Starting Pitcher Leverage (Part 2)
Hi, and welcome back to The Hardball Times’ ongoing series of articles on starting pitcher leveraging throughout baseball history. In part one of this series I introduced the concept of starting pitcher leverage. If you remember it, skip the next paragraph, which briefly summarizes it.
From at least the 1880s until the mid-1960s teams routinely and intentionally used one or more of their starting pitchers against certain opposing teams as often as they could to get a little extra advantage. It could be using an ace pitcher against the best available competition, or it could entail starting a southpaw against a team with unusually strong left-handed hitters. In modern times, managers go with a rotation so rigid that the frequent leveraging of olden times no longer exists. This extinct approach toward using starting pitchers has never been the subject of any thorough statistical evaluation. Until now.
After looking at the best and worst leveraged careers in baseball history in part one, the natural follow up is the best leveraged seasons of all time. Again, AOWP+ is my statistic of choice. If you don’t remember what it is, you can either click back on the link to the first article, or scroll to the bottom and read the section at the end of the article explaining it. Short version: it’s modeled like ERA+ or OPS+, centered on 100, with a higher score indicating the pitcher faced the better clubs more often than one would expect.
One final reminder—though I haven’t figured the AOWP+ for every pitcher in baseball history I have looked at the overwhelming majority of starters from 1876-1969, determining the AOWP+ for over two-thirds of all starts in that season. While I’m sure there are some best/worst single seasons that fell through the cracks, as it’s almost impossible to get every single season without AOWPing every start in history, I have AOWPd over 75% of all seasons where a starter had at least 20 starts from 1876-1969. My best guess is that I have more than 75% of the best leveraged single seasons ever, because the guys pitching the better leveraged seasons should have superior careers, and those are the pitchers I checked on. Alternately, I ought to have less than three-quarters of the worst leveraged seasons.
Best Ever
With all that in mind, out of approximately 5,000 different seasons I AOWPed where a pitcher started at least 20 games, here are the 20 best leveraged single seasons:
Name GS Year AOWP+ R/L 1. Lew Richie 27 1912 117 R 2. Lefty Leifeld 30 1910 116 L 3. Reb Russell 25 1915 116 L 4. Max Lanier 20 1942 114 L 5. Whitey Ford 28 1954 112 L 6. Art Houtteman 25 1949 112 R 7. Thornton Lee 25 1937 112 L 8. Mordecai Brown 34 1909 111 R 9. Ken Heintzelman 24 1941 111 L 10. Mickey Haefner 20 1948 111 L 11. Pete Conway 31 1886 111 R 12. Carl Hubbell 20 1942 111 L 13. Ted Gray 21 1950 111 L 14. Lefty Gomez 31 1931 110 L 15. Bill Phillips 27 1899 110 R 16. Doc White 30 1904 110 L 17. Clarence Mitchell 22 1929 110 L 18. Adonis Terry 34 1886 109 R 19. Carl Hubbell 22 1941 109 L 20. George Mogridge 21 1916 109 L
Actually, the best AOWP+ every goes to Tommy Bond who recorded a 122 in 1875 with Hartford. That’s because there was no set schedule in the National Association, and half the teams (the lousier half) stopped playing midway through the year. Bond became the starting pitcher in the second half. That ain’t leveraging. That’s why in part one I disallowed the NA, UA, and the 1884 AA.
As was the case with the career leaders in part one, southpaws are significantly overrepresented. There are only six righties. Three of them pitched in the 19th century. Two of the remaining three pitched for the early 20th century Chicago Cubs. Art Houtteman is the only other righty.
If you’re wondering how come the six guys at 111 aren’t tied, there’s a reason. Though I present AOWP+ as an integer, the Excel database that calculates this slop gets into tons o’ decimals. There are actually 13 guys at 109, for instance. As for the actual seasons here, every decade from the 1880s to 1950s has at least one season listed. Hank Aguirre’s 1964 just misses the cut as he’s one of those 109s who the excel decimals decimates. In the last 40 years I don’t think anyone makes the top 200 seasons.
Also, the pre-WWI era dominates the top. You could use your pitchers on short rest more with the deadball and the then-new 154 game schedule meant you faced opposing teams far more often than had been the case previously. Let’s have a big hand for Lew Richie. Here’s the best leveraged season ever:
Team Pct GS NYG 682 10 Pit 616 6 Cin 490 3 Phi 480 2 StC 412 4 Brk 379 2 Bos 340 0
That’s so cartoonishly absurd you have to wonder if Tex Avery was the Cubs’ pitching coach. Mordecai Brown went down with an injury that year. If you check the 1912 Cubs, you can’t tell who the ace was. Folks, it was Lew Richie.
