The 2017 Strike Zone Mid-Season Report

Originally published on sonsofsamhorn.com July 20, 2017; scraped from archive.org in 2018.  

Following the introduction of PITCHf/x – and especially following the umpire’s union allowing video review of their ball/strike calls – the MLB strike zone rapidly began to change its size and shape. Between 2010 and 2014, the strike zone became much taller, and somewhat narrower; this was probably one cause of the reduced offense in that period. 

After 2014, the zone has changed slightly, but there haven’t been the dramatic shifts witnessed in the few years before that. In fact, from 2015 to 2016, the zone actually shrank slightly, for the first time in the PITCHf/x era:

We can see that there were some differences between 2016 and 2015. For the first time since PITCHf/x data became widely available in 2008, the strike zone shrank slightly in 2016. Most of the change was at the bottom of the zone, which moved up for both left- and right-handed batters, while the top of the zone also moved upward slightly. The outside parts of each strike zone also shrank a little, while the inside either stayed unchanged or perhaps expanded a tiny amount. Overall, the zone shifted very slightly upward and inward in 2016, and became a little bit smaller, compared to 2015.

In the first half of 2017, the zone has again remained fairly constant. The chart below shows what umpires have called so far. In the blue areas, they almost always called a ball; in the green, almost always a strike. The red areas show the region where a pitch was roughly equally likely to be called a ball or a strike. (This is from the umpire’s viewpoint.) I’ve drawn a polygon (white) just about in the middle of the 2016 red zone, which I call the “strike zone” for 2016, and kept that constant while showing the called pitches from 2008 (before the zone began to enlarge), 2016, and the first half of 2017:

The difference from 2008 to 2016 is very obvious; not so much the difference from 2016 to 2017.

To highlight differences, here is a comparison plot (again from the umpire’s view), with differences between the two seasons shown. Regions in which a 2017 pitch was less likely to be called a strike than in 2016 are blue; areas in which a 2017 pitch was more likely to be called a strike are in red.

Basically, it looks as if the entire strike zone has shifted slightly over to the umpire’s right, without changing size significantly. The difference is quite small – less than an inch – and it may not be real. In 2017, a new pitch tracking system was put into place; instead of PITCHf/x, data are now collected by the TrackMan system. Perhaps on average, the new system is calibrated very slightly differently from the old one? 

In any case, the important point is that the strike zone hasn’t changed its size significantly since last year, so batters and pitchers haven’t had to recalibrate their expectations, which is much more important than the apparatus.

Posted in Uncategorized | Leave a comment

Catcher Framing: The Mid-Season Report


Originally published on sonsofsamhorn.com on July 18, 2017; scraped from archive.org Dec. 2018, possibly with some missing images

A catcher can add – or subtract – several wins to his team by framing pitches, catching pitches near the edge of the strike zone so the umpire is fooled into thinking a ball is a strike, or vice-versa. Catchers who are good at this can add several extra called strikes per game, and over the course of a full season, several hundred extra called strikes add up to a handful of extra wins.

Once a catcher has caught about 40-50 games – a third to a half season or so – we can get a reasonable idea of his true framing ability. I looked at the 32 catchers who have caught over 35 games this season and calculated their framing ability. Briefly, I divide the area around the strike zone into 3-inch squares, and for each catcher, I compare his ball/called strike percentage to the rest of the league. (For more detail, see here.) 

The full list, with breakdowns of framing vs. right- and left-handed batters, can be seen here. The best catcher at framing so far this season has been Tyler Flowers, by a long way. Flowers has added 3.37 extra strikes per game to his pitchers’ total. Overall, Flowers has contributed nearly 162 strikes more than a league-average catcher would have; that is roughly the equivalent of 21 runs that Flowers has added to his team through his framing alone. Since Flowers has also been an unexpectedly good hitter this season, the Braves have received excellent production from their starting catcher so far.

We can look at a chart of Flowers’ framing effects by shading the strike zone red in areas where Flowers is more likely to get a strike than average, and blue where he is less likely to get a called strike:

For both left- and right-handed batters, Flowers has excelled at getting extra strikes from the bottom right corner of the zone (“right” from the viewpoint of the umpire). At the top and left of the strike zone, he has been roughly neutral or even slightly worse than average, but he has expanded that bottom corner to a remarkable extent. 

Just to show what that looks like a little more meaningfully, here are the locations of the pitches that were “mis-called” for Flowers – pitches outside the standard strike zone (the gray polygon) that were called strikes, and those inside the zone that were called balls:

Flowers has been an excellent framing catcher for several years, so his dominance this season is not a surprise. Less expected is the worst framing catcher of 2017 (35 game or more): Jonathon Lucroy. Lucroy was an excellent framer as recently as 2014, and in 2015 and 2016, though his skills were clearly eroding, he was at least roughly league average. So far in 2017, Lucroy has been abysmal as a framer, losing about 2.27 strikes per game compared to the average framer. Lucroy’s pitchers have lost nearly 123 strikes that a league-average framer would have given them:

Lucroy has no good framing areas; he lost strikes all around the strike zone for both left- and right-handed batters. If he has any good spots, it might be the bottom right corner against right-handed batters, where at least he gained a few extra strikes, but he also lost many pitches in that area that should have been called strikes, but that he made into balls:

Rounding off the top of the framing list in 2017 are Martin Maldonado (1.84 extra strikes per game), Yasmani Grandal (1.80), Caleb Joseph (1.74), and Christian Vazquez (1.48). Here are Madonado and Grandal’s charts; both have clear strong and weak zones, in contrast to Flowers:

The bottom five are Lucroy, James McCann (losing 2.22 strikes per game), Cameron Rupp (1.94), Mike Zunino (1.39), and Salvador Perez (0.81). Here are McCann and Rupp, who are at least close to neutral at the top of the strike zone, though they lose many strikes at the bottom:

There has been some speculation that catcher framing might be losing its importance as it has become a standard measurement. With umpires being aware of catchers’ ability to steal strikes, they might be harder to fool, and with managers and coaches being aware, the skill might be so universally taught and selected for, that only good framers are left in the majors. But that doesn’t seem to be the case. The range of framing we see so far in 2017 is pretty much the same as we’ve seen in most recent years. Framing remains a genuine catcher skill that can significantly affect a team’s win/loss record at the end of the year.

Posted in Uncategorized | Leave a comment

How Did the Umpires Call the Strike Zone in 2016?

Originally published on sonsofsamhorn.com in January 2017; scraped from archive.org in 2018.  

The average strike zone in 2016 was similar to 2015’s in size and shape, finally giving MLB players some consistency from year to year. From day to day, though, their strike zone may have changed as different umpires called balls and strikes, each with their own tendencies and preferences. How much did umpires’ personal strike zones differ in 2016?

I took the umpires who called at least 30 games in 2016 and calculated the size of their strike zone, compared to the league average. Briefly, I divided the area around the strike zones for left- and right-handed batters into three-inch squares, asked how many pitches each umpire saw in each zone and how many he called strikes, and then compared to the percentage of strikes in that box across all baseball. The end result is expressed as “extra strikes per game”, compared to what an absolutely average umpire would have called. The full list is here

The three umpires with the smallest strike zones (the most batter-friendly) were Jerry Meals (who called 3.7 fewer strikes per game than the average umpire would have), Joe West (-2.6), and Alfonso Marquez (-2.5). The three largest zones belong to Bill Miller (4.4 extra strikes per game more than the average umpire), Mike Estabrook (3.8), and Jim Wolf (3.6). 

Umpires are generally fairly consistent year over year, with the same umpires tending to have large or small strike zones. Joe West has called a smaller-than-average zone for the past six years; Miller has had a larger-than-average strike zone every year since 2008, when PITCHf/x was fully available. 

We can look at umpires’ calls in more detail by plotting differential strike probability. In the following charts, each square is colored red to indicate if an umpire is more likely to call pitches there a strike, or blue if they are less likely. The intensity of the color reflects the difference in likelihood. 

Here are what Jerry Meals’s and Joe West’s small strike zones look like:

Umpires Call
Umpires Call

Both of them are less likely to call strikes for left- and right- handed batters. West actually is slightly more likely to call strikes inside to righties or outside to lefties, and Meals shows a similar, though weaker, trend. Both of them, but especially Meals, have stronger tendencies for lefties, since the colors on the RHB plots are relatively muted.

Here are the large strike zones of Bill Miller and Mike Estabrook:

Umpires Call
Umpires Call

These are just big zones, to all batters and on all sides of the zones. 

Even for umpires who call a nearly neutral zone overall, there may be preferences for certain regions. Here is Chris Conroy, who has dramatic differences in the way he calls strikes to left- and right-handed batters, but ends up almost exactly average (-0.2 extra strikes per game overall):

Umpires Call

To left-handed batters, Conroy calls quite a large strike zone, but righties have a small one. 

Umpires Call

Most umpires have slight preferences, but few are as dramatic as the ones I’ve mentioned here; the majority are essentially neutral. Here is the distribution of extra strikes by umpire:

Of the 67 umpires who called 30 or more games in 2016, more than half (37) were within one strike per game of being average in either direction, and only 13 were more than two strikes per game away from average.

Check out the complete list to see how umpires’ strike zones have changed over the past several years by clicking here.

Posted in Uncategorized | Leave a comment

What Did the 2016 Strike Zone Look Like?

Originally published on sonsofsamhorn.com Nov. 16, 2016; scraped from archive.org in 2018.  

