Wednesday, May 21, 2008

To Graph or Not To Graph: That is Never a Question

I don't know why, but I have a strange fascination with these sector graphs. Maybe it's because they're graphs, who knows. In anycase, they inspired me to take a moment and put together this bad boy:

(click graph to enlarge)

It's pretty straightforward:

  • The average age of the team is plotted along the X axis
  • The number of games above/below 500 is plotted along the Y axis
  • The size of the circle corresponds to the team's total payroll
  • The color of the circle designates the league or division in which the team plays

After an admittedly brief look at the results, here are my - you guessed it - five takeaways:

      1. Rich doesn't mean good. A lazy eyeballing of the graph doesn't reveal any obvious pattern of bigger circles above the 0.500 line or smaller ones below it. While the free-spending Red Sox, Angels, and Cubs are all safely above water, similarly sized behemoths representing Detroit, Seattle, and of course the Yankees are significantly submerged.

      2. Young does mean cheap. Each of the 11 smallest circles are bunched safely to the left of the centerline (Florida, Tampa Bay, Oakland, Pittsburgh, Washington, Kansas City, Minnesota, Arizona, Baltimore, Texas, and Colorado). Similarly, the 5 biggest circles are all to the right (NY Yankees, Detroit, NY Mets, Boston, and Chicago Sox). The biggest exception to this rule is the Angels, who have the 6th biggest payroll yet the 11th youngest team.

      3. By some bastard child of the transitive property, there is little correlation between age and ability to win. Both the 5 winningest teams (Arizona, Cubs, Boston, St Louis, the Angels, and Tampa Bay) and the 5 losingest (San Diego, San Francisco, Seattle, Colorado, and Detroit...boy, the NL West sucks) are scattered broadly across the age axis.

      4. The Marlins and Rays know something we don't. See that teeny tiny dot waaay off to the left? That's Florida. Despite the oft-noted observation that their entire payroll is smaller than A-Rod's (and less than half the payroll of the 2nd cheapest team), and the fact that they're far and away the youngest team, they still boast the 3rd best record in the NL even after a recent 2-5 slide. And Tampa Bay, with both the 2nd cheapest and 2nd youngest team, are winning at an even greater clip.

      5. The Dodgers are slightly above average in all three dimensions. This is indicated by their position just offset to the upper right of center. Of the 30 MLB teams, they are the 13th oldest, 11th winningest, and 8th richest.

While none of these observations is revolutionary, a properly- conceived graph can tell the story like nothing else can. And yes, it's still early - it'll be interesting (at least to me) to see how this graph looks at the end of the season.

Thanks for reading. Please share any thoughts or questions.


Xeifrank said...

Nice graph Mr Karros. Keep up the good work.
vr, Xeifrank

Dean H. said...


What would be interesting to see would be a graph showing the correlation between average and pythagorean expectation, or maybe the differential between expected and actual wins. We hear a lot about how guys like Nomar and Juan Pierre "know how to win". Well, that would show up in the data. If older teams outperform their expectations based Runs Scored and Runs Allowed, then experience clearly matters. If don't, then ...

Dusto_Magnifico said...

How many of these teams are spending money just to spend money. "We have 45 million dollars to play around with this offseason. Lets spend it all and get overpriced underperforming veterans locked up for the next 5 years and ignore our young kids who cost nothihng for two more years."

I'd rather sign Loney, Kemp, Martin, Bills, Ethier, and Broxton to long term deals before we get another crappy veteran free agent. I know the fans wanted McCourt to get to the 100 million dollar mark to improve the team and at the time we needed it. Our farm system wasn't as deep as it is now.

Unfortunately, McCourt is spending money at an alarming rate, no thanks to Colletti, on players that are overpriced and overrated.

It doesn't seem like Colletti actually has a plan on how to win. He just signs the best apparent player to fill holes. Which I don't mind, but he doesn't consider stats, age, injuryness (?) etc.

Of all the centerfielder to sign He gets Juan Pierre? And then tries to cover up his mistake by getting Andruw Jones?! Both signings biting back.

It gets worse. He also gets Nomar, Schmidt, Gonzo, and who next year. Is Colletti actually trying to build a team that wins? At what point in time do you give up on him?

Orel said...

New maturity has Marlins winning, getting along swimmingly

The Bizarro Supermen (Rays)

Steve Sax said...

EK, could you make the circles pulsate, like they do on espn gameday? Thanks.

Eric Karros said...

Dean: You're in luck, as I do consider requests. I'm no math guy, but I think I know what you're looking for. I'll give it some thought, but no promises - I may end up choosing to instead do another post about the Spice Girls.

Eric Karros said...

Actually Dean, that sounds like an analysis most valuable if done at season's end (see? I really did give it some thought). I do like it though. I guess I could do it for previous years, but that's so yesterday. Remind me around Sept and I'll put it together.

One point of reference though is remember last year the hubbub around the Dbacks winning the West in spite of being outscored over the course of the season? And I believe they tended towards the younger side. So that might be a point against that hypothesis.