Under The Knife: Trage... (07/24)
0 comments have been left for this article. BP Comment Quick Links
| |
|
|
July 24, 2007 Prospectus ToolboxNon-contact, Part OneStatheads and strikeouts…it's an age-old romance. For pitchers, we'll tell you that strikeouts are the biggest predictor of a hurler's future success. When batters go down, though, while we acknowledge that the whiff is an out--a negative result--it's an out we put on a pedestal as one of the Three True Outcomes (along with the walk and the homer). A noble out, I guess. Where does this love for the breezy out come from? Part of it is certainty. One big reason that dissecting baseball data is so rewarding is the peculiar one-on-one matchup between pitcher and batter, a rarity in team sports. The strikeout is the pitcher's ultimate victory in that confrontation, leaving the batter grumbling as he walks back to his dugout, rather than running the bases. It's conclusive. Even better, it's simple. Remember all that trouble we have measuring defense? It's largely academic when the only fielder involved has a catcher's mitt. The issue of whether or not a pitcher is able to control the destiny of balls in play is irrelevant if you keep the batter from making contact with the ball. Now, glovework does still come into play on the strikeout, and it's possible for the batter to reach base even on a humiliating whiff. However, it's exceedingly rare--over the past 47-plus seasons, the batter reached base safely on 0.3 percent of all strikeouts. By way of comparison, this season a batter who hits a ball in play has a more than 30 percent chance of reaching base safely. Strikeout rate is one of the performance metrics that shows the most consistency from one season to the next. Correlation studies, such as this one by James Click, show that once a pitcher shows a tendency to record strikeouts, they're likely to maintain that level--much more likely than they are to maintain their ERA or batting average on balls in play allowed. That's why power pitchers are very generally considered better bets to repeat their performance than pitchers who induce a lot of contact--their strikeouts can be considered a reliable source of high-percentage outs, which they're expected to carry over from year to year. The stat that we most frequently look at when talking about pitcher strikeouts is strikeouts per nine innings (SO/9). This measure is a little quaint--we're pretty close to the point where complete games are so rare that the number of strikeouts a pitcher would record in nine innings is an abstract concept--but it's the standard. Let's take a quick look at the best seasons since 1959, by strikeout rate (min. 162 IP):
Year Name Age IP SO SO/9 SO Rate 2001 Randy Johnson 37 249.2 372 13.41 .3743 1999 Pedro Martinez 27 213.1 313 13.20 .3749 1998 Kerry Wood 21 166.2 233 12.58 .3333 2000 Randy Johnson 36 248.2 347 12.56 .3467 1995 Randy Johnson 31 214.1 294 12.35 .3395 1997 Randy Johnson 33 213 291 12.30 .3424 1998 Randy Johnson 34 244.1 329 12.12 .3245 1999 Randy Johnson 35 271.2 364 12.06 .3374 2000 Pedro Martinez 28 217 284 11.78 .3476 2002 Randy Johnson 38 260 334 11.56 .3227 1987 Nolan Ryan 40 211.2 270 11.48 .3093 1984 Dwight Gooden 19 218 276 11.39 .3140 1997 Pedro Martinez 25 241.1 305 11.37 .3221 2003 Kerry Wood 26 211 266 11.35 .2999 1989 Nolan Ryan 42 239.1 301 11.32 .3047 1997 Curt Schilling 30 254.1 319 11.29 .3162 2001 Kerry Wood 24 174.1 217 11.20 .2932 1995 Hideo Nomo 26 191.1 236 11.10 .3026 2002 Curt Schilling 35 259.1 316 10.97 .3107 2004 Oliver Perez 22 196 239 10.97 .2969 1993 Randy Johnson 29 255.1 308 10.86 .2953 This list is sorted by SO/9, but I've also given you a superior stat, strikeout rate (SO Rate), which does more or less the same thing. SO Rate is strikeouts per batter faced, a measure that gives you a better idea of a pitcher's strikeout dominance than strikeouts per inning does. To illustrate this point, we can look at two players on the top 20 above, Kerry Wood and Curt Schilling. By SO/9, Wood's 2003 was better than Schilling's 1997, and Wood's 2001 was better than Schilling's 2002. Schilling's SO Rate in each of those two seasons (.3162 and .3107) was better than Wood's in either of his (.2999 and .2932), because Schilling was facing far fewer batters per inning than Wood was. The other thing you've likely noticed is that the SO/9 top 20 is dominated by pitching performances from the last 12 years. The highest-ranking pitcher season from the pitching-dominated sixties, Sudden Sam McDowell's 1965 (10.71 SO/9), comes in at number 22 in this ranking. Nolan Ryan's 1973 (10.57 SO/9) is the first representative from the '70s, at number 25. Some of that is a convergence of great strikeout pitchers--Randy Johnson, Pedro Martinez, Wood, Schilling--but another element is that the period since the 1994 strike has seen a historic increase in strikeout levels in the majors.
