It seems like the primary philosophical question being discussed is the relationship between wins and budgets.
What's that? A linear regression? And it's trapped down a well? Where boy where?
On the Basis of this last year alone:There is definitely a causal relationship between budget and wins. The correlation coefficient between 2053's team budgets and their wins is 0.6, which is to say, about 36% of team wins can be predicted by looking at the team's budget (and the rest is skill, rules that help poor teams and random noise). This isn't a bad correlation at all.
Top 5 Budgets and their wins:
1. New York Yankees, 108 wins
2. Cincinnati Reds, 106 wins
3. Washington Nationals, 97 wins
4. Colorado Rockies, 89 wins
5. Oakland Athletics, 60 wins
Bottom 5 Budgets and their wins:
1. Montreal Expos, 65 wins
2. Toronto Blue Jays, 57 wins
3. Miami Marlins, 92 wins
4. Milwaukee Brewers, 88 wins
5. Los Angeles Angels, 82 wins
There's clearly a relationship between the two; it definitely helps to have a higher budget. According to the regression, about $5.4M of budget buys an extra win (or an absence of $5.4M costs a win). There is definitely a relationship here. But of course, one of the five top budgets didn't have a winning record, and 3 of the 5 bottom budgets did. That's a bit of a fluke from the data, but it should make clear that budgets don't equal wins, even if they do have a lot to do with them.
But what about another causal relationship, that of Wins -> Budget Changes?
It turns out the relationship between 2053's wins and the budget change going into 2054 has a correlation coefficient of 0.7, which means about 50% of budget changes can be predicted by wins.
Top 5 Wins with Budget Changes:
1. San Diego Padres, +$22M
2. New York Yankees, +$8M
3. Texas Rangers, +$12M
4. Cincinnati Reds, +$14M
5. Chicago White Sox, +$10M
Bottom 10 Wins with Budget Changes:
1. Toronto Blue Jays, -$2M
2. Seattle Mariners, -$8M
3. Oakland Athletics, -$36M
4. Boston Red Sox, -$8M
5. Pittsburgh Pirates, -$26M
The pattern, both visually and in terms of correlation coefficients is clear; winning has more to do with your budget bumps than your budget has to do with winning. On average one win above average is worth $0.55M or so.
"But wait!" You may ask. "Having a higher budget means more wins, and more wins = more budget. So does this really say anything?"
Now we get to our third relationship. That of wins above expectation and budget change.
Because remember, we can loosely predict how many games a team should win based on their budget. It turns out that the number of games won over expected has a correlation coefficient of 0.814 to budget changes, which means that about 2/3 of predicting budget changes is purely driven by performing above expectations relative to your budget.
Top 6 Wins over Expectation With Budget Changes:
1. Texas Rangers, +$10M
2. Miami Marlins, +$10M
3. Chicago White Sox, +$10M
4. Kansas City Royals, +$28M
5. Milwaukee Brewers, +0
6. San Diego Padres, +$22M
Bottom 6 Wins over Expectation With Budget Changes:
1. Oakland Athletics, -$36M
2. Pittsburgh Pirates, -$26M
3. Baltimore Orioles, -$6M
4. Boston Red Sox, -$8M
5. Seattle Mariners, -$8M
6. Arizona Diamondbacks, -$8M
I want to be clear; the relationship here is crazy spot on. No team that performed above expectations (relative to budget) lost money from their budget and most of them gained budget (I'm not bitter). And every single team that performed more than 3 wins below expectations lost budget. One win above/below expectation will, on average, be worth $0.8M in the appropriate direction.
To recap (based on this year's financial data):
Your Budget this season predicts your wins with 36% accuracy
Your wins this season predict next year's budget change with 49% accuracy
Your wins relative to expectation predict next year's budget change with 66% accuracy
Translation? Higher budgets help teams win. But winning helps budgets more than budgets help winning. So tools that help poor teams get better (like more draft picks, cheaper players, whatever) may actually have a greater effect on winning (and, believe it or not, budgets in the long term) than changes to the budget itself. And that higher budgets create implicit winning expectations; if you have a high budget and can't win the number of games that your budget expects you will be punished, often severely. And if you have a low budget but perform above what you should be able to do, you get rewarded (look at the Marlins). Having a big market helps a ton. But a huge determiner of higher budgets is *outperforming your budget*. A team that always manages to win 5-10 more games than their budget predicts (whatever that budget happens to be) will generally see their budget jump $4-8M every year.
To be clear, I don't intend for the above to be slanted for or against any policy proposal.
It's just one season of data. But you know me. I'd rather have data to talk about than speculate.
Attached is the Spreadsheet, showing the Budget change each team got (in millions) and their wins above expected. I just thought the visual would paint a pretty good picture.
Attachments: