August 14, 2003
Mark Armour & Dan Levitt, Part I
Mark Armour and Dan Levitt turned a series of debates between friends into Paths To Glory, the stories of the building of several interesting baseball teams, published this spring by Brasseys, Inc. Paths looks at some of the greatest dynasties in baseball history, including the A's of the early 1970s, as well as some of the most dramatic missed opportunities, including the Black Sox of the 1910s and early 20s. Armour and Levitt recently chatted with Baseball Prospectus about some of those teams, the men who built them, and the team-building strategies they used to make it happen.
Baseball Prospectus: What first inspired you to write this book?
Dan Levitt: Mark and I had become friends through SABR. Pretty soon we came to realize we were interested in the same thing: What made baseball teams win, and what didn't? We started corresponding by telephone and e-mail, and started doing research as well. We came to realize that not there weren't a whole lot of printed materials on the subject. There were a lot of books on great teams out there. But those books would rank them, or give day-to-day on seasons, stars, and the not-so-great players. But you'd have a really tough time finding a book on say, the 1915 Phillies, and how they won the pennant. We realized there was a void of this type of book. So we set out to put a book together that'd be interesting, about some great teams and some not-so-great ones. It's really a different perspective: not the day-by-day heroics, but rather what were the decisions that went into assembling these teams.
We also wanted to look at what happened to these teams after they were great, what caused their downfall. The '65 Twins for example won the pennant, so why couldn't they win a couple more? In a broader sense, it's asking how did these teams get built, looking at the concept of team building from multiple angles--everything from making the right trades to building rosters to the evolution of reliever usage.
Baseball Prospectus: What can a reader from the teams that were near misses?
Levitt: To understand why some teams win and why some don't, near misses can be just as interesting as the great teams. Why couldn't they get over the hump? You think of a team like the Expos. They were widely acclaimed to be the team of the 80s. I know at the time I felt they'd at least get to the World Series. So what went wrong? The late-40s Red Sox became known as a team with no pitching and no defense--was that really the case? The early 70s Angels had a lot of personality issues, and they also weren't as good as people thought they'd be. The early 80s Expos had serious drug issues. We were always looking for interesting subplots.
Mark Armour: ...and I think that sort of gets to another question: Why did we choose the teams we chose? One thing we tried to avoid is telling a story that we thought had already been well told. Or we'd have a different take on a team, like the late-40s Red Sox. People have asked us, 'why didn't you cover '55 Dodgers?' Because that story's been told very well, and we didn't want to rehash old stories. When you talk about the near misses though, one thing that comes through is that this is a GM's book, as opposed to a manager's book. At some level, as much as people don't like this answer, there's a lot of luck involved. A lot of a GM's job is to get your team into contention. And a lot can go wrong at that point. You can get a great player that has a bad year, injuries, or a hundred other problems that make the difference.
BP: Reading the book, you can see that there are a lot of sidebars and tangents that come up as each chapter progresses. Did you plan the writing out in advance that way, or did it spontaneously go in that direction?
Armour: One thing I always really liked about reading Bill James is the way he'd take literary license. It was somewhat cocky really, saying 'it's my book, I'll do what I want with it.' I'd never really read something done that way before. He'd take the subject seriously, but in the end it's still a baseball book--it's supposed to be fun. When you're writing about the Civil War, you can't dive off into tangents. But if you have a conversation with someone about baseball, this is sort of the way you tend to meander around. You're allowed to change the subject.
Levitt: Plus the things we meander around are topics central to the story. The '49 Red Sox: why is the perception today so different than it was then? Vern Stephens is a microcosm of that, so we talk about Vern Stephens. The Boston Braves: Billy Southworth was really the central figure there, the way he was brought in to help build the team, then after his drinking and personal problems the team fell apart. With the White Sox, Charles Comiskey was the central character; he was second to Ban Johnson in the American League's power structure, and he was also willing to spend his money to build a winner. Same thing with Charlie Finley and the A's or Alex Johnson and the Angels; the guys we meander off on are central characters in that story.
BP: Obviously the team essays dominate the book. But there are also essays on relief pitcher usage, player growth patterns and other topics. that seem to evolve out of team-specific seasons and events. When you first sketched out the book, were these microstudies planned in that form all along?
Armour: We mostly just wanted to look at things that hadn't been covered that much. Although we attempted to draw large themes from all of these stories, one of the main themes you draw is that the rules and context of the game are constantly changing. Lessons from the 1899 Brooklyn Superbas are only marginally relevant today. And one of the the largest ways to look at that difference is asking how a GM would go about building a bullpen. A hundred years ago, he didn't. Fifty years ago, he'd get one good guy. Today, it's seven or eight guys. That's the largest example of how rosters have evolved, and we wanted to examine the effects of that evolution, and if teams are building and using their bullpens in the most efficient way.
Levitt: Talking about player growth curves, looking at the minor leagues, player development, and how players age is a central theme to the story. We make value judgments throughout the book--this team had a bunch of players under 30, they should have had some growth left. This team was older, didn't recognize it, and paid the price for it. We felt we had something new to add to the discussion on aging, so we did some research and tried to find some answers. We needed to address systemically where, how, and why we were making some of these value judgments.
BP: How did you go about doing this?
Armour: The conclusions of the aging chapter are not that different from previous conclusions on the subject. People tend to be really surprised when their team trades for a guy who's 33 and he doesn't do well. The best GMs understand this, but a lot of them don't. They still seem surprised when they get guys at a certain age. The player develops differently than they thought, and they say 'it's not my fault this guy didn't have good year.'
Levitt: The genesis of this is based on a book called Predictions, written by a physicist in the early 90s, that's based on human growth curves. People grow quickly early on, then there's an inflection point start to level off and later start to peter out. You can fit this function to the growth of anything--everything from Brahms compositions to the number of movies Hitchcock has made. What I extrapolated from that was that you can do this with a ballplayer's ability. It's not an easy thing to do of course. If you take a guy who's 24, he might fall off the table at 26, or at 38; there's a lot of uncertainty involved. We worked with the growth curve and compared it to several projection systems out there. This does slightly different things than other methods, and it's another tool to use in terms of projecting players. The more data you have, the more likely you are to better predict performance.
BP: Sticking with some of the studies you did, you use Wins Above Replacement (WAR) level as the standard measure of performance throughout the book. Why WAR?
Levitt: I like statistics that are denominated in runs and wins. If you say that a player is worth two wins above what a replacement-level player's worth, other than maybe defining replacement-level, people understand what that means. That's a concept that's understandable to people who are not that familiar with baseball statistics. It's also accurate. We just felt that if we said the Braves did a good job picking up Jeff Heath, that he was a better solution than what they had the year before better than year before, we should be able to quantify how much better he really was. It's a way to back up what we're saying and give it some weight. We used WAR per 150 games basically because it was a nice round number, and it worked well for 162- per 154-game seasons.
Part II coming soon...