Please welcome Jacob Peterson, our newest contributor and sabergraphic artist at Beyond the Box Score. He will generally be posting bad ass charts and graphs related to statistical awesomeness and other sabermetric weirdness. -ed.
Click the image to enlarge.
While I am never surprised to see Maddux at the top of any modern-era pitching list, particularly one that relates to pitch efficiency, that is simply a gigantic gap between #1 and #2. To put it in perspective, consider this:
I hope your mind is as boggled as mine was when I did that calculation.
A league-average pitcher generally needs around 145 pitches to get 27 outs. That figure was lower before the steroid era, and a little higher in the late 90's; it's been at around 147 in 2011.
During Greg Maddux's career, the league average was 144.1 pitches per 9 innings, yet Maddux himself averaged only 120.6. For reference, no other pitcher with at least 2000 IP since 1988 has averaged fewer than 130 pitches per 9 innings.
Over the course of the 4821.2 IP Maddux tossed from 1988 onward, that equals a total of around 64,600 pitches. The league average pitcher would have thrown around 77,200 in that many innings, so Maddux was nearly 12,600 pitches below average (about 16%). And that doesn't even include any pitches he may have saved in the 186 innings he threw before 1988.
I watched Maddux pitch every 5 days for 11 seasons, and even I was blown away by those numbers.
Another surprise that I encountered when researching this is the high ranking of various marginal pitchers, like Terry Mulholland, Mike Morgan, and Greg Swindell. While these guys were not pitchers of the same caliber as Maddux or Halladay, they did hang around for a long time. Perhaps one of the reasons they were able to have such long careers was that they were so pitch-efficient.
One final note: from 1988 to 1998, the pitch information for a certain number of plate appearances is missing. This usually amounted to only a small proportion of each pitcher's total (Maddux had 320 missing PA out of 19576 total, or 1.6%). To get around this, I simply pro-rated the pitch totals to account for the missing PA.