03 July 2015

Looking at Wei-Yin Chen's Clutch Performance

A while back, I scrutinized Miguel Gonzalez's situational pitching, and concluded that his tendency to consistently strand runners comes as the result of skill (and some luck, as do many great baseball accomplishments). Today, though, I'd like to look at another Oriole, who appears to have developed a similar ability in this regard.

Wei-Yin Chen debuted for Baltimore in 2012, posting a solid rookie season: Across his 192.2 innings, he maintained a 4.02 ERA and a 4.42 FIP. While he did well in many areas, he struggled to prevent runners from scoring, with a 72.8% strand rate and -0.7 LOB-Wins (placing 43rd and 65th, respectively, in baseball). Still, that didn't tarnish his future much — after all, not all pitchers can overperform.

In the two-plus years thereafter, though, Chen's made some changes. Over that span, his 78.2% strand rate (9th in the majors) has translated to 2.1 LOB-Wins, which ranks seventh among pitchers with 300 innings. Notably, that latter mark bests Gonzalez, among others of his ilk. So how did Chen go about attaining it?

Like fellow Oriole starter Chris Tillman, Chen has better limited base runners. In each of Chen's years in the show, he's permitted a lower amount of stolen base attempts:

Year Chen SBA% MLB SBA% Chen SBA+
2012 4.9% 6.5% 75
2013 3.5% 5.5% 63
2014 2.2% 5.7% 38
2015 2.7% 5.8% 47

But this doesn't begin to account for the entirety of the change, because Chen's also posted better results overall when the pressure comes on:

Year Bases Empty wOBA Runners On wOBA
2012 .316 .317
2013 .352 .301
2014 .338 .290
2015 .328 .266

Let's look into this. We'll start with 2012, in which Chen didn't dazzle, and for good reason — he performed worse than league average with runners on, in many regards. He walked more batters...

State Chen uBB% MLB uBB% uBB%+
Bases Empty 6.1% 7.1% 86
Runners On 8.2% 7.9% 104

...didn't strike out as many batters...

State Chen K% MLB K% K%+
Bases Empty 19.7% 20.7% 95
Runners On 17.6% 18.8% 94

...and sacrificed more home runs:

State Chen HR% MLB HR% HR%+
Bases Empty 3.3% 2.7% 122
Runners On 3.9% 2.7% 144

Examining Chen's situational pitch usage gives some clues as to the cause of this:

State Fourseam Sinker Slider Curve Split
Bases Empty 59.9% 4.6% 15.3% 7.4% 12.8%
Runners On 57.6% 8.7% 14.7% 3.7% 15.3%

The sinker and splitter replaced, to varying extents, the four-seamer and curveball. In the case of the latter, that decision likely served him well: Chen's curve went for a strike only 55.3% of the time, induced a whiff 8.0% of the time, and left the park 1.6% of the time — the worst rates of any pitch in his arsenal. The second-worst? Those belonged to...his sinker, which possessed a 60.6% strike rate, 8.3% swinging strike rate, and 1.6% homer rate.

Chen's splitter did well enough in terms of whiffs (12.3%), strikes (63.2%), and long balls (0.7%). On the flipside, his four-seamer went for strikes even more often, at 30.2%, while racking up a fair amount of swinging strikes (10.0%) and going for a home run 0.9% of the time. Overall, the switch didn't work out for him — the sinker's awfulness cancelled out that of the curveball, and the splitter couldn't make up for the four-seamer's absence.

Now, we'll compare Chen's 2012 to his 2013 through 2015. In this three-year span, he's flipped the script in all areas — he has doled out fewer free passes with runners on...

State Chen uBB% MLB uBB% uBB%+
Bases Empty 6.0% 7.0% 87
Runners On 4.3% 7.5% 57

...has punched out more batters...

