Thursday, June 30, 2011
The best pitching seasons since 1960
By David Schoenfield
Who had the best season since 1960? Pedro, Randy and Bob Gibson are in the running.
Last week on his blog, Curt Schilling wrote an interesting post on what he called Pitcher Dominance Factor -- a way to evaluate the best starting pitchers. His formula basically rates a pitcher by comparing his ERA to the league average ERA for starters, and comparing his baserunners allowed per nine innings to the league average.
In many ways the formula works perfectly -- it rewards pitchers with low ERAs and those who don't allow many baserunners. It has a couple primary flaws in assessing overall value: There are no park effects and it doesn't factor in innings pitched, so a 170-inning season can have the same "dominance factor" as a 250-inning season.
Of course, there are other metrics out there as well to measure the best pitcher seasons.
Wins: I think we all agree by now that wins aren't the best way to measure a pitcher. By this method, in 1990 Bob Welch had the second-best season since 1960 with 27 wins -- despite ranking just sixth in his own league in ERA and posting a mediocre 127/77 SO/BB ratio.
ERA: Better than wins, but doesn't adjust for home ballpark, era (a 2.05 ERA in 1968, when the American League ERA was 2.98 is not as valuable as a 2.05 ERA in 1996, when the AL ERA was 5.00) or innings pitched.
ERA+: The Baseball-Reference stat adjusts a pitcher's ERA for his park and era to a scale where 100 is average. Much better than regular ERA, although it still doesn't factor in innings pitched, unearned runs, or the quality of a pitcher's defense. It actually ends up generating a list similar to Schilling's as 15 of the top 25 ERA+ seasons since 1960 occurred between 1994 and 2005. ERA+ also doesn't factor in "dominance" in the sense of what Schilling was looking for, as it's possible to post a low ERA without dominating peripherals.
WAR (wins above replacement): WAR establishes a value to a pitcher's season. FanGraphs and Baseball-Reference have different versions of WAR, and both arguably have a problem in generating a list of the best pitching seasons since 1960.
Under the Baseball-Reference formula, the more innings you pitch, the more value you accumulate. While completely logical (you're more valuable if you pitch more innings), its list ends up skewed towards the late '60s and early '70s, when starters routinely racked up 300-plus innings. Fourteen of its top 25 seasons reached 300 innings (and only four have fewer than 275), so modern pitchers struggle to make the top of the list since they don't pitch as many innings.
FanGraphs' version of WAR doesn't factor a pitcher's ERA or runs allowed. Instead, it extrapolates a pitcher's innings, strikeouts, walks allowed and home runs allowed (adjusted for park and era). It also only has pitcher WAR dating back to 1974. As it turns out, nine of its top 10 seasons occurred between 1995 and 2004, and you also end up with seasons like Schilling's 2002 in the top 10, when he had a great strikeout-to-walk ratio, but finished just 10th in the NL in ERA.
Both the B-R and FanGraphs lists are a terrific guide to greatness. I'm not knocking their lists. I'm just not sure either should be considered the definitive list of the "best" seasons. So here's my subjective 15 best since 1960, trying to account for the changing patterns of the game ... with apologies to the many great pitchers who finished 16th in my personal tally.
15. Steve Carlton, 1980 Phillies
24-9, 2.34 ERA, 304 IP, 243 H, 90 BB, 286 SO, 15 HR The last pitcher to throw 300 innings, Carlton led the NL in wins, innings, strikeouts and SO/BB ratio. He held opponents to a .218 average -- and this despite Greg Luzinski usually behind him in left field.
14. Gaylord Perry, 1972 Indians
24-16, 1.92 ERA, 342.2 IP, 253 H, 82 BB, 234 SO, 17 HR A great season that nobody remembers, although Perry did win the Cy Young Award. While 1972 didn't see much hitting -- there was a reason the AL instituted the DH rule for the 1973 -- Perry's workload was amazing as he had 29 complete games and averaged 8.5 innings per start. That's a lot of spit. From May 6 through Aug. 5, he started 22 games and averaged more than nine innings per start, all while compiling a 1.55 ERA.
13. Randy Johnson 2002 Diamondbacks
24-5, 2.32 ERA, 260 IP, 197 H, 71 BB, 334 SO, 26 HR Johnson captured his fourth straight Cy Young Award with one of his many dominant seasons. He allowed two runs or fewer in 22 of his 35 starts and struck out 15 or more four times. He also tightened up when it most counted, holding batters to a .174 average in "high-leverage" situations, with just one home run allowed in 154 at-bats.
12. Zack Greinke, 2009 Royals
16-8, 2.16 ERA, 229.1 IP, 195 H, 51 BB, 242 SO, 11 HR Greinke hasn't been able to replicate the consistency, focus or, yes, the little bit of magic he had in 2009. He allowed one run or less in more than half of his starts -- 18 of 33 -- but due to lack of run support won just 12 of those 18 games. (Thank you, Royals teammates.) His relative lack of stamina prevents him from ranking higher as he pitched fewer than seven innings in 13 of his starts.
11. Greg Maddux, 1995 Braves
19-2, 1.63 ERA, 209.2 IP, 147 H, 23 BB, 181 SO, 8 HR The shortened season cost Maddux a few starts, otherwise he'd rank even higher with the additional innings pitched. He allowed 39 runs (38 earned) in 28 starts and had a remarkable 0.81 WHIP. His .224 OBP allowed is the second-best since 1960 and his ERA+ is third-best (he ranked even better in 1994, although he did allow nine unearned runs that year.) Maddux allowed more than two runs just four times and more than three runs just twice. He painted the corners, got grounders, didn't give up home runs and dominated without just blowing batters away. Some say he also had a few extra innings off the plate to work with thanks to the kindness of the umpires.
10. Tom Seaver, 1971 Mets
20-10, 1.76 ERA, 286.1 IP, 210 H, 61 BB, 289 SO, 18 HR The only season on our list not to result in a Cy Young trophy, as Seaver lost out to Fergie Jenkins, who went 24-13 with a 2.77 ERA for the Cubs. The vote actually wasn't a travesty; once you adjust for Shea Stadium and Jenkins' 39 more innings pitched, the seasons are close in value. However, Seaver did outpitch Jenkins by quite a bit on the road: 1.63 ERA versus 2.70 ERA, so I rate Seaver's season as better. Seaver's 9.1 K's per nine led the league, at a time when the league average was 5.4, a ratio of +69 percent. In 2011, the NL average is 7.2 K's per nine; a rate of +69 percent would equate to 12.2 K's per nine.
9. Ron Guidry, 1978 Yankees
25-3, 1.74 ERA, 273.2 IP, 187 H, 72 BB, 248 SO, 13 HR My buddy Bish is going to punish me for not ranking Louisiana Lightning her, but this is a tough field. Guidry, all 160 pounds of him, threw 16 complete games and tossed nine shutouts. Guidry didn't lose until July, and in September, with the Yankees battling the Red Sox for the AL East title, he went 6-1 with a 1.19 ERA. He went 3-0 in four starts against the Red Sox. Can we have a re-do on that AL MVP vote?
THE METRIC SYSTEMS
How various systems rate the best seasons since 1960:
1. Pedro Martinez, 2000 Red Sox: 291
2. Greg Maddux, 1994 Braves: 271
3. Greg Maddux, 1995 Braves: 262
4. Bob Gibson, 1968 Cardinals: 258
5. Pedro Martinez, 1999 Red Sox: 243
1. Steve Carlton, 1972 Phillies: 12.2
2. Bob Gibson, 1968 Cardinals: 11.9
3. Dwight Gooden, 1985 Mets: 11.7
4. Bob Gibson, 1969 Cardinals: 11.0
5. Sandy Koufax, 1966 Dodgers: 10.8
FanGraphs WAR (since 1974)
1. Pedro Martinez, 1999 Red Sox: 12.1
2. Roger Clemens, 1997 Blue Jays: 11.1
3. Randy Johnson, 2001 D-backs: 10.7
4. Pedro Martinez, 2000 Red Sox: 10.1
5. Roger Clemens, 1988 Red Sox: 10.0
8. Pedro Martinez, 1999 Red Sox
23-4, 2.07 ERA, 213.1 IP, 160 H, 37 BB, 313 SO, 9 HR Pedro's second of three Cy Young reasons resulted in an opponents' batting line of .205/.248/.288, as he allowed just nine home runs while fanning 13.2 per nine, the second-highest total ever for a starter. He fanned 15 batters four times, 16 once and 17 once. He allowed more than four runs just once, a nine-run disaster after his All-Star start that landed him on the DL. In fact, that's the only strike against this season: He made just 29 starts. Well, that and these gloves he wore during the playoffs.
