As Bert Blyleven is often fond of saying, the double play is a pitcher's best friend. Even with a somewhat improved offense, it looks like the Twins' pitchers are going to be needing all the friends they can get this summer.
Double Play Stats
Double play statistics, as commonly reported, are very difficult to interpret. For instance, in 2003, the Detroit Tigers were leading the league in double plays turned on defense while at the same time fielding a historically bad team. In this case, it was pretty easy to see that it wasn't superb defense and pitching that was getting them so many double plays, it was bad pitching that put a ton of runners on first base.
Fortunately for us, there are people who keep track of these things more closely these days. At Baseball Prospectus, they keep track of double plays for pitchers in this handy report. Before I go through and explain what everything means, I'll give you the last year's data for Twins pitchers who pitched in the majors and are going to make the staff this year:
NAME DP_OPPS DP DP% NETDP
Carlos Silva 132 42 31.8% 24.57
Jesse Crain 80 15 18.8% 4.44
Kyle Lohse 159 24 15.1% 3.00
Francisco Liriano 14 2 14.3% 0.15
Scott Baker 37 5 13.5% 0.11
Brad Radke 142 18 12.7% -0.75
Matt Guerrier 53 5 9.4% -2.00
Joe Nathan 41 2 4.9% -3.41
Juan Rincon 71 5 7.0% -4.38
Johan Santana 122 11 9.0% -5.11
DP_OPPS is the number of opportunities each pitcher had to induce a double play. This would count every plate appearance where there was a runner on first base and less than two outs, or a runners on first and second with less than two outs, etc. DP is the number of double plays induced by the pitcher. DP% is simply the percentage of double plays the pitcher induced per opportunity.
NETDP is a little trickier, but not difficult to understand. A league average pitcher gets hitters to ground into double plays about 13.2% of the time. So if we take Jesse Crain for example, we would expect that given 80 double play opportunities, a league average pitcher would induce 10.56 double plays. Crain induced 15 double plays, so he was 4.44 double plays better than average, or 4.44 NETDP.
So, glancing at the table, we see that Silva was by far and away the best pitcher on the team at inducing double plays. In fact, he was by far and away the best pitcher in baseball at inducing double plays. The next best pitcher in baseball was Brandon Webb with a pretty healthy total of 15.43 NETDP, but still about 9 double plays less than Silva.
How much is a double play worth?
To answer this question, let's take a quick look at some expected runs in some double-play situations, and situations after double plays.
Runners Outs Expected Runs
1-- 0 0.8968
--- 2 0.1075
Runs saved by DP = 0.7893
1-- 1 0.5487
--- 3 0.0000
Runs saved by DP = 0.5487
12- 0 1.4693
-2- 2 0.3502
Runs saved by DP = 1.1191
12- 1 0.9143
--- 3 0.0000
Runs saved by DP = 0.9143
1-3 0 1.8228
--- 2 0.1075 (+1 run scored)
Runs saved by DP = 0.7153
1-3 1 1.1830
--- 3 0.0000
Runs saved by DP = 1.1830
123 0 2.3109
--3 2 0.3718 (+1 run scored)
Runs saved by DP = 0.9391
So, in these instances, you're looking at the double play saving about 0.75 to 1.25 runs, or so 1 run on average. In the case of Carlos Silva, his exceptional ability to induce double plays saved him about 24 runs over an average pitcher last season.
How good would Silva have been without the DPs?
Last year, Silva allowed 72 earned runs in 188.7 innings pitched for an ERA of 3.44.
Had he been league average at inducing double plays, he would have allowed about 96 earned runs. Now, not allowing those double plays makes it much tougher to get out of those innings, and he might not have been able to pitch as many innings. But let's just say he still would have pitched 188.7 innings. That would give him an ERA of about 4.57.
So, all told, those double plays shaved practically a whole run off of Silva's ERA.
Whether or not Silva can repeat this performance in double play situations will be a big key to how valuable he is to the Twins this season.
EDIT: One double play scenario updated as noted in the comments.