Even more impressively, he did it in 27 starts, when most guys on the list have 25 or less. The dominance of those barely qualifying makes sense. The more games you pitch, the more you have to start against everyone. Here’s the all-time Top 10 for those with 30 or more starts.
Name GS Year AOWP+ R/L 1. Lefty Leifeld 30 1910 116 L 2. Mordecai Brown 34 1909 111 R 3. Pete Conway 31 1886 111 R 4. Lefty Gomez 31 1931 110 L 5. Doc White 30 1904 110 L 6. Adonis Terry 34 1886 109 R 7. Murray Dickson 31 1954 109 R 8. Rube Walberg 30 1930 109 L 9. Early Wynn 36 1954 108 R 10. Hooks Wiltsie 38 1908 108 R
The best ever with 40-plus starts was Ed Seward, who posted a 108 AOWP+ in 1888 with 56 starts. That’s not a typo; 56 starts. It was a very different game back then.
Worst Ever
So much for the cream of the crop. Now for the crap of crop. Here’s the 20 worst leveraged single-seasons I know of:
Name GS Year AOWP+ R/L 1. Fred Goldsmith 24 1880 81 R 2. Joe McGinnity 20 1908 84 R 3. Phil Knell 31 1890 86 L 4. Jack Stivetts 20 1889 86 R 5. Bill Swift 22 1935 87 R 6. Kid McGill 20 1890 88 L 7. Frank Smith 23 1904 88 R 8. Harvey Haddix 22 1961 88 L 9. Paul Minner 22 1955 88 L 10. Dutch Leonard 28 1919 89 L 11. Jack Pfiester 25 1909 89 L 12. Chuck Stobbs 20 1953 90 L 13. Joe Bowman 24 1940 90 R 14. Dummy Taylor 27 1906 90 R 15. Ed Siever 33 1907 90 L 16. Eldon Auker 31 1936 90 R 17. Jack Powell 26 1905 91 R 18. Carl Lundgren 20 1903 91 R 19. Bobby Mathews 25 1879 91 R 20. George Hemming 21 1896 91 R
For me, the big shocks are Fred Goldsmith in the top slot and Bobby Mathews in 19th place. When I first looked at the data, I thought 1886 was the big breakthrough for pitcher leveraging. When both AA and NL expanded their schedules, AOWP+ scores skyrocketed while previously no one did better than 104 in a well-organized league.
I’ll discuss this in much greater detail in a later article, but it turns out that leveraging had always existed, but it began inverted. Leveraging initially meant who a pitcher was rested against, and in the mid-1880s flipped to who he was started against. It’s hard to have a really nice AOWP+ when you start over half your team’s games. Goldsmith and Mathews are rare examples of a team’s negative space pitcher getting 20-plus starts.
Another stunner is Hall of Famer Joe McGinnity coming in second place. On the surface it looks strange that the Giants got rid of him after 1908. He’d been terrible in ’07, but seemingly bounced back. Not quite. Only two of his 20 starts came against teams with winning records. Meanwhile, he faced last place St. Louis seven times. After accounting for quality of competition, it was his second straight lackluster season.
In general, these guys have even fewer starts than the best ever. Well, duh. If you don’t have enough faith to start a pitcher against the best teams, why would you give him many starts? Here’s the worst leveraged pitchers with at least 30 starts:
Name GS Year AOWP+ R/L 1. Phil Knell 31 1890 86 L 2. Ed Siever 33 1907 90 L 3. Eldon Auker 31 1936 90 R 4. Vinegar Bend Mizell 33 1953 92 L 5. Tommy Bond 64 1879 93 R 6. Joe Nuxhall 33 1955 93 L 7. Kaiser Wilhelm 36 1904 93 R 8. Bob Caruthers 33 1890 93 R 9. Bob Ewing 32 1906 94 R 10. Vinegar Bend Mizell 32 1960 94 L
And the award for biggest surprise of this article goes to… Tommy Bond! How in Hades do you get an AOWP+ of 93 when he start 64 games for a team that played 84 games? Forget why, let’s talk how. The logistics of the feat boggle the mind. First, as was often the case in the 19th century, it was a badly stratified league with the best team winning three times as many games as either of the two doormats. Bond started all two dozen games against the doormats. Stunningly, he only started four of his club’s twelve games against pennant winning Providence Grays.
