Reading Into Rostersby Bryan Smith
April 21, 2004
While leafing through my 1984 Baseball Abstract, a chapter entitled “Where Does Talent Come From” caught my eye. Bill James started the article off with this set of questions:
Where does a team’s talent come from? Which major league organization has produced the most talent? Which has produced the least talent? Which teams are built basically by trade, and which are basically built through the farm system? Are the most successful teams the teams which are built through the system?
In the ensuing work, James would assign points to each player using what he called the “Value Approximation Method.” Using these points, he tallied up where each player came from originally, how old he was, and how he got to his current team. Players were worth different values, as James sought out to answer the aforementioned questions.
Normally, this is a study to do at the end of a season. The Value Approximation method could be scrapped and one could just use Win Shares, as both statistics do the same inherent thing: universalize players. But, I didn’t want to wait that long, I wanted to redo the study now!
So, I started with this season’s Opening Day rosters. Rather than assign each player a different value, I assumed everyone to be equal. Yes there are flaws with this, Keith Osik credits the Pittsburgh Pirates as much as Barry Bonds does, but we’re too early in the year to attempt to differentiate players. Basically, I took the Opening Day rosters and did research in three areas: age, how they got to their current team, and where they originally came from. This week I’ll deal with the 350 players on Opening Day American League rosters, and next week we’ll look at the National League.
First, let’s deal with the age factor. There are four divisions that I separated the players into: young, prime, past-prime, and old. By definition:
Young: Age 25 and under
Prime: Ages 26 to 29
Past-Prime: Ages 30 to 34
Old: Age 35 and over
Here are the results, separated by division...
TEAM Y P PP O Baltimore 6 10 4 5 Boston 1 5 15 4 New York 1 4 10 10 Tampa Bay 4 10 8 3 Toronto 2 12 9 2Do you notice a noticeable difference between the unanimous top two teams here (Boston and New York), and the rest of the division? New York and Boston have 20 and 19 players respectively that fall in either the past-prime or old categories, while the next closest team stands at 11. How will this work for the AL Central?
TEAM Y P PP O Chicago 5 11 5 4 Cleveland 6 13 5 1 Detroit 5 10 8 2 Kansas City 3 8 10 4 Minnesota 8 8 8 1So once again, the team that I picked to win this division (Kansas City), leads the division with 14 over-30 players. Interestingly enough, the next best team is Detroit, a club with enough early season dominance to spark a Hardball Times article. While I’m hardly buying into the Detroit Tigers as a second-place team, it’s interesting they rank second in average age in the division. The AL West went as I expected, but didn’t exactly go to the rule we seem to be establishing:
TEAM Y P PP O Anaheim 2 13 9 1 Oakland 1 12 10 2 Seattle 3 4 11 7 Texas 7 9 6 3As I thought, the Mariners, with all their post-40 players, rank first in this division. My choice to win the Series, Oakland, comes in second with 12. Seattle’s early season problems have highlighted their problem, age. While I’ve been trying to establish here that the older teams are generally considered the favorites, Seattle just happens to be too old.
Next, let’s examine how these players got to their current clubs. Is it better to buy a team, trade for it, or to build it from the ground up? I have classified each player as coming to his team by trade (T), through the system (S), as a free agent (FA), or through waivers (W). To clarify, if a player was drafted in the Rule V draft, he is placed under the trade category. OK, let’s see how the AL West differs:
TEAM T S FA W Anaheim 3 12 8 2 Oakland 11 8 4 2 Seattle 7 10 8 0 Texas 5 9 11 0While Pat Gillick got a lot of bad press for not stirring the trade waters, it has been Bill Stoneman and John Hart who have been the least active, hardly using trades to establish their current teams. Anaheim has had the luck of developing a lot of players, while Texas is still "profiting" from their 2001 signing binge. Oakland, as we expected from Billy Beane, is built mostly through trades. Generally, either the A’s or Angels are picked in this division, and the two teams were built very differently, so there is no telling number here. Let’s see how the AL Central looks:
TEAM T S FA W Chicago 11 9 5 0 Cleveland 10 7 7 1 Detroit 11 5 8 1 Kansas City 5 6 14 0 Minnesota 7 14 4 0Once again, dramatically big differences across the board. The White Sox, led by Ken Williams, are almost half made up of players that were traded for. Kansas City, considered by some to be a favorite, was developed mostly by cheap free agent signings by Allain Baird. Finally, Minnesota was, more than any other team, built from the ground up. They are becoming the model for what a rebuilding team should follow, and adding a third straight division title will only help that idea. But once again in this division, there doesn’t appear to be a strong correlation with being a favorite and being developed mostly by one of these categories. So how do the big boys out east stack up?
TEAM T S FA W Baltimore 7 7 11 0 Boston 11 0 13 1 New York 8 7 10 0 Tampa Bay 5 6 13 1 Toronto 4 9 9 3This division is mostly made up of spenders, so it’s no surprise to see the double digits in the free agent column, but what shocked me was Boston’s big zero in players developed within their system. Theo Epstein has just recently developed a “Moneyball”-like regime with the Red Sox, so we’ll see how that effects the farm system. The Yankees have a solid corps of seven players that have come up through the farm system, carefully choosing what players to keep and which ones to deal.
So far, our two categories haven’t been too telling. We found that old age generally correlates with success, but as in the Mariners case, too much of anything can be a bad thing. And as far as this past category goes, we haven’t discovered any correlation, yet it still speaks a lot about General Managers and the organizations they run.
So to complete our "study," it’s time to look at what organizations have developed the most Major Leaguers. This is how a player becomes credited to an organization:
A player is considered to be produced by the first organization of which he is a member in which he 1) reaches the major league level or the Triple-A level or 2) is included in a major league transaction, or 3) goes directly onto a major league roster.
I considered Rule V moves to be "major league transactions," so Johan Santana is credited to be an Astro originally, not a Twin (sorry, Aaron). To not show AL dominance, I eliminated all players to have come through the system, and just included the 231 American League players to have come from a different franchise. In order, this is “where the talent came from”:
TEAM # 1 Los Angeles 14 2 Montreal 13 2 Seattle 13 2 Toronto 13 5 Pittsburgh 11 5 Texas 11 7 Oakland 10 8 Boston 9 8 Chicago Cubs 9 8 Cleveland 9 11 Atlanta 8 11 Cincinnati 8 11 Detroit 8 11 Florida 8 11 Minnesota 8 16 Baltimore 7 16 Houston 7 16 San Diego 7 19 Colorado 6 19 NY Mets 6 19 San Francisco 6 19 St. Louis 6 23 Anaheim 5 23 Kansas City 5 23 Milwaukee 5 23 Philadelphia 5 27 NY Yankees 4 27 Chicago Sox 4 29 Tampa Bay 3 29 Arizona 3I’ll detail this issue more when we get the final numbers next week, but as Bill James concluded in 1984, I still see no correlation between developing players and succeeding in Major League Baseball. This may just prove the theory that a General Manager must outwit or outspend his opponents to win.
We’ll dive more into this next week, but it’s been interesting to dive into Opening Day Rosters to see what they really mean. Bill James, 20 years later, still worth the read.
Bryan Smith, co-founder of Baseball Analysts, is a freelance writer with work appearing at SI.com, BaseballProspectus.com and Baseball America. Feel free to e-mail Bryan here, and look for his annual prospect list at SI.com next week.