State Chen K% MLB K% K%+
Bases Empty 17.2% 20.9% 82
Runners On 20.2% 19.1% 106

...and has done a better job of keeping the ball in the stadium:

State Chen HR% MLB HR% HR%+
Bases Empty 3.4% 2.5% 137
Runners On 2.8% 2.3% 122

So what has allowed Chen to reverse his fate in this manner? Again, we can turn to his repertoire for the answers:

State Fourseam Sinker Slider Curve Split
Bases Empty 55.8% 11.3% 13.9% 8.2% 10.8%
Runners On 50.9% 14.5% 16.0% 6.1% 12.4%

The first thing to note here: Chen's still thrown the sinker more in pressure situations, albeit not to the extent that he did in his rookie year. Its velocity has increased a bit, though — rising from 91.2 MPH in 2012 to 92.0 thereafter — and with that, it's improved its rate of strikes (64.6%) and home runs (0.6%).

As much as his sinker has gotten better, though, Chen's seen the biggest difference from his slider. Whereas it carried an average velocity of 81.1 MPH three seasons ago, it's gone 83.1 MPH since then, with more vertical movement to match. Unsurprisingly, Chen's accumulated far more whiffs with it (14.4%) than he has with any other pitch; he also owns a 65.3% strike rate and 0.6% home run rate on the pitch. More of these, along with those sinkers, has granted him more effectiveness with runners on base.

Like with Gonzalez, Chen has benefited from some random variation in this transition. A consistently lower BABIP with men on base (despite mediocre hard- and soft-hit rates) has aided his stranding ability, and we shouldn't expect him to sustain that. Moreover, misfortune likely caused much of Chen's first-year sinker struggles; erasing that has certainly played a role in his turnaround.

Nevertheless, there's enough evidence here for me to foresee continued overperformance in Chen's future. I'd certainly take the over on his 71.8% projected rest-of-season LOB%, and would count on him to leave plenty of men on base in the years to come. If the Orioles do re-up Chen, they might have to worry about his early-season velocity, but his play in the clutch shouldn't concern them.

01 July 2015

June Rundown of MASN Articles

Another month has gone by, which means another month's worth of contributions by Camden Depot for MASNsports.com. Here are those articles for June:

Brach's leverage in O's bullpen is improving: Brad Brach has improved his peripheral numbers this season, and he could find himself with a bigger role in the Orioles' bullpen next season and beyond.

What can be expected of Nolan Reimold? After it was announced that the Orioles would promote Nolan Reimold, I took a look at his career numbers and injury history.

Another scrap heap success story for O's: Building a strong bullpen isn't always easy, yet the Orioles have done a solid job of building one every year under Buck Showalter. Chaz Roe is the latest unexpected relief addition to that successful group.

Delving into Ubaldo Jimenez's revival: Ubaldo Jimenez has been one of the success stories for the O's as the season's midpoint approaches. Why has he been so good? And should we have seen it coming?

Is it time to dump Norris from the rotation? Bud Norris has not pitched well at all this season, and he is blocking Kevin Gausman from making regular starts at the major league level. Gausman demonstrated last year that he's ready to pitch every fifth day in the majors, and the O's still haven't opened a spot for him. That should change.

As always, thanks for reading.

26 June 2015

Why Chris Tillman Has Struggled: Part 3

(This post doesn't include stats from Sunday's game).

In the previous parts of this series, I noted that Tillman was struggling because he’s giving up too many walks with no one on base primarily because he was struggling to throw strikes with his changeup and his fastball against right handed batters and partly because he’s having terrible luck with men on base.  He’s giving up the same amount of production with men on base as he has in previous years but he’s getting hit harder when there are multiple men on base then when there is only one runner on base.

This next chart (data courtesy of ESPN Stats and Information) shows how effective each pitch has been so far this year compared to 2012 through 2014.

The chart shows that Tillman is throwing more fastballs this year with the bases empty than he has in previous years but batters are swinging less often and getting more balls called. Batters are missing more fastballs with no one on base in 2015 than they did from 2012 to 2014. The difference is even more pronounced when looking at this performance against right handed batters with the bases empty. Batters are chasing his fastball at a high rate but they’re also deciding to swing less frequently. His problem has been that right-handed batters simply aren’t hitting his fastball foul.

Meanwhile, he’s only throwing a third of his changeups for strikes as opposed to over half in previous year. Opposing batters have figured this out and therefore are only swinging at about a third of his changeups opposed to nearly sixty percent in previous years. In addition, they’re also missing at about half the rate as they did previously. It seems that Tillman has also realized he’s having problems with his changeup in these situations because he’s throwing it less often down from 13.8% from 2012 to 2014 to only 11% in 2015.

Tillman is also having problems with this pitch even with men on base. The pitch is in the strike zone as often as it has been in past years as measured by his InZone% but batters are chasing and missing the pitch less frequently this year than in past years resulting in the pitch causing more balls and fewer strikes. This signifies to me that his changeup is broken and therefore the Orioles’ pitching coach should be working with Tillman to fix that pitch.

Tillman is also having problems throwing his curveball for a strike with the bases empty as it used to be in the strike zone 37% of the time from 2012 to 2014 but only 26% of the time this year. He’s having no problems throwing it for strikes with runners on base. Opposing batters have put seventeen curveballs into play and have hit six singles and three doubles good for a .529 BABIP.  He’s also struggling with the pitch when the bases aren’t empty as opposing batters have a .467 BABIP against it with men on base and a .324 BABIP against it with men on scoring position. In previous years, opposing batters had a .208 BABIP against it with no one on base and a .240 BABIP against it with runners in scoring position. He may be having fewer problems throwing his curveball for a strike than throwing his changeup for a strike but it appears that his curveball has regressed this year as well as his changeup.

Tillman’s only saving grace has been his cutter. He’s throwing it for fewer balls than he has in previous years. The problem is that batters are more likely to hit it than they have previously regardless of whether men are on base.  So far, this hasn’t been problematic but it is worrisome.

It’s possible to determine how many balls Tillman’s lack of control has cost him for each pitch. The next chart shows the percentage of balls he’s thrown each pitch with the bases empty, how many more balls that it has resulted in him throwing compared to his 2012 and 2014 average and whether the difference is statistically significant.

Basically, the statistics show that he’s throwing significantly more changeups for balls, significantly more fastballs for balls and significantly more cutters for strikes. Despite only throwing 70 changeups with the bases empty, he’s already thrown 18 fewer for balls than would be expected. Needless to say, it’s not surprising that he’s giving up more walks.

Tillman’s problems this year are that he’s giving up more walks than normal with the bases empty and he’s giving up hits at inopportune times. His pitches also aren’t as sharp as they’ve been in the past. The good news that this is something that’s probably fixable because it’s not like his stuff has degraded but rather his control has struggled. The challenge for the Orioles’ pitching staff is finding out what is causing those problems so that he can get back to being the productive pitcher that he was earlier in his career.

How Should the Orioles Open Roster Spots for Chen and Schoop?

Since the promotion of Nolan Reimold and Chris Parmelee, the Orioles have been putting off a roster decision or two. They controversially optioned Wei-Yin Chen last week to get a closer look at Parmelee, and the move worked out well. Not only has Parmelee raked since joining the team (.333/.400/.815), but Kevin Gausman, who made the spot-start in Toronto in place of Chen, pitched effectively, if not efficiently, and helped the Orioles take two of three entertaining games in a competitive series.

Chen is scheduled to pitch tonight, so he's going to need a place on the roster. Jonathan Schoop could also be activated soon. So what should the Orioles do to create roster space?

One way to put off one of those decisions would be to move Adam Jones to the 15-day disabled list. Jones has been dealing with a nagging shoulder injury, and it's in the O's best interest to get Jones as healthy as possible before his return to the lineup. However, Roch Kubatko of MASN noted in the article above that "it wouldn't be shocking if [Jones] returned to the lineup tonight," so that option could be off the table.

T.J. McFarland could also be sent to Triple-A Norfolk, but that would also leave the O's with six relievers in the bullpen. It would be yet another temporary fix, maybe for a game or two.

The current logjam on the O's roster is due to the team's group of corner outfielders. Steve Pearce, David Lough, Parmelee, Reimold, Travis Snider, and Delmon Young are all candidates to be designated for assignment. Jimmy Paredes, who probably doesn't need to carry a glove with him, was a candidate to be designated a few weeks ago but started hitting again and has a wRC+ of 125. So his roster spot appears to be safe.

If I had to rank the six outfielders above from least least likely to get DFA'd to most, it would be Snider, Parmelee, Pearce, Lough, Young, Reimold. That's a complete guess, and my preferred ranking would be close but slightly different: Snider, Pearce, Parmelee, Lough, Reimold, Young.

