clock menu more-arrow no yes mobile

Filed under:

Why the MLB Pitching 'Win' Stat Is Obsolete in One Graph

The number of innings pitched by MLB starters has fallen dramatically over the last century-plus, rendering the stat all-but obsolete.

Eileen Blass-USA TODAY


Baseball has long been a nine-inning game. 100 years ago, starting pitchers used to throw all nine of those innings nearly every start. That has gradually changed over the decades, making starting pitchers more and more dependent on their team's relievers and batters in order to receive a "win" in the box score.

A starting pitcher has to throw five innings to qualify for a win, hence the lowest value on the vertical axis above. As you can see, over the last century-plus, the average number of innings a Major League starting pitcher throws in a game has fallen significantly.

  • In 1905, starting pitchers averaged 8.84 innings pitched (IP) per start (GS). So the stat of pitching "wins" made much more sense back then. Starters regularly threw 9 IP each time out.
  • MLB starters first averaged less than 8 IP/GS in 1939. Starters averaged 7.98 IP/GS that year and again in 1940. Then starters' IP/GS jumped back above 8.0 IP/GS until 1946. MLB starters have not averaged more than 8.0 IP/GS since 1945.
  • MLB starters must throw at least 5 IP in order to qualify for a win. Therefore, the halfway point between a complete nine-inning start and the bare minimum for a pitching "win" is 7.0 IP. 1965 was the first season in which MLB starters' average IP/GS was closer to the minimum of 5.0 IP than a CG. From 1974 to the present day, MLB starters have averaged an IP/GS below 7.0 IP in every season.
  • The Quality Start (QS) stat is based on the criteria of allowing three or fewer runs over six innings or more. The average MLB starting pitcher IP/GS first fell below the 6.0 threshold in 1995. MLB starter IP/GS bounced back up to 6.08 IP/GS in 1998, but then dropped to less than 6.0 IP/GS until 2011, when it rose to 6.02 IP/GS. Starter IP/GS hasn't been above 6.0 IP/GS in either season since. Last year, starters averaged 5.9 IP/GS.

The pitching "win" was a stat that made some sense for much of the first half of the 20th century. However, the decades since have made it so a pitching "win" is greatly influenced by how a pitcher's reliever and batter teammates perform in the innings that are played after he is removed from the game. Simply put, a pitching "win" is no longer as reflective of an individual starter's performance today as it was way back before a lot of us were born. It's a stat that is fast becoming obsolete.