He asked for input on the idea, and here is my response:
Thanks for the tip on the article. I think that this is a very interesting idea. Clearly the percentages Triple Crown is superior to the conventional one in evaluating players. I have a few salient comments that I would like to express re. the idea:
- First, I do believe that Rob Neyer, back when he actually used to write articles, did a bit on this. It may have been a year or two ago.
- As you say, "The Triple Crown is an interesting accomplishment. Not really an award, or even a real title, it has long been a part of the baseball vernacular, an acknowledgement of a season of singular excellence." It is a label, like a .300 hitter. It is basically an anachronism now (given that no one has been so labeled in 35 years).
Baseball stats change over time. The question has always been how do you evaluate the players given the statistics available. Originally, (like 1850s and '60s) runs per game was the stat used to evaluate players. (And what ever happened to calling shutouts Chicago games?) The problem was they would list all players, so someone who played 2 games and scored 15 runs would win the "Batting Crown". (If that is what they called it--I have to check the Dime Guides.) It eventually morphed into run per at bat and then into hits for at bat (sometimes that even included walks sort of an early OBP). The idea of RBI was added to illustrate a players' ability to drive in runs. The conventional statistics became set in stone as baseball history and journalism matured. The idea of a Triple Crown winner formed at about the same time in the primordial slime, if you will, of baseball.
New stats have come around to better answer the questions for which the old statistics were designed. RBI has proved to be more a matter of where the batter fits into a lineup and who bats in front of him than actual production. Slugging percentage is a better gauge of production (or situational hitting stats).
But what does it boil down to? A label. People prefer to use the labels that they are used to even if there are better ones. In Ken Burns' baseball documentary, someone (probably banality mongers Costas or Will) said that a .300 hitter is always a .300 hitter. We know that this is not true over time and place, but people feel comfortable with it.
- A Park-adjusted, league-adjusted OPS is probably the best tool to evaluate a given player irrespective (as possible) of his park and his era. Some prefer Runs Created per 27 Outs or Offensive Win Shares. Take your pick. They are all better at evaluating talent than outmoded labels like Triple Crown winner (even if it is updated with better stats).
Why replace one outmoded label for another. Sabermetrics have progressed so rapidly that it is a pointless intermediate step. The Triple Crown remains as a sort of appendix, an organ that doesn't do much but you still carry around. You don't remove it unless it gets infected, and here the analogy breaks down. But you get the idea.
- By the way, and this is a minor point, you say that the triple crown stats "represent excellence in three fairly distinct categories." In either case, the stats are intertwined. In the conventional Triple Crown, a HR results in, at least, an RBI and it is a hit that affects your BA. In the percentages Triple Crown, a hit affects all three. So they are not entirely independent of each other.
- I do think that Kevin Reichard was a bit harsh in his review. So you can't isolate a player from his environment--"There is, of course, a great flaw in your argument: you can never negate the influence of teammates and opponents." So what? It's a team sport. All statistics are based on situations created by your team and the opponents' team. They are part and parcel of the statistical interpretation of baseball accomplishments. I do not believe that it diminishes your argument one iota.
Also, Bonds playing 16 fewer games clearly has less weight in the at bat differential between him and Berkman than the ludicrous number of walks he has received.