The numbers are ugly, and it doesn't matter if we're talking about ERA (4.95, worst in MLB by nearly a quarter run) or fielding independent metrics (4.57 FIP, 4.43 xFIP, both worst in MLB). A team simply isn't going to win many games with this kind of performance on the mound, even with an above average offense. Which it goes without saying we have not seen this season. But what's behind this huge drop-off in performance, especially given pretty much the same pitchers in the rotation? In order to help answer this question, I have collected copious amounts of Pitch F/X data to attempt to determine whether there has been a change in approach, skill, or if a regression to the mean may be in order. It's going to take a while to go through all of this 2010 and 2011 data, but after the jump I'll let you know what I've found.
1. Throwing StrikesThis one has been painfully obvious. In 2010, the Twins pitching staff was best in MLB at avoiding the free pass, allowing only 2.37 walks per nine innings. This was a full 0.2 walks below the second best team (Philadelphia), and almost a half a walk per game below the second best in the AL, Seattle with 2.83 BB/9. Since the Twins don't tend to strike out many hitters (6.49 strikeouts per nine innings was 26th in MLB last year), limiting walks is a key to success.
This season, the walks have come early and often, as the Twins are second worst in MLB, giving up 3.77 walks per nine innings. It's not as bad as the Chicago Cubs MLB-worst 4.08 walks, but it's a huge increase from last season. Add in a reduction in strikeout rate (down to 5.86 per nine innings), and you get a recipe for disaster. In fact, simply plugging the difference in strikeout and walk rates into the standard calculation of fielding independent pitching (FIP), we get an increase of 0.61 runs per nine innings to the Twins FIP, just about the entire difference between last year's 3.91 FIP (9th in MLB) and this year's 4.57 FIP (worst in MLB).
So we know the Twins pitchers have been walking a lot more and striking out fewer batters this season. As I noted above, I am analyzing Pitch F/X data from the 2010 and 2011 seasons to look at the Twins strike throwing in more detail, including pitch type, location within (and outside) the strike zone, and batted balls given a pitch location. In order to examine pitch location, I split up the strike zone into a three-by-three grid, similar to what one sees in batter's heat maps showing where he likes pitches, etc. As far as throwing strikes is concerned, the Twins haven't seen much of a drop-off since last season. Using the raw Pitch F/X data from MLB's Gameday site, I have calculated the number of strikes, similar to ZONE% one might find on Fangraphs. I can't speak in detail to how ZONE% is calculated, but I suspect the minor differences between my numbers and ZONE% may be due to my accounting for the width of the baseball, if any part of the baseball touches the strike zone I consider the pitch to be a strike.
Year | Strike Zone % | MLB Rank |
2010 | 50.41% | 2 |
2011 | 50.19% | 7 |
As the table above shows, the Twins pitchers have thrown slightly fewer pitches within the strike zone, but not to the point where one would expect an MLB-worst walk rate. The Twins have fallen five places to #7 in MLB at throwing strikes. This is due to pitchers across MLB throwing more strikes (about 1% more) this season, up to an average of 48.47% this season compared to 47.44% in 2010.
2. Umpires Calling Strikes
If walks have increased given a similar strike zone percentage, have umpires been a problem, not calling strikes?
Year | Swing% (in/out of zone) | Called Strike% (in/out) | Swing-Whiff% (in/out) |
2010 | 63.6% / 34.6% | 82.2% / 9.5% | 51.1% / 71.6% |
2011 | 65.1% / 30.7% | 81.6% / 6.1% | 53.0% / 73.3% |
The table above shows that there has been a 0.6% drop in called strikes for watched pitches within the strike zone. Over a total of 1,166 watched pitches, this corresponds to 7 fewer called strikes compared to the 2010 percentage. For watched balls outside the strike zone, Twins pitchers have seen a sharper 3.4% drop in called strikes, corresponding to 79 fewer called strikes compared to 2010. Over a total of 45 games (through Sunday), this is roughly two fewer called strikes than we would have seen last season. Over 6,653 pitches, these 86 called balls contributes to a 1.3% reduction in the overall strike percentage. Enough to be felt, but probably not enough to increase the walk rate to the point we've seen this year.
3. Hitters Aren't Chasing Pitches
The table above shows that batters are chasing far fewer pitches this season than in 2010, Last season, the Twins pitchers got hitters to chase balls at a 34.6% rate, third highest in MLB. This season, hitters have chased only 30.7% of pitches outside the zone, 28th best in MLB. Whatever effect the umpires have had this season, it is more than canceled by hitters laying off balls outside the zone, contributing to the walk rate.
4. Bullpen
As bad as the entire pitching staff has been, the bullpen has been worse. I won't repeat what Seth posted today to TwinsCentric, but the starting pitchers have a 4.72 ERA, fourth worst in MLB, while the bullpen has a 5.41 ERA, second worst in MLB to Detroit's 5.84 bullpen ERA. When your highest leverage situations are being pitched by your poorest pitchers, it has a large impact on winning ball games. From a win probability added (WPA) standpoint, the rotation has contributed a total of -0.84 WPA compared to the bullpen's -1.40 WPA. Of course, both pale in comparison to the lineup's overall -5.77 WPA, but I'm focusing on the pitchers today. However, I should point out that (last night's game notwithstanding) the bullpen has blown only seven saves this season (including setup guys blowing leads), right around league average. Of course, when one compares to only 10 saves and few opportunities this still isn't very good.
5. Defense
I'm going to try not to get on a rant here, but defense has to play a role in the Twins poor performance on the mound this season. Overall, looking at UZR/150 metrics, the Twins defense has slipped but is still above average, down to +1.3 from +3.2 UZR/150 last season. This corresponds to about a 20 run difference over an entire season, so a drop-off but not off the table. However, one needs to look no farther than the middle infield to think defense may be getting into pitchers' heads. Last season, with Orlando Hudson and J.J. Hardy, middle infield defense was a key strength. At shortstop, the Twins overall +13.2 UZR/150 was best in all of baseball, and at second base the Twins overall +9.9 UZR/150 was fourth in MLB. This season, the Twins overall shortstops -21.1 UZR/150 is easily worst in baseball, and second base's -8.8 UZR/150 is 23rd in MLB. And it's not just a matter of range, as errors have contributed to the poor defensive metrics. With this kind of a sieve behind me, I'd hesitate to throw strikes as well...
I could see the Twins start to reverse the walk rates a bit, if Francisco Liriano rediscovers his mojo, and if the Twins can patch together three or four bullpen arms who don't stink on ice. But at this point, with the number of holes in the staff as well as defensively, I don't see run prevention becoming a strength for the Twins any time soon.