Saving the King’s Arm

There was a theory born in the comments section of my last post suggesting that better run support for Felix Hernandez in the upcoming years could lead to reduced innings. That would be a good thing, so I checked out data from the 66 games in which both he has started and Coach Wedgie has been in possession of the hook.

It was first interesting to discover that Felix is actually a part of quite a few friendly blowouts. In his 66 starts since 2011, the M’s won fifteen of those games by four runs or more. That’s about 23 percent of his total starts, and fits between Justin Verlander’s 21 percent and C.C. Sabathia’s 36 percent in 2012. In other words, it’s not like the M’s haven’t given Wedgie some excuses to rest the horse.

If we plot runs scored by the Mariners to Felix’s innings pitched, here’s what we get.

 Felix run support plot

I don’t see a strong relationship there, but the upper right corner is definitely promising. Of the six starts* in which Felix received at least eight runs of support, he never went longer than seven innings. However, it only saved him three pitches per start. Small sample sizes and all, that is not a lot of evidence to suggest that higher run support will lower his innings and pitch counts.

*There are two data points on one of those 9’s, as well as some other points.

Finally, I put together a comprehensive model including components to indicate how well Felix was doing in each game and how close the game was. The model essentially suggested that increasing run support from his usual 3.6 runs per game to even 4.5 runs per game might only reduce Felix’s average innings by one or two tenths of an inning. Not even one third of an inning. When I ran the model to try to predict pitches thrown, that same change in run support only led to about one less pitch per start.

The number one thing that can reduce the King’s innings and pitch counts is smart decision-making by Mr. Eric Wedge. He has been afforded opportunities to cut back on pitch counts in the past, and has seemingly decided against it. It would be wise to save Felix some added stress during blowouts, but I’m not so sure that a bump in run scoring will make much of a difference. I think it’s mostly in the hands of the man with the mustache.

Followup: Run support here came from the final score of the game. There is definitely an argument to use the run support and run deficit at the time Felix exited each game. That data is harder to put together, but could explain more about whether or not we can expect reduced stress on Felix’s arm.