Thursday, May 08, 2014

Swings and Misses: Surprising Changes in 2014 Hitter Strikeout Rates

Todd Frazier is striking out far less often in 2014.
Photo Credit: Andrew Mascharka
It's May.  And when it comes to trying to draw anything from individual player statistics, we're still in small sample size land..  But short of just reporting what players have done thus far, most of us are at the point that we want to start using the 2014 data to actually infer something about changes in talent.  But with most everyday players only logging between 100 and 150 PA's, what can we actually say?

In Pizza Cutter's now-classic study (which he re-did here), he found that the first statistics to stabilize in a season--meaning the first stats that can provide some meaning beyond the noise--was strikeout rate.  Specifically, he found that when you have 60 PA's, you can use those PA's to explain roughly 50% of the variation in that player's next 60 PA's.  All other hitting statistics are more volatile than that, requiring even more PA's to get a large enough sample to provide that level of prediction.  

Strikeout rate is a funny thing for hitters, because higher strikeout rates are often correlated with increased performance when you compare across players, rather than decreased performance.  Yes, strikeouts are generally bad things.  But players who strike out more tend to be more willing to take pitches they can't handle (this leads to more walks, and sometimes better BABIPs), and tend to swing harder (this leads to more power).  

Nevertheless, given that most everyone has 60 PA's by now, I thought it would be interesting to look at who has seen the biggest shifts in strikeout rate across major league baseball...and how that has correlated to their performance thus far.  I'm only looking at players that have more than Pizza Cutter's requisite 60 PA's, and am comparing their 2014 strikeout rate to their projected strikeout rate (based on an average of their Oliver, Steamer, and ZIPS preseason projections...rarely did those systems show substantial differences among them, but I decided to take the average anyway).

Largest Increases in Strikeout Rate Across MLB


This is the group that people tend to worry about: players who experience a big spike in their strikeout numbers.  And we see a lot of cases like that on this list.  Junior Lake, for example, has been striking out an absurd 42% of the time for the Cubs...and there's no question that he's struggled.  Pablo Sandoval--who I thought was supposed to have a good year based on his offseason weight pictures?--has similarly seen a big spike in his strikeout rates, and has had a horrific start to the season.  

But there are some others on this list who it's hard to get too concerned about.  Mike Trout is second only to the amazing Troy Tulowitzki (my long-time man-crush) at the top of the WAR leaderboards, and there's a sizeable gap between him and the rest of the field.  He's striking out more this season...but honestly, is anyone worried about him?  Justin Upton, similarly, is off to another great start this year.  Sure, his BABIP and HR/FB rates are probably not sustainable, but that's to be expecting when you're wOBAing 0.420.  I don't see anyone sounding alarms about his increased strikeout rate.

Overall, it is the case that more players on this list--and remember, these are the extremes--have underperformed than overperformed, to the tune of about 0.025 wOBA points.  As a group, these players were expected to be slightly above average (0.319 wOBA), but they've been below-average thus far.  So the lesson?  A large spike in strikeout rate can be reason to be concerned about a player, but it's also not a reason to panic.


Largest Decreases in Strikeout Rate Across MLB


If the first list had players we should worry about, this is the group that many of us will want to get excited over.  And at the top of this list, at least, you do see some of the early season's most exciting and surprising performances.  Charlie Blackmon has been a revelation in Colorado, and has gone from solid to brilliant in ability to avoid the strikeout while performing at MVP-caliber levels.  Miguel Montero similarly looks to be rejuvenated this year, and Derek Norris seems to be breaking out.  

Most players on this list, however, have shown more modest changes in performance.  Wilin Rosario, Jeff Baker, and Travis d'Arnaud are struggling despite making more contact.  Players like Pedro Alvarez, Brandon Moss, and Ike Davis--guys who have historically had strikeout "problems" aren't yet showing any increase in their performance as a result of their apparent increase in ability to avoid the strikeout rate.  Overall, players with a dramatically reduced strikeout rate in the season are showing only a slight increase in their overall performance this year (an increase of ~0.014 wOBA units).

Overall Relationship: do changes in strikeout rates lead to changes in performance?

With the caveat that wOBA is not at all stable right now, I did do a quick comparison between the difference in projected and actual strikeout rate versus the difference between projected and actual wOBA.  Here's what I found:


So, it's not a strong effect.  If I was doing this right--on full-season data, for example--it would be a stronger effect, because wOBA would have stabilized better.  But even now, we can see a small, predictive effect: unexpectedly low strikeouts leads to higher wOBA, and vice versa.  The correlation is -0.26.

2014 Cincinnati Reds Hitter Strikeout Rates



For all of the complaints that I keep hearing about the Reds offense, on the whole the team's regulars have hit better than projected, wOBAing 0.342 thus far compared to 0.319 expected.  Most of that is driven by the stupendous performances of that catching corps: Brayan Pena and Devin Mesoraco have been amazing.  I'm also not weighting the averages above by PA's (due to independence concerns), so they are probably (definitely) having too much impact on those totals.

Still, most Reds players are right about where you'd expect them to be based on preseason their projections.  Brandon Phillips pops up as the only Reds player to see more than a 5% increase in his strikeout rate: a subject that C. Trent Rosecrans wrote about recently.  He's actually hitting better since C. Trent's article (see May split) but Phillips has done little this year to suggest that his struggles last year were entirely driven by injury.  I'm hopeful that he can still be productive, but he's probably the hitter that I'm most worried about on the Reds (and I'm not the only one).  My little pseudo-study here of 2014 players, however, at least suggests that a spike in strikeouts is not necessarily indicative of disaster.

On the other side of the coin is Todd Frazier.  While he just hits the + 5% cutoff I arbitrarily set for a "green dot" in my spreadsheet, he's seeing virtually the same change in his strikeout rate as Brandon Phillips...just in the better direction!  I haven't seen Frazier getting a lot of attention for his very nice start thus far, but it's been great to see him anchor down the hot corner.  Between his offense and his excellent defense, it's no surprise that Frazier is at the top of the Reds' WAR leaderboard.  Based on his performance thus far, he's been the Reds' MVP!

Given how unreliable wOBA is right now, I don't think we can conclusively link the changes in the players' performances to their changes in strikeout rates.  But I'm comfortable saying, at the least, that when a player has a sharp increase in his strikeout rate in a month's time (~60-100 PA's), that is reason to be somewhat concerned.  And when a player shows a sharp drop in strikeout rate, there is reason to be optimistic.  

So with that in mind, all signs look good on Todd Frazier.  And, unfortunately, the opposite is true of Brandon Phillips.  

No comments:

Post a Comment