Category Archives: Research

A repository for research articles

Baseball, Commentary, Paid Content, Research

Strike while the iron is hot

Leave a comment

While we’re on the subject of called balls and strikes in Nippon Professional Baseball, @Student_murmur has done a nice study in Delta Graphs about how the count affects umpires calls, which you can find in Japanese HERE.

Lacking pitch tracking data and having to depend on Mark I eyeball technology, the researcher drew a line around the strike zone in different counts where calls went 50-50. It’s no surprise that the smallest zone accompanies the 0-2 count, and the largest the 3-0 count.

The data from 2017 and 2018 collected by Delta Graphs looks fairly consistent and are consistent with the findings of John Walsh in 2010, which you can find HERE. Walsh found that the area of the strike zone’s vertical plane in MLB was 3.52 square feet with a 3-0 count and 2.42 square feet with an 0-2 count.

The last post here dealt with pitch locations derived from Nikkan Sports’ pitch-by-pitch data in 2018. These consisted of designated as inside the edge of the zone and outside the edge. I labeled them all borderline pitches, either borderline ball locations or borderline strike locations.

Looking at pitches in these two border zones that were called balls and strikes on 3-0, 0-2 and all other counts in 2018 produces the following tables, where 78 percent of all borderline pitches were called balls in counts other than 3-0 and 0-2. In 3-0 counts it was 60 percent. In 0-2 counts it was 92 percent.

We need to understand that in this study and the Delta Graphs one, the data points are being observed by someone else’s eyeballs and recorded by someone else’s hands. Also, pitchers can worry less about balls and strikes in certain counts. Unfortunately, the same seems to go for umpires.

Borderline called NPB balls and strikes in 2018 with 0-2 counts

CountBorder ZoneBallStrikeTotal
0-2Strike Zone1182183
0-2Ball Zone2,12242,126
0-2Total2,1231862,309
Percent91.94%8.06%

Borderline called NPB balls and strikes in 2018 with 3-0 counts

CountBorder ZoneBallStrikeTotal
3-0Strike Zone0349349
3-0Ball Zone5243527
3-0Total524352876
Percent59.82%40.18%

Borderline called NPB balls and strikes in 2018 with all other counts

CountBorder ZoneBallStrikeTotal
OtherStrike Zone211,06011,062
OtherBall Zone39,50413539,639
OtherTotal39,50611,19550,701
Percent77.92%22.08%
Baseball, Commentary, News, Old stuff, Paid Content, Research

Problems with punch-outs

Leave a comment

The other day HERE I tried to answer the question whether foreign hitters in Nippon Professional Baseball have larger strike zones or not by looking at the percentage of strikeouts that are decided by a called third strike.

Having done that, I realized that individual variation makes such an analysis really, really murky. Some players hack, some are more disciplined. Pitchers that lack a good swing-and-miss pitch should conceivably have a higher CST (Called Third Strike) percentage.

Still, the study did lead to an interesting observation about the nature of Japan’s two leagues. As some of you know, either the Pacific League is the stronger of Japan’s two leagues or it’s just really, really good at hiding that fact, considering how poorly Central League teams do in interleague play and in the Japan Series.

The umpire merger

From 2003 until 2010, Pacific League position players were taking called third strikes in 20.53 percent of their strikeouts. In the Central League, the percentage was 22.84. The PL was dominated at the time by huge ballparks, where home runs were less frequent.

Since the umpires of the two leagues merged from the start of thew 2011 season, the PL called-third-strike percentage rose to 21.57, while the CL’s dropped slightly to 22.73.

The managers

In the previous article, I suggested that managers might be affecting how often called third strikes went their teams’ ways. But that was probably incorrect, for the same reason that judging individual hitters is fraught with danger. Unless you have the photographic evidence of the pitches in question, you can’t really tell.

Managers WILL effect the number of third strikes called against their team because of their batting and pitching policies. A look at how each manager’s team did relative to its league, shows that from 2003 to 2018 shows some interesting stuff, but it’s just that: interesting stuff.

Consider curmudgeonly “kantoku” Katsuya Nomura. His Rakuten Eagles struck out 3,369 times over his four seasons in charge, and his players went down on called third strikes 6.03 percent more often than the league, his Eagles were taken out of at-bats by umpires 203 extra times.

Anyway, the manager whose teams have ostensibly benefited the most from the umpires’ calls were Hisanobu Watanabe (Seibu Lions) with 92 fewer called third strikes on his hitters and 106 extra called strikes for his pitchers. No. 2 on this list (2003-2018) is Trey Hillman overall +194, and Koichi Ogata (+186). At the other end are: Nomura (-242), Hideki Kuriyama (-239), Akinobu Okada (-137) and Bobby Valentine (-136).

That’s interesting, but if you look at those Eagles hitters, what do you see? Tons of walks, few strikeouts. That was a team led on offense by Takeshi Yamasaki, a power hitter who frustrated managers and teammates by taking tons of called third strikes. He rarely swung at a two-strike pitch if he thought it might be outside of his zone, often putting his fate in the umps’ hands.

Again, I don’t think there is anything the least bit instructive about those. I just thought they were fun. But as mentioned above, managers can have a real effect on how their teams play. Take the DeNA BayStars, for example.

The Alex Ramirez effect

When Alex Ramirez took over the DeNA BayStars in 2015, his most public policy was telling his players to “swing at the first strike.”

What happens when batters execute this tactic? Here’s what happened in 2018, comparing the results of 17,792 plate appearances started by a swing or a first-pitch ball (as the 2015 BayStars were instructed to execute), and the the 48,046 PAs in which the first pitch was taken.

OptionPAAvg.OBPSlug
Swing 1st-pitch strike, take 1st-pitch ball17,792.272.305.424
Take 1st pitch regardless48,046.251.336.392

There is overlap of course, since first-pitch balls fall into both camps. But those looking to drill the first strike hit for average and more power, but paid for it in more outs and a lower on-base percentage. This goes a little bit in explaining the career of Ramirez — a guy who hit for good average, became the first foreign-registered player with 2,000 hits, and hit for good power, but didn’t draw walks unless he had to.