Now, why would you treat your ace like that? My first hunch would be platoon leveraging. Bond was a righy and the team’s other pitcher, Curry Foley, was a lefty. Providence had two of the best hitting lefties in baseball in Tom York and Joe Start. Looks like platoon leveraging… maybe. There’s something else going on, though.
Bond pitched for Boston, who came in second place. They ended the year with a six game series against the Providence Grays, the game’s first ever winner-take-all series for a pennant. Boston needed some big pitching to succeed as they trailed Providence by three games in the standings. So what did they do? Well, they didn’t give the ball to Bond. Or Foley. Never mind that those two accounted for every start on the season.
Instead Boston’s manager, the great Harry Wright, found some guy named Jim Tyng, who had never played an inning of major league baseball in his life, and let him start the first three games of the series. Had it not been for a four inning relief appearance nearly a decade later, those three starts would’ve been his entire career. The Palmer/Gillette Encyclopedia doesn’t even know if he was a righty or lefty. Baffling. It’s like letting lucky ticket holder in Aisle 517, Row 4, Seat 101 get the nod.
Tyng actually won the first game, but lost the next two, ending their chances. Then Wright tabbed Lee Richmond for the next start. He was the game’s first quality southpaw, but this was his MLB debut. Finally, he let Foley and Bond take a start each to end it. Truly bizarre.
Looking at the full season, it makes a bit more sense. Harry Wright did engage in some platoon leveraging earlier in the year. In the previous six game series against the Greys he’d let Foley start half the games. Providence had two of the best left handed hitters in the game in Tom York and Joe Start. Foley lost all three games. So he didn’t trust Foley.
Bond had won two out of three against Providence in that earlier series, though. However, by the end of the year, I can only assume he’d broken down. In late July, Boston had been 21-19, in fourth place, closer to last than first. Wright decided to use Bond as an iron man pitcher at that point, having him start 35 consecutive games. He won 28 of them, putting them in the thick of the race. In a league averaging over five runs a game, he held teams to well under half of that.
Then his arm crapped out. He only pitched well in one of his last four starts. He only started one of the three games immediately prior to the big end series, and pitched poorly in it. Wright, desperate, scraped up some scrubs in a desperation ploy to save the season. Not surprisingly, it failed. Bond’s low AOWP+ was a fluke caused by a bit of platoon leveraging early in the year, an ill-timed injury late, the huge variations in won-loss percentages that gave one team unusual importance, and finally the nature of the 1879 schedule where Boston only faced their archrival in two separate extra-long series.
Yeah, like I said, it was a very different game. Aiding the theory that Bond injured his arm, he never was that good again after his fantastic run in 1879. Those 35 consecutive starts were too much strain for the 23-year-old’s arm. Sorry for going on for so long, but I found that really interesting.
Big Heaping Gobs of Data
One last thing I want to share—the best and worst known AOWP+s for each season from 1876-2005. (I did this study before 2006 ended and I see no point checking on it because pitchers ain’t leveraged anymore). Most of the numbers are self-explanatory, but please note the number on the far left indicates how many pitchers with at least 20 starts I AOWPed that year. Aside from 1884, the info relates to all MLB pitchers. (For 1884 it’s only the NL, and the number on the far right for 1884 also refers only to NL pitchers).
