Blog Posts

Steamer Update 2/12/14

The first thing that you should know (and hopefully already do know) is that we’re up on Fangraphs. You can download the hitter and pitcher sheets from there by clicking “Export Data.”

The playing time for these projections is based on the Fangraphs Depth Charts which are updated daily by a team of experts and should get more and more accurate as the season approaches.

Yesterday, we updated our projected lineups. During the season this year, our projected lineups will update based on teams’ actual lineups with recent games weighed most heavily and our “rest-of-season” R, RBI and playing time projections will reflect these updates.

We’re hard at work on our code (some genuine improvements along with some spring cleaning) so you can expect a couple of additional updates within the week.

The Hollow Under the Knee

Dave Cameron recently suggested that the rise in strikeouts over the last 6 years could be at least partially due to changes in umpiring.  On The Book Blog forum, mcbrown focused in on 0-0 counts with the bases empty and looked at called strike rates, ball rates and swing rates by year.  I thought it might be interesting to follow his lead and look at pitches taken on 0-0 counts with the bases empty (right-handed hitters only) and see if the called strike zone has changed.  Brian Mills and David Golebiewski have looked at this before.


It’s not obvious or dramatic but it looks like the bottom of the called strike zone has moved down a bit matching what David Golebiewski found when looking at this after the 2011 season.  Look at 2008 and 2009, the center of the bottom dashed line falls in the sky blue range indicating something like a 35% chance of a strike call.  By 2011 and 2012, it’s in the green, suggesting roughly a 50% chance of a strike call.

There’s another little bit of trickiness based on how I collected the data above.  Hoping to get a less hazy look at the bottom and top of the strike zone, I limited the data to pitches on which the mlbam stringer had entered an “sz_bot” and “sz_top” (indicated the bottom and top of the strikezone) that were within 0.1 feet of the average values over the five year span.  This would throw out the tall/upright and short/crouching batters.  But maybe all we’re seeing in the images above are changes in how the stringers enter the bottom of the strike zone?  Brian Mills looked at this issue as well as the issue of systematic bias in strike zone designations.  He suggests that there were a changes in how sz_bot and sz_top were assigned some time in 2007.  I left 2007 out of this analysis completely.  To check for moving strike zone designations, I looked at batters who stepped up to the plate in all five years (339 of them) and looked at the average bottom and top that had been assigned to each batter in each year and then the average of those averages for each year.

2008 2009 2010 2011 2012
bottom 1.59 1.63 1.68 1.61 1.60
top 3.44 3.44 3.46 3.43 3.45

It looks like something a little funny happened in  2010 where for the same group of hitters the average sz_bot snuck up by about an inch but there doesn’t appear to be a trend in designated strike zone size.

So, has the bottom of the strike zone moved down?  And how much difference could this make?