#1 Starters and the World Series
Overall, just over 20% of games are started by #1 starters. For World Series teams, that percentage is almost doubled, 37%. The typical World Series team has two starting pitchers that you would tend to describe as #1 starters.
The 1911 World Series is outside the scope of this research, but in 1911 both teams, the New York Giants and the Philadelphia Athletics, had strong starting rotations. I remember reading in some old book, I believe the 1912 Reach Guide, that in the matchup between John McGraw and Connie Mack, Mack had the advantage because he had three aces, whereas McGraw had only two. John McGraw had Christy Mathewson (26-13, 1.99 ERA) and Rube Marquard (24-7, 2.50), but no strong third starter. Mack had Jack Coombs (28-12, 3.53), Eddie Plank (23-8, 2.10), but also Chief Bender (17-5, 2.16). (Coombs ERA had jumped that year from 1.30 to 3.53, but Coombs was also a tremendous hitter, hitting .319 that year, driving in 23 runs and scoring 31 (!!). Can you imagine a pitcher scoring 31 runs in a season? He did pinch hit a few times, but all of his runs scored came when he was pitching.)
Anyway, that line has stuck with me for 25 or 30 years since I read it, as something will once in a while, and has helped to shape my thinking. This is the first time I have had the data organized to kind of put it to the test.
Historically, there is no apparent advantage in the World Series to the team that has more #1 starting pitchers, as identified by this method. Teams which have won the World Series have had, on average, 37% of their games started by #1 starters; teams which have lost the World Series have had 37%. Teams with a larger number of starts by #1 starters have lost the World Series as often as they have won it. Even when there was a large separation in the number of games started by #1 starters, there has been no advantage in the World Series. There have been 51 series in which one team had at least 30 more games started by #1 pitchers. The team with the advantage has won 25 World Series, lost 26.
The teams that had less starting pitching got to the World Series for some other reason. The illustrative case is the 1971 and 1979 matchups between the Orioles and the Pirates. The Orioles had far greater depth in starting pitching, measured here at 112-33 in 1971, and 100-30 in 1979. But the Pirates had Willie Stargell and Roberto Clemente or Dave Parker and Al Oliver and Richie Hebner and Manny Sanguillen. The Orioles had fantastic fielders and could counter Roberto with Frank Robinson, but the Pirates had more firepower. I’m not saying it was inevitable that the Pirates would win; if you could replay those series I think the Orioles would win as many as they lost—but not more.
Twelve teams have made the World Series with no pitcher who makes the list as a #1 starter. Four of those teams won the World Series; eight lost. The last team to do that was the 2013 Red Sox. Clay Buchholz was 12-1 with a 1.74 ERA, but made only 16 starts; by the time of the World Series his fastball was in the mid-eighties. He pitched four shutout innings in the World Series, but. . .go watch the tape. He was hanging on by a fingernail. Jon Lester was "our" #1, and scores here as a strong #2 but you know. . .Nine starting pitchers were mentioned in the Cy Young voting that year. He wasn’t one of them.