Later on in this series of articles (not right away), I’m going to trace the history of leveraging over the decades. Consider this an opening salvo of some what’s to come. It’s always based on GS, not IP, even when the latter’s available. Here’s the year-by-year extremes:
# Leaders AO+ Year Last AO+ 5 Cummings, Candy 101 1876 Spalding, Al 95 2 Bradley, George 103 1877 Bond, Tommy 102 3 White, Will 101 1878 Ward, John 95 7 Ward, John 103 1879 Mathews, Bobby 91 10 Richmond, Lee 104 1880 Goldsmith, Fred 81 10 Keefe, Tim 103 1881 Welch, Mickey 97 16 McGinnis, Jumbo 104 1882 Galvin, Pud 98 22 Mathews, Bobby 104 1883 Buffinton, Char 96 8 Buffinton, Charli 103 1884 Ferguson, Charl 96 26 Buffinton, Charli 103 1885 Conway, Pete 95 33 Conway, Pete 111 1886 Henderson, Hard 94 35 Smith, Elmer 105 1887 Cushman, Ed 94 35 Seward, Ed 108 1888 Staley, Harry 94 41 King, Silver 107 1889 Stivetts, Jack 86 47 Mullane, Tony 107 1890 Knell, Phil 86 40 Stivetts, Jack 107 1891 Ehret, Red 93 36 Keefe, Tim 104 1892 Breitenstein, T 94 33 Dwyer, Frank 104 1893 Keefe, Tim 92 29 Hutchinson, Bill 106 1894 Gleason, Kid 94 28 Stivetts, Jack 107 1895 Gumbert, Ad 96 26 Terry, Adonis 104 1896 Hemming, George 91 26 Orth, Al 105 1897 Cunningham, Ber 93 34 Griffith, Clark 104 1898 Weyhing, Gus 95 39 Phillips, Bill 110 1899 Donahue, Red 94 29 Fraser, Chick 104 1900 Donahue, Red 97 47 Callahan, Nixey 108 1901 Dugglesby, Bill 95 46 White, Doc 105 1902 Garvin, Ned 95 44 Taylor, Iron Jack 106 1903 Lundgren, Carl 91 50 White, Doc 110 1904 Smith, Frank 88 54 Reulbach, Ed 109 1905 Powell, Jack 91 56 Mathewson, Christ 107 1906 Taylor, Dummy 90 50 Flaherty, Patsy 108 1907 Siever, Ed 90 40 Wiltsie, Hooks 108 1908 McGinnity, Joe 84 40 Brown, Mordecai 111 1909 Pfiester, Jack 89 40 Leifeld, Lefty 116 1910 Willis, Vic 91 42 Brown, Mordecai 107 1911 Reulbach, Ed 95 42 Richie, Lew 117 1912 Brown, Buster 97 41 Camnitz, Howie 108 1913 Packard, Gene 92 64 Collins, Ray 103 1914 Bender, Chief 93 63 Russell, Reb 116 1915 Cicotte, Eddie 92 48 Mogridge, George 109 1916 Shawkey, Bob 95 51 Ehmke, Howard 105 1917 Hendrix, Claude 93 37 Coombs, Jack 105 1918 Gregg, Vean 96 43 Benton, Rube 108 1919 Leonard (I), Du 89 56 Marquard, Rube 105 1920 Ayers, Doc 95 45 Nehf, Art 104 1921 Myers, Elmer 92 48 Alexander, Pete 104 1922 Morton, Guy 95 46 Aldridge, Vic 103 1923 Ferguson, Alex 95 45 Nehf, Art 107 1924 Shawkey, Bob 97 53 Davis, Dixie 106 1925 Rommell, Eddie 94 50 Hoyt, Waite 106 1926 Thomas, Tommy 96 52 Shaute, Joe 106 1927 Blake, Sherriff 94 51 Uhle, George 105 1928 Haines, Jesse 94 53 Mitchell, Clarenc 110 1929 Haines, Jesse 95 53 Walberg, Rube 109 1930 Faber, Red 94 50 Jones, Sad Sam 108 1931 Hoyt, Waite 93 56 Gomez, Lefty 110 1932 Pipgras, George 91 48 Walker, Bill 107 1933 Swift, Bill 95 46 Stewart, Lefty 107 1934 Lucas, Red 91 46 Clanton, Cy 107 1935 Swift, Bill 87 42 Knott, Jack 109 1936 Auker, Eldon 90 50 Lee, Thornton 112 1937 Dietrich, Bill 91 46 Gomez, Lefty 106 1938 Whitehill, Earl 94 47 Smith, Eddie 109 1939 Allen, Johnny 92 45 Butcher, Max 105 1940 Bowman, Joe 90 44 Heintzelman, Ken 111 1941 Schumacher, Hal 94 41 Lanier, Hal 114 1942 Bridges, Tommy 92 41 Davis, Curt 103 1943 Newsom, Bobo 96 40 Roe, Preacher 105 1944 Bowman, Joe 97 35 Newhouser, Hal 105 1945 Wyse, Hank 95 42 Gromek, Steve 109 1946 Haefner, Mickey 94 46 Schmitz, Johnny 108 1947 Hearn, Jim 