Years MLB SO Rate
1960-1969 .1509
1970-1979 .1348
1980-1989 .1406
1990-1999 .1592
2000-2007 .1675
One way we can adjust for the change in strikeout rates across the eras is to normalize these values to a single standard. This is exactly what the Davenport translations do, presenting normalized strikeout rates (EqSO/9) on the DT Player cards. This way, we can look at the best strikeout seasons of all time, on an even playing field:
Name Year Team EqIP EqSO EqERA EqSO/9 Pedro Martinez 1999 BOS-A 259.2 382 2.15 13.2 Rube Waddell 1900 PIT-N 173.1 252 3.58 13.1 Rube Waddell 1902 PHI-A 220.2 316 2.49 12.9 Dazzy Vance 1926 BRO-N 168.1 237 3.90 12.7 Johnny Vander Meer 1941 CIN-N 225 310 3.36 12.4 Sandy Koufax 1960 LA-N 175 240 4.22 12.3 Rube Waddell 1903 PHI-A 270.1 370 3.43 12.3 Randy Johnson 1998 SEA-A 184.1 249 4.20 12.2 Randy Johnson 1992 SEA-A 224.2 302 4.17 12.1 Randy Johnson 1995 SEA-A 273.2 368 2.33 12.1 Randy Johnson 1997 SEA-A 235.2 316 2.29 12.1 Randy Johnson 2001 ARI-N 284.2 383 2.59 12.1 Dazzy Vance 1925 BRO-N 258 347 3.24 12.1 Pedro Martinez 2000 BOS-A 261 349 1.76 12.0 Rube Waddell 1904 PHI-A 264.2 351 2.92 11.9 Dwight Gooden 1984 NY-N 240.2 316 3.44 11.8 Sam Jones 1956 CHI-N 182 238 4.05 11.8 Sandy Koufax 1961 LA-N 270 350 3.63 11.7 Dazzy Vance 1924 BRO-N 299.2 391 2.58 11.7 Rube Waddell 1907 PHI-A 216.2 282 4.53 11.7 Lefty Grove 1926 PHI-A 255.2 329 2.99 11.6 Van Lingle Mungo 1936 BRO-N 289.2 374 3.42 11.6 Johnny Vander Meer 1942 CIN-N 242.2 312 3.26 11.6 Randy Johnson 1993 SEA-A 274.1 352 3.05 11.5 Cy Seymour 1898 NY-N 258.1 330 4.01 11.5 Dizzy Dean 1933 STL-N 274.1 347 3.74 11.4 Randy Johnson 2000 ARI-N 278 352 2.72 11.4 Nolan Ryan 1989 TEX-A 258.2 328 3.55 11.4 Kerry Wood 1998 CHI-N 184 234 3.33 11.4 Nolan Ryan 1978 CAL-A 223 279 4.16 11.3 Using the normalized figures, we see the addition of a number of seasons that were outside of the previous search, but we also see a good bit of movement within the previous list's time frame. Most notably, Sandy Koufax's 1960 season pole-vaults past Randy Johnson's cluster of dominance to fall into sixth place (based on raw stats, it was 45th); Koufax's 1961 and Nolan Ryan's 1978 both make it into the top 20. Beyond enabling us to take a better look at individual pitchers' seasons across eras, EqSO/9 is also a fine tool for comparing team's strikeout rates over the years. We'll pick up there in part two of Toolbox's tribute to the strikeout. This article would not have been possible without research contributed by William Burke and Clay Davenport. Further Reading Rany Jazayerli, "Doctoring the Numbers-Ballardesque Strikeout Ratios": A look at pitchers with extremely low strikeout rates. Nate Silver, "Lies, Damned Lies-Strikeout Rate, Redefined": The debut of Nate Silver's signature stats column dealt with the difference between strikeouts per nine innings and strikeout rate. James Click, "Baseball Prospectus Basics-Statistical Consistency": This study showed the level of year-to-year consistency of various statistical measures by measuring their correlation coefficients over the period from 1991 to 2003. Among pitching statistics, SO/9 and groundball/flyball ratio were the ones that showed significant year-to-year consistency. Nate Silver, "Lies, Damn Lies-Secret Sauce": Explaining the relationship between a team's normalized strikeout rate (EqSO/9) and their postseason success. For more on this subject, you can also review Silver's chapter (with Dayn Perry) in Baseball Between the Numbers, entitled "Why Doesn't Billy Beane's [Expletive Deleted] Work in the Playoffs?"
Derek Jacques is an author of Baseball Prospectus.
|
