The Immaculate Regression
August 25, 2009 11:11 AM
Okay, cheesy title, I admit. But David Archer’s blog post on “immaculate innings” (when a pitcher strikes out the side on nine pitches) got me to thinking about some of the conclusions drawn.
Archer offers a graph to suggest that the frequency of immaculate innings occurrences has increased over the decades. While it is true that the total number of immaculate innings per year has increased, the number of innings pitched has, too, increased, as has the number of strikeouts per innings pitched. (Archer does note that “a lot more teams play today, meaning more innings pitched and more opportunities for immaculate innings.”)
As one of the commenters on the blog noted, perhaps it would be more accurate to “normalize” per total innings pitched:
So, perhaps an increase indeed in frequency…but is it a statistically substantial one? I truly do not know. Maybe some of you more credible statisticians can weigh in.
The blog post also (sort of) offers one possible reason as to the increased frequency of immaculate innings”:
As one friend pointed out, the best explanation for the increase in recent decades appears to be the advent of the modern reliever, especially the flame-throwing, one inning closer (more immaculate innings have been thrown in the 9th inning (eight) than in any other inning)…
I am not sure about this assertion, either, given the following graph:
Since 1950, 18 immaculate innings have occurred after the fifth inning, 15 before the fifth inning, and one in the fifth inning. This does not seem to be a huge difference between early- and late-game occurrences.
As far as starter vs closer, I made a feeble attempt at categorizing the pitchers who have tossed immaculate innings since 1950:
| Year | Inning | Pitcher | Primary Role |
| 1959 | 9 | Jim Bunning | starter |
| 1964 | 8 | Bob Bruce | starter |
| 1977 | 9 | Bruce Sutter | closer |
| 1984 | 9 | Ron Guidry | starter |
| 1987 | 8 | Jeff Robinson | middle reliever |
| 1989 | 8 | Rob Dibble | closer |
| 1990 | 8 | Jeff Montgomery | closer |
| 1992 | 9 | Trevor Wilson | starter |
| 1994 | 9 | Mel Rojas | closer |
| 1997 | 9 | Mike Magnante | middle reliever |
| 2002 | 9 | Jason Isringhausen | closer |
| 2002 | 8 | Byung-Hyun Kim | closer |
| 2004 | 8 | Brandon Backe | starter |
| 2004 | 9 | LaTroy Hawkins | closer |
I concede that the primary pitching role may not be entirely accurate, but it seems like only about half of the late-inning immaculate innings were thrown by true closers.
My uninformed take on all of this is that any increase in the frequency of immaculate innings is most likely a factor of the increased frequency of strikeouts in general. As to why that is the case, please discuss!
Still, the idea of an immaculate inning is fun, and I appreciate David Archer bringing up the topic.
Disclaimer: I know nothing about statistics and significance.
Well, duh.
More strikeouts, more immaculate innings. It would surprise me only if the number if immaculate innings went down.
People don’t care about strikeouts like they used to. Albert Pujols strikes out relatively infrequently for a power hitter, but he still has not had one season where his strikeout total was less than the number of home runs hit. Contrast that with Joe Dimaggio, who in 13 seasons in the majors had 7 years of more homers than K’s. In two of Yogi Berra’s three MVP seasons, he accomplished that feat.
Bobby Bonds held the season strikeout record of 189 from 1970 until 2006. Since 2006 that number has been eclipsed 6 times. Mark Reynolds, the current record holder at 204 should shatter the mark he set last year – he’s projected to have 222. But he has 38 homers and is on a pace to hit 50. Ryan Howard should keep pace with his last two years when he struck out 199 times. He should also have 45 homers.
This is what people want. Remember, Chicks dig the long ball.
Ruthian homers. Ruthian strikeouts and more of ‘em.
Comment by Tommy T — Tuesday, August 25, 2009 @ 1:30 PM