Saying Meaningful Things (About How Defense Affects Pitching)

What bothers me the most about the coverage on nationals.com is Ladson's penchant for the meaningless.  I understand that a bit of superficiality is necessary when generating content for mass consumption.  I also understand that disguising your opinions in bland platitudes and unsupported assertions doesn't constitute analysis.  So Adam LaRoche is better defensively than Adam Dunn.  OK.  What does that mean, though?  Nationals.com is a wasteland of relativity: 'better', 'worse', 'more', and 'less.'

The preceding rant was triggered by months of claims that better defense will result in better starting pitching.  Sounds logical.  Unfortunately, it also sounds meaningless.  How much better will the Nationals defense be this year?  What are the ways in which an upgraded defense creates better starting pitching?  Have other teams experienced these effects?  If so, what level of improvement can we expect (measured in ERA+, wins, etc.)?

The great thing about baseball analysis in 2011 is that the answers are out there.

As long as 'out there' doesn't mean 'nationals.com'

It turns out Vince Gennaro has an answer to a very closely related question.  The link - containing his PowerPoint presentation and an audio file of his presentation at SABR 39 - can be found here.

Poor defense requires starting pitchers to record additional outs.  Given these extra outs, the opposition will score additional runs. A secondary problem is that those bonus outs also shorten the starting pitcher's outing.  This produces even more runs because if middle relievers (who absorb a good portion of the additional innings) were as good as the starting pitchers they are replacing, they wouldn't be middle relievers (sorry Chad Gaudin).

You can argue that extending outings through better defense still doesn't count as an improvement to the starting pitching.  It's certainly a plus if starters pitch deeper into games, but they are not actually performing any differently.  My sneaking suspicion is that Ladson had the psychological benefits of a better defense in mind.  This would gel with his overall preference for unquantifiable crap.  This psychological benefit may exist, but I don't understand why it is necessarily stronger than the benefit of having an offense capable of providing early leads or knowing your team has the ability to come back from deficits.  For example, John Lannan is a huge wuss.  He's going to pitch scared regardless of whether he doesn't trust his defense or he thinks he can't give up a run without losing the game.  He's going to sulk regardless of whether someone makes an error behind him or he's getting no run support.  It doesn't matter.

Jordan Zimmermann is not a wuss.

Returning to things that do matter, Gennaro's conclusions are that:

• Top quartile defensive teams get their pitcher through 0.5 innings more per game than bottom quartile teams
• Those 81 innings fall largely to the dregs of the bullpen
• The range of this impact (from the best defensive team to the worst defensive team) is somewhere in the neighborhood of 3-4 wins
• The specific impact on a given team depends on a few items:
• The quality differential in your bullpen - if every reliever is equally effective, then this doesn't matter
• What league you're in - the impact is muted in the NL where pitchers are removed prematurely for pinch hitters (i.e., if an error caused 15 extra pitches but the starter leaves the game early for a pinch hitter, the bullpen did not have to do extra work because of the error)
• Your defensive ranking - exiting the bottom quartile is valuable; entering the top quartile is valuable; shifting your ranking somewhere in the middle is not so valuable

It's possible to calculate how many fewer runs the Nationals would have given up last season with a league-average defensive efficiency rating (DER) enabling more efficient bullpen utilization.  If I was a paid sportswriter for the Nationals website, I would gladly do it.  Maybe in a future post.