This is a very important tool for measuring the overall competence of a team's front office.

"The median was slightly higher due to the fact that average was skewed by the New York Yankees' record payroll of $218.3M.

Outliers don't skew the median, they skew the mean. If anything it seems like it's the PIT, WAS, FLA, TB teams at the bottom (all ~.5 or less of the mean) that are skewing the mean more than NY (the only team that's at 2x the mean).

The median was slightly higher [than the average] due to the fact that average was skewed by the New York Yankees' record payroll of $218.3M.

I don't understand this. The Yankees skewed the average upward but not by that much. If they had cut their payroll to the Red Sox level, the MLB average would have dropped by about $2 million but the median wouldn't have changed at all.

I realize that "ouliers don't skew the median" and, in fact, said the *average* (or mean) was skewed by the Yankees, not the median. That is a true statement. However, given the fact that the median is above the mean, your point is well taken that the teams below the median are negatively impacting the mean more than those above the median.

I have changed the language in that sentence to set the record straight. Thanks.

Rich,

When measuring the payroll efficiency in MLB wouldn't an objective study note that there is NO SALARY CAP. Therefore, Steinbrenner/Cashman & Company are PLAYING "BUY" THE RULES!!!!

Regards,

Hank & Hal

Excusing the Red Sox is a little too results-oriented for my liking. Sure they won it, but what share of the odds did they have at winning it? I'm tempted to suggest a graph of payroll versus run differential to give a better idea of who had the best chance at winning it all... but I've read that Arizona actually did deserve some of their supposed "luck" in wildly outperforming their expectation there.

Someone on the Hardball Times, I think David Gassko, did a nice writeup of the increasing value of wins over 81, and their decreasing value at some point (for example, if you have a 115-win team, is it really worth $30 million to sign Santana for a season, or do you just stand pat?). That might be the very bestest method to analyze payroll compared to expected return.

Also, I'm torn on flipping the y-axis to go from $250 "up" to $0, as that would make "up and to the right" data points the best ones, as we're used to seeing :-) Anyways very cool, I'm not being picky but am just thinking outloud.

Rich,
Do these numbers include what the Red Sox paid to negotiate with Dice K in a yearly rate added onto his contract? While I doubt they do, it certainly should as that money went to player payroll.

Mike,
I had totally the opposite reaction you did. When I started reading the article, I thought this is good stuff, but the point of the whole deal is to hold the trophy at the end then the article had this:

"Due to the fact that it's the goal of all teams to win the World Series, I'm going to excuse Boston from any list of inefficient clubs. Yes, the Red Sox paid up for their success, but it's hard to argue with the fact that they won it all. While Boston may not have been the most efficient in terms of regular-season wins vs. payroll, John Henry, Theo Epstein & Co. were clearly the most efficient in terms of winning World Championships – especially in view of the competition within the division."

And I thought, fair enough. Sure they spent alot, but they give themselves a chance of winning it all every year, and as a fan I can't ask for more then that.

God, I love this website.

Also, I really like Rob Neyer's suggestion of a multiple season format to assess a franchise's efficiency over time.

Thanks

Purpura was actually fired in August. The Astros' season was over at that time but they did play the rest of the schedule as required by the league.

Yea, the Red Sox payroll doesn't include the $50 million to negotiate with Dice-K, but it should. Their payroll wouldn't look like such a 'bargain' in that case.

One correction which I'm sure you aware of:

Both the Phillies and the Angels lost in the division series of their respective leagues, not in the championship series.

I think it would be interesting to see the chart using player salaries as a percent of team revenue, rather than the raw dollar amount. Without a salary cap, teams are only limited by their budgets - I think it would be interesting to see who maximized their budgets (e.g., if the Yankees' revenue is $400 million and Tampa Bay's is $50 million, the Yankees actually spent less as a % of team revenue and thus did a better job of maximizing their budget).

IMO, the posting fees should be included for both Dice-K and Igawa (ugh), which would drive up the expenditures of both the Red Sox and Yankees.

Wonderful tool, Rich. The one thing that is bothering me of late is that many analytical studies, whether for individuals or for teams, do not consider the post-season.

Major league teams, especially at the high end, are becoming more and more cognizant of what can win in the post-season and play the regular season accordingly. The Red Sox clearly throttled down the regular season in order to be ready for the post-season. The Yankees have done this for years, especially with Rivera. Although the Red Sox and Indians both won 96 games, I think both teams would admit that they achieved a different level of success, and that the difference is not "luck".

I am not sure exactly how to make this adjustment, or even whethere is appropriate in this case. Just wanted to make the point that the goal of most good major league teams is not to win the highest number of regular season games--it is to win what you need to win, and then kick ass in October.

Thanks, Mark. I agree with you that the analytical community generally underplays the importance of the postseason. At a minimum, players should be given credit for their stats and teams for their wins.

In the case of Boston, I tried to do so qualitatively by giving the organization its proper respect for being "the most efficient in terms of winning World Championships." There isn't a team in baseball that wouldn't trade seasons with the Red Sox.

As for Matsuzaka's contract, I believe the posting fee of $51 million could be amortized over the six-year life of his contract if one really wanted to account for the total cost of his services. However, the $51M was not treated as salary by MLB and, as such, not subject to luxury taxes.

Right, but does that make the British guy who put his life savings on black in roulette (and won) the best investor of the year? Sure, there's no investor that wouldn't trade RESULTS with him, but if we did the spin all over again, would anyone trade their conservative mutual fund investment with his strategy?

I generated a graph where an arrow between team A and team B means: team A won more games than team B while having a lower payroll than B (but I didn't create arrows that can be obtained via transitivity, because they would make the graph look too busy).

Basically, you shouldn't look bad for spending more than other teams if you also manage to win more than them, and you shouldn't look good if you're penny-pinching while not winning more than others (looking at you, Rays and Marlins!). Such teams end up not showing up in my graph at all as a result (there is no arrow linking them to other teams, they're on an "island"). This way, a team like Boston doesn't look bad for getting the expected result for their spending, and a team like Tampa Bay or the Marlins doesn't look good for lucking into a few wins with a dismal payroll.

Enjoyed the writeup. What's helpful is to isolate FA spending including retaining homegrown guys beyond year 6. Payroll/wins really measures two things - teams' ability to field Ryan Howards for $750K and how effiecient they spend their FA dollars. Dave Studeman at THT was able to improve the coefficient to .77 by accounting for the impact of pre-FA players (http://tinyurl.com/25gn6n).