Starting in 2010, the MLB strike zone has changed size and shape dramatically, probably because umpires started using (and being evaluated on) electronic review of their strike calls. There were large changes in the zone between 2010 and 2011, and from 2013 to 2014. However, there was very little change in the size or shape of the strike zone from 2014 to 2015. Have the umpires finally settled on a zone they’re comfortable with?

To evaluate the strike zone, I use PITCHf/x data to collect the location of all the regular-season called strikes and balls, and calculate the probability of a strike being called in each sub-region around the strike zone. The area in which the probability of a strike being called is around 50-50 marks the edge of the de facto strike zone. This is what it looked like in 2016 for left-handed and right-handed batters. Areas in blue have a very low probability of being called a strike; those in green were very likely to be called strikes; and the region in which balls and strikes are equally likely is red. These charts are from the umpire’s viewpoint, so the batters would be standing in the middle of the two charts:

2016 Strike Zone

That’s not very informative on its own. Here we can compare 2008 (before the zone started to expand) to 2015 and 2016:

2016 Strike Zone

The 2015 and 2016 zones are not identical, but there isn’t a lot of difference between them; both are much taller and narrower than the 2008 zone was.

We can also directly compare the zones and highlight differences. Here, regions in which a strike was more likely to be called in 2016 than in 2008 are shown in red, and regions in which strikes were less likely to be called are blue. The intensity of the color represents the probability:

2016 Strike Zone

Again, this highlights the taller, narrower zone that was called in 2016 compared to 2008. The zone expanded most at the bottom while the sides shrank equally dramatically, especially for left-handed batters; the top of the zone also moved upward, but didn’t change as much as the bottom did. 

Now doing the same comparison for 2016 to 2015:

2016 Strike Zone

We can see that there were some differences between 2016 and 2015. For the first time since PITCHf/x data became widely available in 2008, the strike zone shrank slightly in 2016. Most of the change was at the bottom of the zone, which moved up for both left- and right-handed batters, while the top of the zone also moved upward slightly. The outside parts of each strike zone also shrank a little, while the inside either stayed unchanged or perhaps expanded a tiny amount. Overall, the zone shifted very slightly upward and inward in 2016, and became a little bit smaller, compared to 2015.

Posted in Uncategorized | Leave a comment

Why doesn’t Joe Kelly dominate?

Joe Kelly, with his 100 mph fastball, looks like he should be an absolutely dominant pitcher, but his actual results this year have been mediocre at best.  At sonsofsamhorn.com, I looked at Kelly so far in 2015, focusing on his fastball.  I think that

Kelly this year seems to be be trapped by his fastballs. If he throws them as hard as he is capable of, they are less likely to turn into hits, but more likely to become walks. If he slows his fastball down, improving his ability to throw strikes, the pitches become much more hittable.

Joe Kelly fastballs: Balls per 100 pitches, sorted by pitch speed

Posted in Uncategorized | Leave a comment

Henry Owens debut

I took a look at the major-league debut of the Red Sox’ top-50 prospect, Henry Owens, on Aug 4/2015.

Henry Owens strikes out Jacoby Ellsbury on a slider well outside the strike zone

Posted in Uncategorized | Leave a comment

What happened to pitching in 2001?

Most of the trends in these pitching charts are easy to understand:

ERA and WHIP since 1960

It’s easy to see the Year of the Pitcher in 1968, the recovery as baseball adjusted the pitching mound, the sudden increase during the Steroid Years, and the recent crash in offense partially associated with the increase in the size of the strike zone.

But there’s an a abrupt drop in WHIP and (to a lesser extent) ERA between 2000 and 2001, and I don’t know of any rule change, or other change in the game, that would drive that.  Steroids were still winked at, offense remained at a very high level, the strike zone wasn’t officially changed — what happened?

Posted in Uncategorized | Leave a comment

Justin Masterson’s velocity

As of his game today (June 28) he’s faster than his last few games for the major-league Sox, but barely back to where he started the year, and nowhere near where he was a couple years ago.

Justin Masterson pitch speed, 2015

He probably needs to pick up another couple mph at the least to have a chance at consistent effectiveness at the major-league level.

Posted in Uncategorized | Leave a comment

What’s up with Rick Porcello?

Rick Porcello hasn’t been pitching very well for the Red Sox so far.  Although there’s a meme going around that Porcello needs to use his two-seam fastball more, in an article I wrote (Location is the Key for Rick Porcello) I suggested that his pitch mix is not the problem, while the location of his four-seam fastball might be.

Rick Porcello: Pitch characteristics (2015)

Posted in Uncategorized | Leave a comment

Are Rookies Having a Harder Time?

There have been suggestions from the Red Sox management that the gap between AAA and the major leagues is wider today than ever before. I looked at the claim (Are Rookies Having a Harder Time?) and didn’t find much evidence to support it.

Rookie wRC+ by year

Posted in Uncategorized | Leave a comment
« Older