clock menu more-arrow no yes

Filed under:

Bill James Interview

New, 1 comment

I was able to conduct an interview with Bill James, Red Sox consultant, and the man who's been called "The Guru of Baseball Statistical Analysis." James is a founding member of SABR and the writer of "Bill James' Historical Baseball Abstract." His influence on baseball has been great, as a lot of general managers and others follow his statistical analysis.

Over The Monster: What is exactly is your role within the Red Sox organization?

Bill James: I try to avoid definition. Saying what my role is and what it isn't is a way of self-limiting, and I've always tried to avoid that. I just do whatever I am asked to do.

OTM: How much of a say do you have with trades, signings and other transactions within the organization?

James: Well, I don't have any "say" in the sense that anything is my decision. Theo has decisions, the ownership has decisions, the scouting system has decisions, the minor league operation has decisions. I have opinions. When the guys who make decisions are in the process of making them, they sometimes ask for my opinion, and I give it. I try to figure out ways to be helpful.

OTM: What is your role during the amateur draft?

James: It changes from year to year. This year I spent a month or more prior to the draft studying the records of college players, and drawing up lists of how I would rank them, based strictly on their performance. Then I sat in on ten days of long, long, long meetings with the scouts, in which we reviewed information and ranked players.

I tried to limit my input, because these guys have much more experience in evaluating young players than I do, but I did participate, and I enjoyed it, and I felt that my participation was useful.

OTM: What is your opinion on how many pitchers a team should carry? Are 12 pitchers too many or just right? Are 10 pitchers too little?

James: Well, it depends on who you have, obviously. If you have starters who go seven or eight innings, that's one thing; if you have starters who go six, that's another. If you have lefties in your bullpen who can work either side of the plate, that's one thing; if you have pitchers who are vulnerable one way of the other, that's another thing.

When I was a kid, teams typically carried nine pitchers, sometimes eight, and I can remember teams going through parts of the season with seven. Now we debate eleven or twelve and occasionally wind up with thirteen. There are advantages both ways.

I have always thought that, if I was managing a table-top team, Strat-o-Matic or whatever, I would prefer to have more pinch hitters available. But real baseball is different from table-top baseball in at least two relevant ways. First, in table-top games, a pitcher with a 3.50 ERA has a 3.50 ERA, period. In real life a relief pitcher who might not be able to pitch effectively for three innings might be very effective throwing 15 or 20 pitches an outing, because he is able to focus, limit his repertoire, and maximize his energy.

Second, in table-top games the platoon differentials are all over the map, and the platoon differentials for hitters tend to be as large or larger than those for pitchers. In real life the platoon differentials for some pitchers are much larger than the platoon differentials of almost any hitter, so that it becomes more effective to work the platoon differential from the pitching side than from the hitting side.
That said, I'd probably prefer to carry fewer pitchers. But you've got to do what you've got to do.

OTM: When evaluating pitchers, which stat do you think is the most important to be aware of? Which stat do you tend to look at the most to judge if a pitcher is successful or not?

James: Well, I start with the strikeout to walk ratio. But knowing how well a pitcher has performed in the past is easy. What we're essentially asking is, "what is the likelihood that this pitcher will perform as well or better next year than he did last year?" There are three things which provide information about that, which are:
  1. His health,
  2. His "stuff", and
  3. Mis-matches or discrepancies in his data.
I'm not a scout, so I can't get too much out of the stuff. What I'm always looking for something in the data that is misaligned. If a pitcher has a good strikeout rate but has given up too many hits, maybe he's just been unlucky, and maybe you can bet on his luck to improve. Of if he has struck out 35 men in 80 innings but given up only 75 hits, probably he has just been lucky, and you should avoid him. Or if he has given up 85 hits and 40 walks in 80 innings but somehow has come out of it with a good ERA anyway, probably he can't sustain that, whereas if his other ERA components are better than his ERA, maybe he is due for a better year. Or if he has gone 17-11 but with a 4.80 ERA, maybe you shouldn't bet on him to sustain that. Or if he has a 5.20 ERA but it breaks down as 6.15 before August 1 and 2.80 after, maybe he has figured something out, or maybe he has just had a lucky streak, and you have to try to figure out which it is.

Or if a pitcher has only pitched 45 innings in 48 games and gave up 21 runs but 14 of them were in four bad outings, maybe he is better than his stats show. Or if he pitches in Colorado maybe he is better than his stats show, but if is the odd case where he has a 3.80 ERA in Colorado but 7.52 on the road, maybe he isn't better; maybe he is worse.

The reality is that we're not dealing with one of these problems; in a typical case we're dealing with 15 or 20 of them, so it is almost impossible to figure out what the true level of ability is behind all the moving screens in the statistics. So we're usually wrong.

OTM: When evaluating hitters, which stat do you think is the most important to be aware of? Which stat do you tend to look at the most to judge if a hitter is successful or not?

James: The trick is not to look at one stat, but to balance them all. I don't think there is anything in a hitter's record that you can ignore.

OTM: How largely have stats -- other than the normal batting average and ERA -- influenced front office personnel and scouts? Do front office
personnel and scouts tend to rely more on stats nowadays, or does their personal judgment of a player still play a large role? How have the Red Sox changed in that aspect since you joined the organization?

James: Major league executives have always relied very, very heavily on statistics. They don't do so any more now than they ever did.

There has never been a time in baseball history when 10-game winners were paid as much as 20-game winners. There has never been a time when .250 hitters were paid as much as .300 hitters. There has never been a time when outfielders who drove in 100 runs didn't make the All-Star team, or when outfielders who drove in 60 runs did. There has never been a time when you would trade a 30-homer guy for a 20-homer guy, unless there was something else in the deal.

The stats have always played a huge role in how players are evaluated, and a huge role in every decision. What has changed is two things. First, there are more stats around. And second, there is some change in the emphasis on different stats. On base percentage is more important than it used to be.