Wednesday, September 3, 2008

Overperforming/Underperforming Baseball Teams According to Pythagorean Expectation

In this blog, I have decided to look at which baseball teams records that reflect their overall performance, and which teams have records that don't.

Using the Pythagorean Expectation formula used by Bill James (win pct=RS^2/(RS^2+RA^2)), I determined the expected win percentage of all MLB teams. However, I replaced the coefficient of 2 with 1.81, which is supposedly more accurate. Understand that while the Pythagorean Expectation is quite accurate, it doesn't take into account good managing and clutch performances. Still, it is a useful and telling statistic for the most part. It could help to answer questions such as if the Rays are the real deal or not.
Here is the data:
AL East RS RA PCT P1.81 Diff
Tampa Bay 639 546 0.618 0.571 0.047
Boston 729 580 0.587 0.602 -0.02
NYY 671 630 0.536 0.528 0.008
Toronto 603 531 0.518 0.557 -0.04
Baltimore 704 739 0.457 0.478 -0.02
AL Central RS RA PCT P1.81 Diff
ChiSox 696 611 0.558 0.559 -0
Minnesota 703 623 0.558 0.554 0.004
Cleveland 669 629 0.489 0.528 -0.04
Detroit 695 703 0.478 0.495 -0.02
KC 562 686 0.423 0.411 0.012
AL West RS RA PCT P1.81 Diff
LA Angels 633 573 0.613 0.545 0.068
Texas 768 836 0.486 0.462 0.024
Oakland 533 574 0.457 0.467 -0.01
Seattle 581 690 0.391 0.423 -0.03
NL East RS RA PCT P1.81 Diff
NY Mets 684 604 0.561 0.556 0.005
Philly 669 576 0.547 0.567 -0.02
Florida 647 681 0.504 0.477 0.027
ATL 639 669 0.432 0.479 -0.05
Washington 535 686 0.381 0.389 -0.01
NL Central RS RA PCT P1.81 Diff
ChiCubs 750 566 0.612 0.625 -0.01
Milwaukee 663 577 0.58 0.563 0.017
St. Louis 673 621 0.54 0.536 0.004
Houston 624 653 0.525 0.479 0.046
Cincinnati 588 686 0.442 0.431 0.011
Pittsburgh 618 754 0.423 0.411 0.012
NL West RS RA PCT P1.81 Diff
Arizona 634 609 0.507 0.518 -0.01
LAD 578 568 0.496 0.508 -0.01
Colorado 656 708 0.471 0.466 0.005
SanFran 530 649 0.428 0.409 0.019
San Diego 533 649 0.384 0.412 -0.03

As you can see, most of the actual winning percentages are extremely close to the Pythagorean Expectation. However, a few things stick out to me.
1. LA Angels Overperforming--The team that differs the most from its actual winning percentage are the Angels. According to the theory, they are actually only the 5th best baseball team in the AL despite having the 2nd best record. The cause for this is probably a combination of a few things: great managing by Mike Scioscia, winning a lot of close games (K-Rod has had 59 save opportunities), timely hitting, and getting lucky. The Angels are a good team with playoff experience, but don't be surprised if they fall quickly in the postseason.

2. Rays the Real Deal, Kind of--Are the Devil Rays truly a playoff caliber team this year? Yes. Are they the best team in the AL (currently they own the best record)? No. Behind the Angels, they are the 2nd most overperforming squard in the MLB, and haven't been playing as well as the Red Sox (who are slightly underperforming). Unless David Price provides a major spark for the Rays this month, I can't see them getting past the much more experienced, underperforming Red Sox team.
3. Cubbies are Heavy Favorites in the NL--The Cubs are even slightly underperforming, while the Brewers are overperforming. Their Pythagorean (with 1.81 exponent) Win % is >0.06 pts ahead of the Brewers. This should be the year, but of course Bill James doesn't take curses into account.
Conclusion--Admittedly, this analysis didn't show anything really eye-opening, but I found it interesting nonetheless. The postseason is short, teams get hot, players get injured, and there is a lot of natural deviation, so it's impossible to have a 100% positive pick on which team will win the pennant or a series. Most of the time, the best team in the MLB won't win the World Series.

Natural deviation has never been as apparent as last year, when the Diamondbacks scored less runs than they allowed but had the best record in the NL and swept the cubs in the divisional round.

Just adding this in as an annoyed Yankees fan: The firing of Joe Torre was a ridiculously ignorant decision. Fact is, you can't completely evaluate how good a team or a manager is by the results in a short postseason. A manager could easily lose despite making smart, educated moves, and that is what happened to Joe Torre. There is too much deviation. Bringing a team to the playoffs that many years in a row is an amazing accomplishment. If you're going to fire somebody based on postseason performance, evaluate their decisions and their preparation, but not their results.

No comments: