The Oakland Athletics have signed Hideki Okajima to a minor league deal. Okajima is a Japanese native that pitched in the NPB from 1995 to 2006 before pitching for the Red Sox from 2007 to 2011. The left-handed reliever proved to be very effective in the Majors before taking a big step back in 2010 and designated for assignment early in 2011 and spending most of the year in AAA. The Yankees then signed him to a minor league deal, but the physical raised some concerns about his shoulder and he was released. Okajima then went back to Japan and pitched for the Softbank Hawks, his 3rd different NPB team. He threw 47.2 innings, and walked just 6 batters and had a .94 ERA. Of course, defensive independent metrics liked him less, as he had a 3.41 kwERA and 2.33 FIP. That is a 79 kwERA – and (using Softbank’s 3 year park factor) 74 FIP -, so even though Okajima wasn’t as historic in 2012 as his ERA seems, he was still really good in the NPB.
I’ll be honest, a couple of months ago, I was surprised to hear that Okajima was interested in coming back to the states because he had lost so much fastball velocity. NPB Tracker says he averaged 85.39 MPH on his fastball in 2012. He has never had a “good” fastball, but averaged close to 88 MPH on his fastball when he first came to the Majors. Even in 2011, when he struggled, he averaged 86.9 MPH on his fastball (and this is according to MLB AM data, according to Brooks Baseball, he was about a MPH higher each time). That is a drastic drop in fastball velocity, not to mention that this velocity doesn’t really play in the Majors. In 2012 in the Majors, Livan Hernandez was the only non sidearm pitcher that threw softer than Okajima’s 2012 velocity, and he was horrible. He also seemed to wear down as the year went along, as Okajima averaged under 85 MPH in his last two outings, and 3 of his last 4.
Of course, especially with a change of systems, there is the problem of pitch classification. MLB AM data says he throws 5 different pitches, though Brooks Baseball and NPB Tracker agrees that he throws 3 different pitches, a fastball, a changeup, and a curveball. As his fastball did, his curveball and changeup lost over 2 full MPH. He also changed his pitch selection some, going with more changeups (throwing it a 3rd of the time) and throwing the curveball less. In the Majors, his change was the slightly better pitch, so this may be a change for the better. I wanted to dig deeper into Okajima’s season though, and see if I could gather some more advanced data.
I used NPB Tracker data and looked at every outing and plugged in his heat maps and results for every pitch (he threw 785). I did it manually, so there may be an occasional mistake. The heat map looks really ugly, but it gives you the general idea. The middle 9 zones are the strike zone. For the results, I changed line drives to flyballs and just conflated the two.
As you can see, the most common location was the arm side middle part of the strike zone. His least frequent place in the strike zone was right down the middle. Overall, his least frequent locations were high and arm side, arm side middle out of the zone, and glove side middle out of the zone. According to the maps, he threw pitches in the strike zone 51.6 % of the time. For comparison, the average zone % of MLB pitchers was 44.9% in 2012. Okajima’s career zone % is 49.7%, suggesting that he threw more pitches in the strike zone in 2012 than he did in the Majors. When you look at just pure strike percentage, by adding all the batted balls and strike swingings and lookings and dividing it by the total + balls, Okajima threw strikes about 69.6 % of the time. In the Majors, he threw 63.2 % of the time, right around the league average (in 2012) of 63.5 %. I think a lot of this had to with the fact that he got swinging strikes 11.46 % of the time, higher than his MLB career average of 9.9%. Even though I am sure that he knew he would be able to get away with it more since he was facing inferior competition in 2012, it is hard to imagine that the now 37 year old (36 during the 2012 season) was suddenly able to throw more pitches in the strike zone. This makes me think that the heat maps might have a little swing/no swing bias in them, meaning that the program may put pitches that are not actually in the strike zone, on the fringe of the strike zone if they are swung at. Obviously, there is no way to prove this without operating the program itself or going back and looking at every single pitch (or at least a large batch of pitches). Since full NPB games are not really available to watch on demand in the United States, this would be pretty much impossible (not to mention extremely tedious if it were possible). According to the heat map, he liked to throw glove side (43.9 %) more than arm side (40.6%), and liked to throw in the lower portion of the strike zone the most. Up high above the strike zone was his least favorite place to throw the ball height wise. How does this compare to how he threw when he pitched in the Majors (just the data from Brooks Baseball, normalized, without the graph part)?
Obviously this is much different. The place he threw the most pitches (though less than league averages) was right down the middle. It appears he threw more pitches in the middle part of the plate than in the lower part of the plate. Compared to league averages, he threw a lot of pitches up high, and he didn’t throw a lot of pitches low compared to league average.
So what else does the data tell us? First off, he gave up what seemed like a large amount of foul balls. MLB average Fouls/Strikes was 27% in 2012 (I keep using MLB averages as comparison because we don’t have this stuff for the NPB). Okajima’s was 33.7 %. I usually think that giving up a lot of foul balls is a bad thing, as it means that you can’t put away hitters. Just from a cursory look at MLB pitchers, there seems to be some validity to it. Bruce Chen gave up the most foul balls by percentage in 2012, and Phil Hughes and Travis Wood were 2nd and 3rd. However, Jason Marquis had the lowest foul/strike percentage in 2012, and Edinson Volquez, Derek Lowe, and Joe Blanton also were extremely low. Matt Cain, Blake Beavan, Cliff Lee, Justin Verlander, Jeremy Hellickson, and Jordan Lyles all had about the same foul/strike %. That is a really diverse group, and this seems to point to the fact that fouls don’t mean a lot in predicting how a pitcher does. With that said, Okajima’s was higher in the NPB that Bruce Chen’s (the leader out of qualified pitchers) was in the Majors. This is probably not a good sign.
While Okajima’s official season statistics say he gave up just one homer, the NPB tracker had two different homers he gave up. I think the 2nd one was in the playoffs. Out of the 59 flyballs Okajima gave up (flyball outs + flyball singles+ home runs), just 3.9% of them turned into homers. I don’t like looking at HR/FB %, especially for other leagues, just because there are so many variables. Okajima could have been just better at limiting good contact, there is obviously less power in the NPB than in the Majors, and it could just be luck. There are too many options to explain the difference without a clear way to decipher which one it is. For what it is worth, Okajima was good at limiting homers in the Majors, with a 8.2 HR/FB %, and .95 HR/9IP.
Okajima had a 43 GB % on batted balls, which is usually below average. In the Majors, he was a well below league average ground-ball guy, around 36 %. As mentioned above when talking about his defensive independent metrics, Okajima was “BABIP lucky”. His BABIP on groundballs was .183, but .315 on flyballs. So he had a disproportionate amount of groundballs eaten up, while he had an inordinate amount of fly-balls turned into hits. This is a little strange considering how many infield flyballs Okajima got when he was in the majors.
Oliver, the projection system, really likes Okajima for 2013, projecting him to have a 3.36 FIP. Only his first year in the Majors was better than that. Of course, projecting someone to pitch in Oakland will make him look better than a pitcher pitching in Boston. Of course Okajima is older now, and has struggled in his last two years in the Majors. Using kwERA to tease out the park might help. He is projected to have a 4.25 kwERA in 2013 by Oliver, compared to his career average of 3.91 kwERA and 3.38 kwERA in 2007. So it does seem that the big difference is the park. Okajima already limits homers, and Oakland will only help this.