7. Randy Johnson, 2001 Diamondbacks
21-6, 2.49 ERA, 249.2 IP, 181 H, 71 BB, 372 SO, 19 HR And if you want to give him extra credit for winning three games in the World Series, please do so. Johnson struck out 10 or more in 23 starts, including a record-tying 20 on May 8 against the Reds. His 13.4 K's per nine is the best ever for a starter, left-handed batters hit one home run off him all season and you didn't dare dig him against him: he also hit 18 batters. Yes
6. Sandy Koufax, 1966 Dodgers
27-9, 1.73 ERA, 323 IP, 241 H, 77 BB, 317 SO, 19 HR You could flip a coin between any of Koufax's three Cy Young seasons (1963, '64, '66). His strikeout and hit rates were better in 1965 than 1966, but he also allowed 14 more runs in the same number of starts. He threw 11 shutouts in 1963, but had a big home park advantage that year. In 1966, he had 1.52 ERA at home and 1.96 on the road, his most even split, so I give '66 the nod. This was his final season, as he pitched in so much pain doctors were injecting steroids directly into the elbow joint, according to Jane Leavy's "Sandy Koufax: A Lefty's Legacy." Down the stretch, with the Dodgers battling for the pennant, he started seven times over the final 26 days, posting a 1.20 ERA.
5. Steve Carlton, 1972 Phillies
27-10, 1.97 ERA, 346.1 IP, 257 H, 87 BB, 310 SO, 17 HR If you want to argue that Carlton's season was the best ever, I won't mount much of a disagreement. Carlton famously won nearly half of his team's 59 wins as he completed 30 of his 41 starts and threw eight shutouts. He held batters to a .207/.257/.291 line while making 31 starts on three days' rest. He allowed one home run to a cleanup hitter all season. And did you check the innings pitched total? The National League OPS that year was .680, not far below the .704 so far this season. Carlton's 12.2 WAR ranks No. 1 on Baseball-Reference's list since 1960.
4. Roger Clemens, 1997 Blue Jays
21-7, 2.05 ERA, 264 IP, 204 H, 68 BB, 292 SO, 9 HR While the AL was batting .271/.340/.428 in 1997, Clemens dominated with a .213/.273/.290 line. He won the pitching Triple Crown, leading the league in wins, ERA and strikeouts while allowing zero runs or one run in 20 of his 33 starts. Clemens had many great seasons -- seven Cy Young Awards, seven ERA titles -- but 1997 stands out as his best. Red Sox fans just punched the wall in disgust.
3. Dwight Gooden, 1985 Mets
24-4, 1.53 ERA, 276.2 IP, 198 H, 69 BB, 268 SO, 13 HR He allowed just a .201 average and .270 slugging percentage. Left-handed batters had a lower OPS off him than right-handers. In September, with the Mets fighting the Cardinals for a division title, he went 4-0 in six starts with a 0.34 ERA. His strikeout rate of 8.7 per nine may not appear to stand out now, but that's 58 percent better than the NL average. He even hit .226. And he did all this at 20 years of age while partying with Darryl Strawberry.
2. Bob Gibson, 1968 Cardinals
22-9, 1.12 ERA, 304.2 IP, 198 H, 62 BB, 268 SO, 11 HR Yes, 1968 was the Year of the Pitcher, but even in a league where the league ERA was 2.99, Gibson's 1.12 ERA ranks as fourth-best ERA+ since 1968. He threw 13 shutouts (meaning he was just 9-9 if he didn't throw up a zero). You can find areas to nitpick: He allowed nine unearned runs, he averaged just 7.9 K's per nine (although that was second-best mark in the NL). But this is what stands out most to me: He pitched seven innings his first two starts of the season ... and then at least eight every start after that. The man literally didn't have a bad game all season.
1. Pedro Martinez, 2000 Red Sox
18-6, 1.74 ERA, 217 IP, 128 H, 32 BB, 284 SO, 17 HR Can a guy who pitched 87 fewer innings than Gibson have had a better season? (Vote in the poll!) Pedro's opponent batting line is just sick: .167/.213/.259. Absolutely incredible. He was throwing 95-mph Wiffle balls that year, unhittable heaters and changeups and curveballs with precise location. The AL ERA in 2000 was 4.92, giving Pedro the best adjusted ERA since 1960. I think I answer it this way: If I wanted one of these guys pitching at his peak in a game to save the future of mankind, I'd take 2000 Pedro Martinez.
(Statistics from Baseball-Reference.com and FanGraphs.com.)