Kill the win

In my post last week in which I shared my thoughts on the Hall of Fame selections, I made an offhand comment regarding pitching wins and Pedro Martinez. I included a link to the active leaders in wins, but sometimes a link isn't good enough -- this chart shows the top 20 active pitchers in wins:

Three active pitchers have over 200 wins, Mark Buehrle will by the end of April, and then there's a huge dropoff to A.J. Burnett with 155, roughly two-plus seasons worth. Pedro Martinez was elected despite having "only" 219 wins in his career -- he'd be the active leader if he were still pitching. Forget about 300 wins; 250 is going to be hard enough to reach for modern pitchers.

Three pitchers won over 20 games in 2014 -- Clayton Kershaw, Johnny Cueto, and Adam Wainwright, which is two more than in 2013 when Max Scherzer was the only one. Ask yourself an easy question -- how many 20-game win seasons have there been in baseball history? I'll even give you some help -- there have been almost 2,800 team seasons since the National Association was formed in 1871. Starting rotations have expanded from one to the current five and have been at least four as long ago as 1900, so there have been approximately 10,000 or so opportunities for pitchers to win 20 games.

This chart should help:

There have been 1,203 seasons in which a pitcher won 20 or more games, almost 400 of which were prior to 1901 and the formation of the American League. The bulge in the 1970s was probably a result of expansion and a 162-game season as much as an era of good pitching, but those days are over.

This chart explains a great deal of the reason:

Pitchers receive no decision in around thirty percent of their starts. There are some interesting inflection points in the charts as well as a brief gap from around 1973-1980 in which the introduction of the designated hitter appeared to change pitcher utilization by league, but even that has mostly gone away. It's difficult to accumulate wins when pitchers aren't staying in games to finish what they started.

This chart shows the percent of games that are complete games since 1871:

It's a bit out of date, but nothing happened in 2014 to change it dramatically. Quick, name the leaders in shutouts and complete games in 2014 -- they were Kershaw with six shutouts and Wainwright, Henderson Alvarez, and Rick Porcello with three complete games. We don't reference these numbers much anymore because their relative scarcity makes them meaningless in making useful comparisons -- it's been almost forty years since pitchers finished even twenty-five percent of their games and almost one hundred since they completed half or more. With fewer starts, fewer complete games and more no-decisions creating the perfect recipe, the win is likewise losing its relevance.

None of what I've written is new or novel. I've only added some reference to an issue that has already been rendered moot by a great number of people. So what should we do, just ignore the win? That would be my first suggestion, but that won't happen overnight, and in seasons like Kershaw's and Scherzer's in 2013 the win can be useful -- Kershaw got those 21 wins in only 27 starts, and that's an impressive feat not matched by many in baseball history.

Back in the day when I was still blogging on my own (that would be 2013) I wrote on this issue a couple of times (here and here, and my favorite one is here, just for my title alone). I've learned a couple things since then, and my next post will have some suggestions for what we can replace wins with. It won't be perfect, but as pitcher utilization patterns change, the average fan is going to need something better to use as a quick guide to pitcher effectiveness.

All data from Baseball-Reference and FanGraphs. Any errors in gathering and processing it are the author's.

Scott Lindholm lives in Davenport, IA. Follow him on Twitter @ScottLindholm.