Originally published at The Tenth Inning Stretch

----

"Clutch" is a word that has been used to describe the players who go above and beyond the norm and perform highly in difficult situations in close games: The guy who comes into the game with 2 outs and the bases loaded, down three in the bottom of the ninth and hits a grand slam off a closer who never blew a save. The pitcher who comes in with bases loaded and no outs, up one on the bottom of the ninth and strikes out the side to save the game. Whether or not there are players out there who are "clutch" and consistently outperform themselves in those type of situations has been a vast subject of debate. Here is a good summary of a few historic studies and here is a list of lots of links on the subject, if you feel like reading more.

About 5 years ago, David Appelman devised a measurement called, well, "Clutch" that tries to describe clutch hitting and pitching. Basically what it does, is it looks at the difference of someone's performance (based on Win Probability Added) in High Leverage situations only, versus his overall performance. Players with positive Clutch outperform their overall performance in those situations and players with negative Clutch underperform. This does not mean that a player with a higher Clutch is a "better" player than one with lower Clutch. If someone hits .220 with the bases loaded and .200 overall, is not a better player than someone who hits .320 with the bases loaded and .340 overall, despite the Clutch numbers.

Fangraphs picked the stat now and I sorted the Twins hitters, relief pitchers and starting pitchers based on Clutch. This is an exercise just for fun. I don't think that it means that much (see example above), but it is interesting and it potentially be an additional tool for player evaluation, but not a stand alone tool. Situational and High Leverage performance definitely might tell something about relievers as a group, but we are not about to jump into any conclusions here. Just a demo

Here are the 2013 Twins' hitters ranked by Clutch (60 PA min) :

Here are the 2013 Twins' relief pitchers ranked by Clutch (10 IP min):

Here are the 2013 Twins' starting pitchers ranked by Clutch (35 IP min):

Pitchers in both roles are listed on both tables for that role only. A couple of observations:

• The Twins definitely do not use or value the metric, since about half of the players with positive Clutch numbers are gone
  
• I am not surprised to see Dozier and Duensing lead the hitter and the reliever groups respectively. It is my recollection that they really excelled in pressure situations
  
• Worley leading the starter group may be a surprise, until someone considers that, in order to pitch in a high leverage situation, a starter needs to get himself into trouble to make that situation high leverage. Thus, Clutch, is not very useful at all for starters (other than looking at fun results like the above...)