Winning causes payroll: study
Suppose you're a Martian who has just immigrated to North America, and you have no idea how baseball works. All you've got is a database full of statistics, and a black-box graph theory algorithm to try to figure out cause-and-effect relationships. What would you conclude?
The answer can be found in a new JQAS paper, "Dependence Relationships between On Field Performance, Wins, and Payroll in Major League Baseball."
The author applies the algorithm to a bunch of stats, and comes up with a bunch of dependencies for what causes what. Most of them don't make a lot of sense. For instance:
-- Walks depend on: OBP, Runs, and SO.
-- Total bases depend on: AB, 2B, HR, IP, and Runs.
-- Earned runs depend on: ERA, hits against, and HR against.
... and so on.
Be that as it may, the real purpose of the paper is to look at payroll and wins. What does the graph theory algorithm say about those? This:
--Winning percentage depends on: Fielding percentage, On-base percentage, and Saves.
--Payroll depends on: Fielding percentage, Pitcher strikeouts, and Winning percentage.
In any case, after that, the author looks at how an increase in payroll or wins affects the future. He finds:
-- If you bump a team's payroll 10%, it wins an extra 2.5 games this year, but returns to normal afterwards.
-- If you bump a team's payroll 10%, payroll drops slowly from +10% to +2% over the next 10 years.
-- If you bump a team's winning percentage by 10%, it returns to baseline in subsequent years.
These are no big deal. However:
-- If you bump a team's winning percentage by 10%, it bumps the team's payroll by 10% immediately. Then payroll rises to +25% over the next three years, settling back down to +10% by year 10.
So, according to the algorithm, payroll doesn't seem to have long-lasting effects on winning. But winning appears to have long-lasting effects on payroll! Therefore:
"... while we found some evidence that winning affects payroll and payroll affests winning, the evidence suggests the effect of winning on payroll is the more direct, larger, and more lasting in magnitude one."
In summary: winning causes payroll.
That's what the black-box algorithm says. But, to any member of the Martian-American community who may be reading this, I would respectfully suggest: you're better off not putting too much faith in the results of this particular paper.