After picking off Anthony Gose on Saturday, Ryan Feierabend has 11 career pickoffs in a mere 107.1 innings. That's good for a rate of only 9.8 innings per pickoff, the best among any pitcher for whom we have data with at least 100 career innings pitched. Here at BtBS, we specialize in minutiae, so what follows is an analysis of some of the best in the business at catching runners leaning.
I started by looking at the inning/pickoff numbers for all post-1954 pitchers (the sac-fly and IBB era) that threw at least 100 career innings. Greg Smith comes in second at 14.3 IP/PO, a difference of 3.6 from first place. After Feierabend, the next active players are Aaron Loup at 16.2 IP/PO and Joe Beimel at 18.7. However, most of the leaders all threw very few innings, and the first pitcher on the list to break 1,000 innings is Darold Knowles, way back at 24.3 IP/PO.
I wanted to know if the smaller sample sizes were reliable or not, so I did a split-half analysis similar to the one Russell Carleton (Pizza Cutter) did a while ago on pickoffs, splitting them by whether they were indexed as even or odd in my Retrosheet database. Here's a table of the correlation coefficients between the two groups for all data up to 2013, starting in each of the last few years:
|Start year||Total pickoffs||r|
Only when 2008 is included – meaning that we're examining six full years of data – do pickoff rates stabilize past the 0.70 correlation limit that Pizza Cutter set out. With around 185,000 batters faced in the majors each year, and just less than a third reaching base (and discounting home runs), which gives us over 300,000 total batters faced. The average pitcher last year faced 272 batters, so that means we're looking at about 1100 batters faced before pickoff rates are reliable. At last year's rate of 4.2 batters per inning, that means 260 innings pitched is a much better indicator of a pitcher's pickoff rate.
If we use that as our cutoff, Beimel comes out on top, just ahead of Jim Umbarger and Jerry Garvin, all of whom were regular relievers at some point in their careers. In fact, the first player on the list who started at least 70% of his career games is Jeff Locke, all the way down in eighth. There are some recent starters who rank relatively high, though: Dallas Braden and Clayton Kershaw are 19th and 20th, Julio Teheran is 32nd, and Mark Buehrle is 34th. Kershaw's ranking is most impressive, though, because he allows so few baserunners. If you don't let anyone on, you can't pick them off, meaning that Braden's 1.325 career WHIP gave him many more opportunities to nab runners than Kershaw's 1.072, Teheran's 1.135, or Buehrle's 1.279.
Opportunities/pickoff is probably a more interesting number, but is also much more difficult to calculate, since you can't pickoff more than one runner at once, and some situations (runner on first, nobody out) are more likely to induce pickoffs than others (bases loaded, 1 out). For now, innings will have to do as a reasonable proxy for opportunities, although it wouldn't be too difficult to fold WHIP into the equation somehow as well.
Unsurprisingly, the list is heavy on lefties: Bob Stoddard is the only non-southpaw to crack the top 10, and only 3 of the top 21 (14%) are right-handed. Beimel holds the single-season record for pitchers with at least 50 innings. In 2003, the then-26-year-old Pirate threw 62.1 innings, picked off seven runners, had five caught stealing, and didn't allow any to successfully steal, in one of the best baserunner-control seasons ever. Making it even more impressive is that he didn't benefit from a great arm behind the plate: Jason Kendall, the Bucs' primary catcher that year, was known more for his hitting and threw out only 27% of runners, below the league average of 31%.
It's worth noting that pitchers known for their moves, such as Andy Pettitte and Mike Marshall, don't rank particularly high. Moves like Koji Uehara's to end Game 4 of the 2013 World Series seem to be more fluke than skill. While pickoffs remain an interesting, if underappreciated, part of the game, they're still a valuable weapon for guys like Feierabend.
. . .
Steven Silverman is a contributor at Beyond the Box Score and a student at Carnegie Mellon University. You can follow him on Twitter at @Silver_Stats.