Tuesday, January 05, 2010

Open letter to Jon Heyman (re: Bert Blyleven, HoF)

Dear Jon,

I read with interest your explanation of why Bert Blyleven is unworthy of induction into Baseball's Hall of Fame. I understand that you have your reasons, and I appreciate that you are sharing them with the rest of us. I even understand your reasons, but there are a couple of them that I find objectionable.

I know that you've heard the Blyleven arguments. I'm not going to repeat them, as I can't top what Joe Posnanski, for example, has done. There are some who are unwilling to listen, and you're obviously not of that group, and that's a good thing.I'm even confident, based on your position on Tim Raines, that you've listened. I don't know that I can offer anything that you haven't heard before. Certainly, I don't have any new statistics, new numbers to make Blyleven's case. Frankly, I think that the numbers are overwhelming, but you don't, and that's your prerogative. But I do want to offer some perspective on a couple of your objections.

You start your demurrer by saying that
My contention regarding Blyleven is that almost no one viewed him as a Hall of Famer during his playing career, and that is borne out by the 17 percent of the vote he received in his first year of eligibility in 1998, followed by 14 percent the next year. Blyleven obviously had an excellent and extremely lengthy career that looks a lot better to many with a decade to review it. And it doesn't hurt that he's the favorite of the Internet lobby.
That's a fair point. Indeed, for some definitions of "Hall of Fame" it would be dispositive, given that we're talking about a Hall of Fame. But I think it is, perhaps, a bit short-sighted if one views the Hall as an honor to be bestowed upon the best baseball players as opposed to just an acknowledgement of fame. There is a reason that there's a five year wait and a fifteen year balloting process. That gives one a chance to develop some perspective, and presumably, that perspective would be on the actual performance of the player, not just the contemporaneous perception of his performance. If we allow that to be dispositive, there's no point in waiting until the career ends, never mind five years afterwards.

So I don't think that "he didn't seem like it at the time" is an adequate reason. What else is there?
I look at numbers, too, and while my numbers may be slightly more simplistic than WHIP, WAR or VORP, I think they tell a story of a pitcher who was extremely good, consistent and durable but not quite Cooperstown-worthy. Blyleven was dominant in a lot of at-bats (thus, the 3,701 strikeouts) and even a lot of games (60 shutouts). But he was never dominant for a decade, a half decade or even a full season.
Really? In 1973, he led the AL in shutouts, strikeout-to-walk ratio and ERA+, while pitching the 4th most innings. That's not a dominant full season? From 1971-1975, he was third in strikeouts (behind Ryan and Seaver), 6th in ERA, and 7th in innings pitched. That's not a dominant half decade? From 1971-1980, he was second in shutouts, fourth in strikeouts, and tied for fifth in ERA while being 4th in innings pitched. That's not a dominant decade? Really, it would be hard to make a case that Blyleven wasn't the dominant pitcher of that particular 10-year span.

Unless you excessively penalize him for whatever his teammates did or didn't do.
Only four times in 22 seasons did he receive Cy Young votes (he was third twice, fourth and seventh once), only twice did he make the All-Star team and only twice did he win more than 17 games. I tend not to vote for players who I see as great compilers rather than great players, which is why I don't see Lee Smith or Baines as Hall of Famers, either. Baines and Blyleven compiled similarly in some key areas, with Blyleven finishing with four percent short of 300 victories at 287, and Baines four percent short of 3,000 hits with 2,866. And actually, a case could be made that Baines had more greatness, as he made six All-Star teams, three times the number of Blyleven.

Some will say that Blyleven's career was equal to Hall of Famer Don Sutton's but I say it is just short of Sutton's. They both had big totals in other categories but Sutton wound up with 37 more victories, going over the magic 300 mark by 24.


