Comparing pitchers across generations

My last post discussed ways in which hitters can be more accurately compared across generations while acknowledging these comparisons won't be perfect. This post focuses on pitchers.

Comparing pitchers from different eras is basically a fool's exercise, which is why I'm undertaking it. Since organized baseball began in 1871, teams have gone from using one pitcher to as many as twelve or thirteen. The number of starts has decreased, and it's important to note just how long it's been since pitchers finished even 25 percent of the games they started:

Complete games dropped below fifty percent of starts almost one hundred years ago, and dipped below twenty-five percent in the 1970s. Couple that with the move to five-man rotations, and what used to be 38-40 starts for an ace has dropped to 32-34. As a result, wins are much more difficult to accumulate, so using wins as a way to compare pitchers falls short. ERA doesn't fare much better, since it's predicated on the run environment. Different measures like strikeouts, WHIP and others fall short as well due to changes in hitting patterns.

Having written this, there is one statistic that can help put all pitchers in one place to see how they stack up: Fielding Independent Pitching Minus (FIP-). Since successful pitching is largely the result of reducing hits, and since FIP- values are park and league adjusted, Walter Johnson's 5.3 K/9 rate won't pale in comparison to Nolan Ryan's 9.5 -- both will be compared to their respective leagues, parks and eras.

FIP- shows the percentage a pitcher was better than the rest of baseball. If a pitcher had a FIP- of 95, he was five percent better than the league in the given period, and a FIP- of 105 was five percent worse. These are the pitchers with the best career FIP- in baseball history:

Other traditional pitcher measures are shown to add context, but the person I'll be writing about a lot in the next couple of months tops the list. A quick review of the list suggests several things -- mainly, that FIP- works as a measure of determining the best pitchers, since the players we would expect to be represented are indeed there. It works as a cross-generational test because pitchers from almost all eras are included, except for baseball's infancy. There are some anomalies (Noodles Hahn?), but in general the list lines up well with those widely regarded as the best pitchers in baseball history.

There will be no shortage of Hall of Fame voters who will bemoan that Pedro won "only" 219 games in his career, and that this will somehow sully the Hall of Fame. Those people would be wrong on two counts. First, the win has lost its place as a useful measure of pitcher success as pitchers have fewer opportunities to work themselves out of jams. Second, take a look at the wins leaders among active pitchers and mentally determine where each pitcher will likely top out when his career is over. Forget about 300 wins -- without drastic changes in starting pitcher utilization, 200 wins will be an achievement in itself. Of course, by then hopefully the win will have reached its demise as a baseball stat, but that's a post for another day.

Nick Shepkowski of 670 The Score in Chicago wrote a post last winter after Greg Maddux wasn't a unanimous choice in last year's Hall of Fame vote. Other than the very specific facts stated in the post, Pedro Martinez' name can be inserted in place of Greg Maddux in every place and reach the same conclusions. Pedro Martinez put up his numbers in the midst of one of the greatest offensive surges in baseball history --  he put up Dead Ball Era numbers during a hitter's era, the very definition of a Hall of Famer. He's certainly in the discussion of the best pitchers of his generation, and FIP- places him at the top in career achievement. Does this mean he was the best pitcher ever? No way I'll make that kind of blanket statement, because baseball really does change enough to make statements like that meaningless, but I'm very comfortable stating he belongs with the best that ever were.

*** Look at the data yourself. This shows the FanGraphs data, limited to pitchers with at least 2000 career innings -- download the data and reach your own conclusions. You can also view a Tableau data viz and play around with it to see the data represented visually.***

All data from Baseball-Reference and FanGraphs

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