Minnesota Twins and Run Prevention: Pitching vs. Defense

Jesse Johnson-USA TODAY Sports

Sometime after the 2012 season, I wrote a post about how the Twins could improve their run prevention by roughly 100 runs just by replacing their worst pitchers with some league-average ones. Unfortunately, that post seems to have disappeared into the vapors of the internet, but the general premise was this: 2012's debacle was primarily due to a handful of especially terrible pitching performances (think Jeff Gray, Jason Marquis, Nick Blackburn, Matt Maloney, etc); further, replacing those innings with league-average pitchers would provide a large improvement.

Despite efforts to fix the pitching staff, the 2013 season was another uninspiring effort in run prevention. The front office brought in several new faces to the pitching staff, but sadly many these new pitchers (Vance Worley, Pedro Hernandez, Kyle Gibson, Mike Pelfrey, etc) were decidedly below average, and many of the returning faces were terrible as well. Instead of a 100 run improvement, they were only able to shave off 44 runs, going from 832 to 788. At the same time, baseball as a whole got better at preventing runs in 2013, as the league average for runs allowed for the season dropped by 17 runs between 2012 and 2013. So from a net perspective, the team only improved by 27 runs, and they were still among the worst teams at giving up runs, ranking second-to-last in the American League.

From an ERA standpoint, the pitching staff improved from 4.77 to 4.55. However, when I looked up the ERA numbers on Fangraphs, I noticed that the improvement in Fielding Independent Pitching (FIP) was much better. While the team’s ERA improved by only 0.22 between 2012 and 2013, the team’s FIP improved by 0.43, which translates to roughly 75 runs over the course of the season. Fangraphs fielding-independent pitching numbers indicate a much bigger improvement than the actual change in runs allowed - by roughly 30 runs. It could be that the defense was worse than last year, and that contributed to the large difference. However, it is known that FIP calculations overstate the ability of pitchers with below-average stuff (which would include most Twins pitchers). So who is to blame for the worse-than-expected performance? The defense or the pitchers? Or both?


By and large, the pitching staff as a whole performed as good or better in 2013 than they did in 2012. Overall, they were still terrible, but there was improvement. As I mentioned above, their ERA and FIP both dropped, and overall they allowed fewer runs. Also, their underlying rate stats were fairly stable between the two seasons:

Stat (MLB rank, high-to-low)



Strikeout %

15.2% (30th)

15.7% (30th)

Walk %

7.5% (22nd)

7.3% (25th)

Ground Ball %

45.7% (14th)

43.7% (21st)

Line Drive %

21.1% (9th)

21.2% (13th)

Fly Ball %

33.2% (20th)

35.1% (11th)

Infield Fly Ball %

8.9% (25th)

9.9% (14th)

Home Run per Fly Ball %

12.7% (5th)

10.2% (16th)

Overall, only three numbers had major changes between past two seasons: Ground Ball, Fly Ball and Home Run rates. The pitching staff ended up giving up more fly balls, which could be bad since fly balls turn into home runs and extra-base hits. But a good portion of the additional fly balls stayed in the infield (see the increase in Infield Fly Ball %). Since infield pop-ups are almost guaranteed outs, the overall increase in fly balls probably didn't hurt the staff too much. Additionally, even with the increase in fly balls, they cut back fairly significantly in their home runs allowed (30 fewer homers than 2012), and that showed up in the lower home run rate. So overall, I don’t see any strong evidence that 2013 pitching staff was worse than the 2012 pitching staff, and if anything it was slightly better.


There are three major defensive measures that attempt to assign a run values to defensive performance. On Baseball Reference one can find Total Zone (Rtot) and Defensive Runs Saved (Rdrs), and Fangraphs has Ultimate Zone Rating (UZR). All three metrics estimate how many runs the team or player is above or below average. Advanced defensive metrics are an inexact science, and they shouldn’t be used to draw too strong of conclusions. Looking at the Twins defensive values between 2012 and 2013, I think it is fair to say that the conclusion is rather stark.

Stat (MLB rank)




Defensive Runs Saved

30 (6th)

-8 (20th)


Total Zone

-7 (16th)

-42 (24th)


Ultimate Zone Rating

1 (15th)

-43 (26th)






The Twins went from being an average to above-average defensive team in 2012 to a below average team in 2013, and the impact of that change was approximately 40 runs. This matches quite well with the "missing" 30 runs of improvement expected from FIP. Who is responsible for the deterioration in defense? Well, anyone who watched a few games this season probably could come up with a good guess. Here are the same defensive values, this time just looking at the outfield:

Stat (MLB rank)




Defensive Runs Saved

21 (4th)

-35 (28th)


Total Zone

29 (3rd)

-35 (28th)


Ultimate Zone Rating

20 (4th)

-35 (28th)






The Twins outfield defense went from being one of the best in the league in 2012 to one of the worst, with a net change of approximately 60 runs. Now I think everyone expected some drop-off in defense when switching from Ben Revere and Denard Span to Aaron Hicks and Chris Parmelee; however, I don’t think anyone predicted exactly how terrible the change would be (or how many innings would be given to non-outfielders like Ryan Doumit and Chris Herrmann).

In order to drive home the point that the defense was terrible, here are some more stats to take in:

  • Fangraphs keeps track of Balls In Zone (BIZ) for each position, the number of plays made on those balls, and the ratio between the two values. Looking specifically at the outfield, the Twins were 2nd in MLB in 2012 at making plays on BIZ. In 2013, they dropped to 28th.

  • Looking at the different between ERA and FIP, the Twins were roughly league average in 2012 (17th), but had the 25th largest difference in 2013.

  • Bill James developed a Defensive Efficiency metric which attempts to measure how effective a team is at turning batted-balls into outs. In 2012, the Twins ranked 18th; in 2013, they dropped to 29th.

  • League average BABIP is .295. In 2012, the Twins pitching BABIP was .294. In 2013, it was .309, which was the second highest value in the majors.


At the conclusion of the 2013 season, I assumed that the Twins pitching staff took a small step forward from the disaster that was 2012, primarily from the bullpen. Definitely not enough improvement to approach respectability, but at least it was moving in the right direction. However, after digging into the defensive numbers, I think that there is a good argument that the pitching staff took two or three small steps forward, but the defense took a big step back. When trying to improve the Twins this offseason, it is important that both areas are improved. To put the -39 runs into perspective, that is roughly the difference over the course of the season between Scott Diamond and Shelby Miller.

When analyzing run prevention - either for a team or individual pitchers - it is always important to remember to look at both sides of the run prevention coin. Unfortunately, it is really, really difficult to properly untangle the credit or blame between the pitchers and defenders. The defensive numbers I’ve shown here are definitely not 100% precise; however, they do paint a consistent picture. One can’t say with 100% confidence that the Twins defense was EXACTLY 39 runs worse than last season, but it does provide a good rough estimate for determining the impact of the defense, and that impact was undoubtedly negative. Most importantly, these numbers match the eye test. Anyone who watched a handful of Twins games this past season saw Willingham limp around on one good leg, Parmelee not get to a ball down the line, Arcia misplay a routine flyball, or Plouffe lose track of the number of outs. Defense matters, and if the Twins want to return to respectability, they need to improve in that area.