Some will say Blyleven was handicapped by playing for a string of horrific teams. But his many teams combined for a record of slightly over .500. For the most part, they were mediocre. While his career mark of 287-250 is clearly better than his teams' overall record, it isn't that much better.
This the stuff that I'd really like to address. In those excerpts, you list four different criteria that you're looking at for evaluating Blyleven's career.
  1. Cy Young votes
  2. All-Star appearances
  3. Wins
  4. Winning percentage vs. team winning percentage
Here's the problem - none of those is a measure of pitching performance. Each one of them is a proxy. Proxies can be fine, and sometimes, they're the best that we can do. There are aren't any temperature monitoring stations that have been in constant operation for 100 years, so we use proxies like tree rings and ice core samples to attempt to construct a temperature record. But we have an enormous amount of information available as to how well Blyleven performed as a pitcher. And despite that fact, you're focusing on proxy values that add tremendous amounts of noise and variation to the performance information.

Those proxies are all flawed. It is obvious that Wins and Losses, as important as they are for teams, make a poor proxy for pitching performance. You said
Many stat people suggest wins are not important in evaluating careers. But until wins don't decide who's in the playoffs and who's out, who makes the World Series and who doesn't, I will continue to view them as important
but I suspect that you recognize that as the strawman that it is. No one has ever suggested that wins and losses don't matter, only that they represent a poor, or at the very least an overrated and imprecise, way to evaluate pitchers.
A pitcher's goal for each game is to win the game, not to strikeout the most batters.
A pitcher's goal may be to win the game (how does that differ, exactly, from the left-fielder's goal or the shortstop's goal?) but he only has influence over one half of the performance that determines that outcome. And even in that half, there's a tremendous amount of influence that falls to the defenders behind him.

As you are well aware, a pitcher can pitch well and lose; he can pitch poorly and win. As a general rule, good pitchers will win more games than bad ones, and do it with better winning percentages, but there's an enormous amount of "noise" in that statistic.

As for the "winning percentage vs. team winning percentage," I'm sure that you can see the flaw there. The one good pitcher on a good offensive team with lousy pitching will tend to greatly exceed his team's winning percentage. If the team isn't good because the offense is lousy, that's not the case. We're talking about multiple levels of error in the proxy value, of course.

So you're starting with a flawed proxy, and adding to it more proxies which are dependent on it. Cy Young votes and All Star appearances are not only subjective measures, they tend, particularly during the time when Blyleven performed, to be based largely on Wins and Winning Percentage. If a pitcher ends up with relatively fewer wins than his performance would dictate because of the vagaries of defensive and offensive support, you're going to penalize him for it three times.

Harvey Haddix took a loss in a game in which he allowed no base-runners for 12 innings. Many pitchers have lost games without allowing any hits. As late as 1990, Bob Welch won a Cy Young award, despite a vastly superior pitching performance, on the strength of tremendous run support from the team behind him. There was nothing that pitchers control that Roger Clemens didn't do far better than Bob Welch, but Welch's teammates scored a ton of runs every time he pitched, and he was credited with nine more wins than Clemens, and he won the Cy Young because of it. Isn't it obvious that there are serious flaws in using that to determine a Hall of Fame vote? If the sportswriters screwed up in 1974 and 1977, that should count against Blyleven in the Hall of Fame balloting? Again, sometimes we have to resort to flawed proxies, but we don't have to here, because we know how Blyleven pitched.

Bert Blyleven allowed three runs or fewer in a game 422 times. He was award a win in just under 60% of those games. If he'd had the defensive/offensive/bullpen support to get the win in 63% of those, as Catfish Hunter did, he'd have won 13 more games, finishing at 300 even. Without changing his actual pitching performance one iota. Would he be a Hall of Famer then?

He didn't get enough wins, he didn't get enough Cy Young votes (because he didn't get enough wins), he didn't make enough All Star teams (because he didn't get enough wins) and he didn't exceed his team's won-loss record by a large enough margin (because he didn't get enough wins.) Basically, your argument boils down to "he wasn't awarded enough wins."

And you're smart enough to realize that that's a lame argument.

Labels: , ,



Post a Comment


<< Home

Links to this post

Links to this post:

Create a Link