players will do better or worse than their current career numbers or 2017 projections. Namely, players like Haniger, Segura, Zunino, and Martin. In Pt. 2 of this article, I will explain why I expect players to deviate from these numbers.

Because of injuries, fatigue, “hot/cold streaks”, and maintenance days, the lineup isn’t always going to look like this. And really the lineup will probably never ever look like this. No managers in the MLB are nearly as progressive as many sabermetricians, and so even though some unconventional tactics are utilized, many still use concrete roles in the bullpen and outdated methods of lineup construction. As an example, managers are unlikely to move star players such as Seager or Cano down in the lineup even if they are significantly worse against a certain handedness of pitchers. Due to the egos and psyches of players, this may be the correct way to go about things at times, but it doesn’t always make sense. That said, all things relatively equal, it’s generally safe to go with the more experienced (or emotionally demanding) player. So Seager will unlikely bat as low as sixth like he is in my “vs. LHP” optimized lineup.

For the first time since October, there will be regular season Mariners baseball tomorrow. The Mariners play the AL West favorite Astros, and Felix will look to recover from his past two seasons in which he has failed to look like King Felix. Last season he really failed to look like Felix Hernandez. I can support the blonde beard. Not so much the slicked back blonde hair.

Pt. 2 of my article explains my lineup construction briefly and the Mariners season in context!

Mariners solidify rotation with a pitcher you (probably?) haven’t heard of

The Mariners went via the trade route just like we all knew they would. Albeit a much more low profile acquisition than many thought, Chris Heston is a nice find. Heston was had for a player-to-be-named-later. This continues the Mariners’ recent trend of acquiring pitchers who are either groundball pitchers, good K/BB pitchers, or have high spin rates. (Depending on the pitch, it is very advantageous to have a high spin rate. With pitches like cutters/sinkers and changeups, a high spin rate is bad). 

Right away, John Sickels compared the likes of Chris Heston to Matt Shoemaker. After a solid spring, Sickels projected that Heston would have a Matt Shoemaker-esque coming out season. While that didn’t exactly happen, Heston did have a very solid season which was highlighted by two complete games, one being a no-hit performance. His rookie season Heston went 12-11 with a 3.95 ERA (3.98 xFIP; 4.07 SIERA). Additionally, he owned a passable 7.14 K/9, 3.24 BB/9, and a HR/9 of 0.81. (Just for fun, Heston’s 2015 HR/9 is 1.00 less than Taijuan Walker’s 2016 HR/9. Boy oh boy was Walker terrible.)

The big difference between the two is Chris Heston is an extreme groundball pitcher while Shoemaker is surely not. And as Sickels notes, Shoemaker’s beard is far superior.

While Heston owns a healthy line drive rate, he owns a GB% (56.2%) that is about 12% above league average and among the highest in the MLB. His FB% (26.3%) is nearly 10% below league average, while managing to keep his HR/FB rate better than league average – that’s important. His batted ball profile is strikingly similar to Martin Perez, Kendall Graveman, Mike Leake, Francisco Liriano, Edinson Volquez, and Chad Bettis. Oh, and not to mention Noah Syndergaard and Jake Arrieta. Unlike the latter two, Heston does not throw hard and will never reach that type of level of domination. Heston owns the highest GB/FB rate, and is tied for best in GB%. Similarly, Heston’s career GB% ranks second in the MLB in pitchers’ 2015-2016 GB%.

batted-balls

Brooks Baseball’s amazingly convenient beta feature offers this description of Chris Heston’s 2016 pitches compared to other RHP as follows:

His sinker is so slow that it is substantially gravitational and is an extreme flyball pitch compared to other pitchers’ sinkers. His fourseam fastball comes in below hitting speed and has heavy sinking action. His slider comes in below hitting speed, generates an extremely high number of swings & misses compared to other pitchers’ sliders, sweeps across the zone and is an extreme flyball pitch compared to other pitchers’ sliders. His curve comes in below hitting speed, generates an extremely high number of swings & misses compared to other pitchers’ curves, has little depth, results in somewhat more flyballs compared to other pitchers’ curves and has slight glove-side movement. His change (take this with a grain of salt because he’s only thrown 8 of them in 2016) has below average velo, generates more whiffs/swing compared to other pitchers’ changeups, results in more flyballs compared to other pitchers’ changeups, has slight cut action and has some natural sink to it.

The TL;DR/summary:

  • His fastball sits at or below 90 but has a great amount of movement/heavy sink.
  • His slider generates A LOT of swings and misses, but it is an extreme flyball pitch.
  • His curve also generates a lot of swings and misses, but it is less so of a flyball pitch.
  • He rarely threw his changeup.

Heston is a innings-eating command pitcher who embodies Dipoto’s philosophy of controlling the strike zone. Because of his approach, he relies heavily on his defense since he does not generate strikeouts and has walked more batters than expected thus far. Right now he is a solid fourth or fifth pitcher in a rotation, but he has the potential to be more than that. Kind of similar to what we felt about Nathan Karns before last season, although… don’t let that scare you.

At AT&T Park, Heston had one of the most pitcher friendly stadiums in the MLB (second to Safeco Field). Now that Safeco has moved in their fences, it may be less pitcher friendly, especially to flyball pitchers. Not a problem for Heston.

With the Mariners’ increasingly rapid shift towards prioritizing defense, Heston stands to gain more benefit than any other M’s pitcher. With Martin patrolling center and Gamel and Haniger likely to see a lot of time in the corners, the outfield defense has improved by leaps and bounds in just a couple months. Seth Smith also looks to be on his way out of town due to his penchant for literally only being able to hit right-handed pitching. Seager and Cano are coming off of very good defensive seasons, and Segura looks to solidify a shortstop position that has struggled to be occupied by a player who can hit or field. The last player was the defensive-savvy Brendan Ryan. Just as important, Seattle has one of the MLB’s premier pitch framers in Mike Zunino and theoretically a better backup in Carlos Ruiz than Chris Iannetta turned out to be.

Heston bolsters a rotation that also features Felix, Kuma, Paxton, and one of Nathan Karns, Rob Whalen, Ariel Miranda, or a player-to-be-acquired in the coming days (or weeks). While I think this is a rather savvy pickup by the Mariners’ front office, the Mariners should try to add on to the competition for the 5th starting spot, as well as continuing to build depth.

Not only is this a low-risk move, but the Mariners also saved the $10M-$12M they would have had to pay a mid-to-back-of-the-rotation free agent.