There has apparently been some worry about starter James Shields this year who now sports the second highest ERA of the rotation. Tommy went over his thoughts yesterday and believes that Shields may have been placing his fastball over the heart pf the plate more frequently as of late. I noted in the comment section that overall, his BABIP has been extremely high and much higher than his career average. Take a look at the year by year numbers:
Other than his first half season in in 2006, this season is clearly the outlier. He now has the sixth highest BABIP among qualified starters this season. That doesn't mean Shields has been pitching badly this season. More of a luck thing as the DIPS stats suggest that he is pitching about the same if not better this season.
Back to the original stat subject, I ran the numbers and found the BABIP for each pitch type for Shields compared to the numbers from the 2008-2009 seasons to this season. This should tell if his fastball is giving up most of his hits which he may have been placing over the plate too often as noted in Tommy's article.
The four seam fastball is higher this season but it didn't increase as much as the two seamer which shot up about .150 points. Maybe it is the small sample size, but it has been put into play only just slightly higher as with all his pitches except his cutter and curve, which has been put into play less this season compared to the previous two.
Without the Hit f/x data, we won't know the true nature of those hits and truly find if he is lucky or not. Instead, maybe some metrics like In Wide Zone (IWZ) or contact rate might tell us enough if Shields has been more prone to the base hit this season.
Looking at the raw numbers, Shields four seam fastball as been in the wide zone (two feet long, Gameday operators adjusted height) about 4% more this season and has been made contact with about 3% more. The two seamer has changed more dramatically but it hasn't been thrown as much as the four seamer so it may be noise. Only the changeup has been in the zone less and has been missing more bats this season. Overall, Shields has been in the wide zone only about one percent more this season than the previous two (.533 to .541). And the contact rate is only up about one and a half percent over the same time span.
BABIP is a stat that is expected to regress over the course of a season to around .300. Of course there are deviations both ways with Shields deviating towards the high end so far this season. The reason I expect him to regress downwards is that his career average is around .310 now and he had an average of .301 during his three previous full seasons (2007-2009). And after his horrendous start against the Florida Marlins last weekend, Shields rebounded nicely against the Atlanta Braves yesterday with only five hits on 18 balls in play. Like Tommy said yesterday, it is only a "short-term blip on the radar". Shields is smart enough and has the experience to figure out what is wrong and work to correct it to provide strong starting for the Rays.
Data from Fan Graphs and MLBAM.