Let's start with Delmon Young, who should be the first player to be removed from the roster. Young and Reimold are redundant. Young is a better career hitter against left-handed pitching than Reimold (112 wRC+ vs. 104 wRC+), but Reimold is also not a platoon bat, is a better outfielder, and is better on the basepaths. The Orioles re-signed Young to hit, and he has not done that this season (73 wRC+).

To me, the next move comes down to Lough or Reimold. It's important for the Orioles to keep at least one outfielder on the roster besides Jones who's capable of holding down center field defensively. That's especially important with Jones on the mend. I give the clear defensive nod to Lough, who's both a better outfielder overall than Reimold and also has more experience playing center field in the majors. Reimold is superior offensively to Lough, but simply put, he is not a center fielder.

Lough career as OF: +30 DRS, 31 UZR
Reimold career as OF: -2 DRS, -13 UZR

It wouldn't be ideal to jettison two right-handed bats at the same time, but the Orioles are also getting Schoop back. His return should help balance the lineup when Ryan Flaherty isn't playing.

Really, you can make a case for anyone in this group to be designated for assignment. The Orioles could also work out a trade to alleviate the roster pressure, similar to the trading of Alejandro De Aza to Boston. But Snider should stay because he's hit well enough and is also under team control for 2016. Parmelee should hang around because he's been on fire and it's worth seeing how good he really is. Pearce should stay because while he hasn't hit well, his peripheral numbers aren't that bad and he's hit into some tough luck (still hitting the ball hard; low BABIP). Plus, he can also fill in at first base and even second base and field well enough. Still, with Schoop close to returning, Pearce's ability to now play second base matters slightly less, and Parmelee can also handle first base duties.

Yet another alternative would be to send Flaherty to the minors. Schoop shouldn't be considered an everyday option right away, necessarily, but Pearce could fill in on days when Schoop sits.

But the Orioles seem to revel in making unconventional decisions pushing the limits, so it's wise to expect the unexpected in any case. Considering the crafty roster moves the Orioles have made the last few years, any player with an option is a candidate to be temporarily removed from the active roster. As Jones accurately said last week, "I don't know how they do it. I'm just glad I'm out of options."

24 June 2015

The Non-Wild Pitch

In the first inning of the May 27 Syracuse Nationals - Norfolk Tides game, Syracuse loaded the bases with one out via two singles and a hit-by-pitch. Michael Bowden's 0-2 pitch to Ian Stewart bounced away from catcher Audry Perez. Darin Mastroianni tried to score, but Perez got to the ball quickly and threw to Bowden covering home plate. Bowden tagged Mastroianni before he touched home and Mastroianni was called out. The other runners, Tony Gwynn and Matt Minicozzi, advanced to second and third base, respectively.

How would you score that play? Most fans - even most hardcore baseball fans - would automatically score that play as a wild pitch on which Mastroianni was thrown out. They reason that you have to account for Gwynn and Minicozzi's advance somehow, and that it somehow seems right to charge the pitcher with a wild pitch when he did, in fact, throw a wild pitch.

Those fans are wrong. The correct scoring decision doesn't, to the best of my knowledge, have an official name but it's occasionally referred to as an "Out Advancing" or as a Fielder's Choice. The guiding principle is easily understood - neither a wild pitch nor a passed ball can be charged if a runner is put out without having successfully advanced a base. (The second part of the condition allows a wild pitch to be charged when a runner is thrown out trying to advance a second base - from second to home or from first to third.) Regarding the other runners, they are treated the same way as runners on a double-steal are treated when one of the runners is caught - they are not credited with a stolen base but are considered to have advanced on a fielder's choice.

It's not only fans, but even also the occasional official scorer who doesn't understand this play. A few years ago, there was a similar play at Harbor Park. With runners on first and third, the pitcher's pitched landed in the dirt and bounced a few feet away. The runner on third broke for home, realized that he wasn't going to make it, and tried to return to third. The catcher threw to third in time to retire the runner. While this was going on, the runner on first advanced to second. The game's official scorer insisted that a wild pitch be charged to the pitcher, despite the arguments of myself and the team's media relations director that the scoring rules prohibit it. Eventually, we alerted Major League Baseball Advanced Media that the official scorer wasn't listening to us and that we would have to change the play after he submitted the box score. (MLBAM can unilaterally change official scoring decisions if they contravene the rules or if they are contradicted by facts.)

