November 22, 2002
The Daily Prospectus
Balanced Lineups Redux
Soon after yesterday's installment of "Aim For The Head" appeared on the web site, my e-mail starting getting reader comments like the following:
Jonathan Bernstein wrote:
Fun piece, and an interesting finding, but I don't think you really answered the "how to build a team" question.
I'm sorry, but did you come anywhere near answering the question?
Keith, in regards to the question of whether balanced teams or lopsided teams are better...did you ever get around to answering it?
I don't think that answered the question.
Another reader wrote:
?Truncated? It looks like the article was chopped at the bottom.
When so many many readers respond so quickly with similar comments, it's clear that I didn't do a good job in writing the article.
Presenting the original question again:
Based solely on offense, expected runs created - given the scenario that your total starting lineup team OPS was fixed at a certain number. Would you be better off building a team with a few superstars, balanced off with some truly horrible players, or a team of mostly mediocre players?
The AFTH I wrote focused on the issue of lineup balance. Does a balanced lineup produce more runs than an imbalanced lineup of the same aggregate performance? In doing so, I chose to neglect the broader question about what the implications for team building are, which is ultimately the practical application of answering the question.
As for answering the balanced lineup question, the feedback in my inbox makes it clear that I didn't state the conclusion well. So let me try again: based on historical results, the "superstars & scrubs" teams scored slightly fewer runs than did "balanced" teams with the same OPS. However, the size of the effect was almost always less than 10 runs, and it doesn't make much practical sense to consider the run scoring efficiency of lineup balance as a factor in building a team.
As for the strategic team-building implications, the short answer is that it's probably better to focus on the "superstar & scrubs" approach despite the minor run-scoring inefficiency. Replacing scrubs is typically much cheaper than replacing average players. There is more available talent, and less tendency to be overpriced the way the middle of the market often is. Because there are only 25 (or 40) spots in the lineup, and a limited number of plate appearances to go around, there is value attached to each roster spot.
While having two 45 VORP players in the lineup may be slightly better than an 80 VORP MVP-caliber player and a 0 VORP AAA journeyman in those same spots, the latter team is easier to upgrade with even a small expenditure in resources. Finding a 15 VORP player for $1 million is relatively easy compared to pulling off a deal of all-stars to improve on a 45 VORP player. Being able to concentrate value in a limited number of roster spots opens up more room for overall improvement.
For those reasons, I prefer the imbalanced team approach. Get the best players you can to fill some of the holes on your team, taking your chances with the waiver wire for the other holes, rather than filling all of the holes with mere adequacy.
There are several other comments and questions on the topic of lineup balance that I may follow up on in a future AFTH, but given the overwhelming dissatisfaction with the conclusions (or lack of them), it seemed prudent to address those immediately. Thanks to everyone who wrote in.