95 51 Haefner, Mickey 111 1948 Simmons, Curt 95 49 Houtteman, Art 112 1949 Meyer, Russ 92 48 Gray, Ted 111 1950 Stobbs, Chuck 92 42 Raschi, Vic 106 1951 Trout, Dizzy 93 45 Pierce, Billy 107 1952 Minner, Paul 94 43 Staley, Gerry 107 1953 Stobbs, Chuck 90 46 Ford, Whitey 112 1954 Byrd, Harry 92 49 Pierce, Billy 107 1955 Minner, Paul 88 45 Nixon, Willard 107 1956 Nuxhall, Joe 94 48 Hoeft, Billy 107 1957 Garver, Ned 92 47 Lary, Frank 103 1958 Wilson, Jim 95 49 Sanford, Jack 103 1959 Herbert, Ray 96 51 Ford, Whitey 106 1960 Pascual, Camili 93 53 Perry, Jim 105 1961 Haddix, Harvey 88 58 Simmons, Curt 108 1962 Washburn, Ray 94 63 Mahaffey, Art 105 1963 Williams, Stan 93 58 Aguirre, Hank 109 1964 Jay, Joey 94 65 Lemaster, Denny 104 1965 Bouton, Jim 93 64 Ellis, Sammy 103 1966 Giusti, Dave 95 61 Maloney, Jim 103 1967 Ellis, Sammy 95 64 Krausse, Lew 103 1968 Odom, Blue Moon 97 57 Hunter, Catfish 105 1969 McGlothlin, Jim 97 54 Bolin, Bobby 105 1970 Peters, Gary 96 48 Short, Chris 104 1971 Dobson, Chuck 95 41 Downing, Al 104 1972 Pappas, Milt 95 40 Wise, Rick 102 1973 Pappas, Milt 97 35 Wilson, Don 103 1974 Siebert, Sonny 97 32 Seaver, Tom 103 1975 Cuellar, Mark 96 31 Perry, Gaylord 104 1976 Torrez, Mike 97 27 Perry, Gaylord 105 1977 Jenkins, Fergie 97 30 Jenkins, Fergie 103 1978 Lonborg, Jim 98 29 Guidry, Ron 103 1979 Flanagan, Mike 97 30 Tiant, Luis 103 1980 Knepper, Bob 97 28 Reuss, Jerry 104 1981 Niekro, Phil 95 32 Blue, Vida 105 1982 Seaver, Tom 98 31 Flanagan, Mike 103 1983 Guidry, Ron 97 31 Gooden, Dwight 102 1984 Eckersley, Denn 96 29 Langston, Mark 105 1985 Sutcliffe, Rick 95 36 Ryan, Nolan 103 1986 Welch, Bob 98 37 Martinez, Dennis 104 1987 Hough, Charlie 98 32 Flanagan, Mike 102 1988 Knepper, Bob 96 33 Hershiser, Orel 102 1989 Flanagan, Mike 97 32 Appier, Kevin 102 1990 Brown, Kevin 97 30 Glavine, Tom 103 1991 Maddux, Greg 97 33 Witt, Bobby 104 1992 Welch, Bob 97 35 Schilling, Curt 103 1993 Gooden, Dwight 95 25 Smoltz, John 104 1994 Hershiser, Orel 95 25 Morgan, Mike 105 1995 Clemens, Rogers 95 25 Moyer, Jaime 103 1996 Key, Jimmy 97 23 Glavine, Tom 103 1997 Cone, David 98 23 Martinez, Pedro 103 1998 Gooden, Dwight 96 23 Clemens, Roger 104 1999 Erickson, Scott 96 16 Appier, Kevin 103 2000 Burkett, John 98 13 Maddux, Grege 102 2001 Burkett, John 97 15 Erickson, Scott 103 2002 Pettitte, Andy 94 13 Mussina, Mike 103 2003 Burkett, John 97 11 Brown, Kevin 103 2004 Glavine, Tom 98 11 Mussina, Mike 104 2005 Rogers, Kenny 97
References & Resources
What the heck is AOWP+?: The stat I invented to judge pitcher leveraging. It’s AOWP/TOWP*100. AOWP is Average Opponent Winning Percentage. TOWP is Team’s (Average) Opponent Winning Percentage. To figure AOWP for a single season, you take the number of starts a given pitcher had against each opposing team, and multiply that by the team’s winning percentage. After doing this for all rival squads, add up the products and divide by the pitcher’s total GS. The result is his AOWP. The same logic applies to TOWP, only here you look at how many games the team played against all rivals. If a pitcher’s used evenly, his AOWP will be the same as the TOWP, and he’ll have an AOWP+ of 100. If he’s used more against better teams, he’ll have a higher AOWP+. I calculated AOWP+ for 659 pitchers who started 182,000 games, including over two-thirds of all games from 1876-1969.