In fairness to all concerned, the official rule book does not clearly deal with this situation. The rule book is filled with many different scoring rules, principles, and examples, but does not explicitly state the scoring rule or provide an example of this play. A simple declarative subpoint - perhaps something like "Neither a wild pitch nor a passed ball shall be charged if a baserunner is put out without having advanced a base safely." - would clearly instruct the scorer. And, to make it more emphatic, the rule book should include examples of at least three scenarios - a basic wild pitch/passed ball; the out-advancing scenario; the scenario in which a runner is out trying to advance two bases - and the play would be clear.

23 June 2015

Why Chris Tillman Has Struggled in 2015: Part II

(This post doesn't include stats from Sunday's game).

In my previous post on the subject, I noted that one of the major reasons why Chris Tillman has been struggling this year is because he has given up a large percentage of walks with the bases empty primarily because he has thrown a larger percentage of balls with the bases empty this year than in years past. In this post, I wanted to see whether he was having any problems with a specific pitch.

As an affiliate of the ESPN SweetSpot Network, writers for Camden Depot have access to ESPN’s TruMedia Site and can use it to create heat maps that show Tillman’s accuracy for each of his pitches (defined by ESPN as fastball, changeup, curveball and slider (actually his cutter)) in 2015 and from 2012 through 2014 with the bases empty.

Below are charts showing where he's thrown his fastball with the bases empty from 2012 to 2014.


And here's a picture with the bases empty in 2015.


Tillman has been throwing strikes with his fastball albeit a bit up in the zone in 2015 with the bases empty. Also, his low fastballs have been out of the strike zone more often this year than in years past. But things more or less look good, right? Well, it turns out that he's having some struggles throwing his fastball against right handed batters with no one on base.

This shows how he did against righties from 2012 to 2014.

And this shows how he's done this year in the same situation.


His problem has been that he’s been throwing his fastball outside more often this year than in past years against right handed batters. He’s thrown 20.3% of his fastballs outside this year as opposed to 18% in previous year. It's not a major problem but Tilman throws his fastball more often than any other pitch and therefore even a minor problem can have a huge impact. I didn't notice any problems with his fastball against left handed batters.

Here's Tillman's change-up with no one on base from 2012 to 2014.

And here it is in 2015.


I think it's safe to say that Tillman's changeup has been a disaster with no one on base this year because that chart is ugly. I mean, he's not coming close to the strike zone most of the time. His problems become clearer when we further filter the results by whether the batter is a lefty or a righty.

Here's how he's thrown his changeup against lefties in 2012 to 2014.


And here's the pitch in 2015.


From 2012-2014, Tillman was able to place the pitch in outer part of the strike zone. This year, he's pretty much been solely throwing the pitch outside and nowhere near the strike zone. That's simply not going to work in the majors.

This is how his changeup looked from 2012 to 2014 with the bases empty against right handed batters.

And here is how it has looked against right handed batters in 2015.


His changeup was being located properly in 2012 to 2014 but this year it's simply been thrown low and outside of the zone. The bottom line is that opposing batters don't need to worry about his changeup because he can't locate it. If they hold off, then it will likely be a ball anyway. Obviously, this puts Tillman at a disadvantage.

Surprisingly, as we'll see in the next post, he's only having this problem when there aren't runners on base. It's reasonably accurate once runners are on base which could indicate a possible mechanical problem. In any event, this seems to be a major reason why Tillman is struggling.

Next up is the curveball. Here's how it looked from 2012 to 2014 with no one on base.

And here's how it looked in 2015.


Tillman hasn't had much accuracy throwing his curveball for a strike this year with no one on base and more of these pitches are balls this year compared to other years. However, there are two mitigating factors. The first is that Tillman hasn’t thrown very many curveballs this year and therefore it has less impact than pitches thrown more regularly. The second is that this pitch isn't necessarily meant to be thrown in the strike zone for a called strike but rather to get a batter to chase a bad pitch (that he'll hopefully miss). His curveball could use some improvement but probably isn't his major problem.

The final pitch is his cutter. Here's how it looked from 2012 to 2014 with no one on base.


Here's how it looked in 2015.