The available shot-attempt data doesn’t go back an entire decade. It’s all very useful, but historical debates over players are incredibly hard to quantify. Many analytically inclined fans have things they figure based on what kinds of players are over or under rated today, but that’s far from a perfect method. It’s not just that we lack data, but having not lived through these games, we often lack context.
In his 1995 book Whatever Happened To The Hall of Fame
, Bill James discusses this at length. He notes that as time passes since a player’s retirement, especially once they’re involved in the hall of fame argument, debates become more statistically oriented. It’s harder to feel about a player exactly as we felt watching them, and statistics become our crutch. As the godfather of sabermetrics, you’d imagine James would welcome this. He does the opposite.
Take the chapter on “Gordon vs. Doerr.” Joe Gordon and Bobby Doerr were two of the premier second basemen of their era. Gordon played for the New York Yankees, Doerr for the Boston Red Sox. Both were largely overshadowed by legendary teammates in their careers. Gordon by Joe DiMaggio, and Doerr by Ted Williams.
James lamented that Doerr had been elected to the Hall of Fame . Doerr had a .288 batting average to Gordon’s .268. Gordon hit about 30 more career home runs, and Doerr had far more doubles. The basic statistics lean Doerr, but left out context.
Today we have all sorts of park adjustments, but the short version of the story is they were both right handed bats. That means that Gordon’s long fly balls were going to left field, and Yankee Stadium had a far left field wall. Doerr played at the smaller Fenway park, and while the Green Monster can rob you of some home runs, bounces off the wall cause a lot of would-be fly balls to become doubles. OPS+, a park-adjusted offensive catch-all that’s popular today, measures a player as a percentage of an average player’s expected offensive contribution in that park, league, and year. Gordon’s career OPS+ was 120, and Doerr’s 115. Gordon was elected to the Hall of Fame in 2008, 13 years after this book was written, and 30 years after Gordon died.
James’ point is that people at the time, without relying heavily on statistics, would know this, and account for it. Considering Gordon’s MVP award, one can see his point. We have other examples, like this great article
about Joe Namath. Young, stats-loving fans assume Namath was overrated because he won a big Super Bowl. In actuality, Namath’s low completion percentage is because he never took sacks. He always managed to get the ball away, and while his completion percentage is lower, he wasn’t losing yards like other quarterbacks who didn’t get the ball away as well.
This is about hockey, and baseball and football talk might be lost on you. What we learn from experience in other sports is that very often when we use statistics, we’re looking at a player in a very basic way. Instead of looking at a highly-rated player’s negative Corsi Rel and declaring him overrated trash, use those situations as a hint at what guys you should be looking closer at. No single number, not even any Wins Above Replacement model, can tell you all you need to know to discuss the player in any informed way.
There are a lot of players I’ve looked at the basic numbers, information, and circumstances of, and decided are probably overrated. Maybe Grant Fuhr was just a guy standing behind the best team ever, but maybe the Oilers took a huge number of penalties. Maybe shot quality mattered more back then. Bob Pulford and Dick Duff might be two guys on great teams inducted out of an over-represented era. They might also have been the 1960s equivalent of Patrice Bergeron and Brad Marchand. I honestly have no idea about any of this, and I plan on eventually doing a better job of figuring it out. If the push towards data has shown us anything, though, it’s that we shouldn’t be dismissive of things we haven’t done the work on.