I decided to make my comments on the Game Averaged PythagenPat win estimator a regular, updated feature of DOV, complete with PythagenMatt for every Major League team.  I'll update this each night with the latest games played for each team included and leave comments open to discuss the strengths of teams as they rise and fall on the chart.  Here are the 30 teams in order of adjusted PythagenMatt W%

For those not familiar with PythagenMatt, I'll give the short explanation here and a link to the full explanation appears in comment four below.

We know that the biggest problem with ordinary seasonal Pythag is that is puts too much emphasis on a team's production in blowout games.  Generally, once the game starts to get out of hand, the losing side starts to send out the reserves, especially the reserve pitchers.  They in essence become a weaker team for the rest of that game.  This often leads to additional run scoring that has essentially no meaning (or very close to none).  I believe that the best way to cancel out the impact of blowout games is to put each game on a pythagorean scale.  If you win by 8 runs or by 12, you can still only win one game at a time, and the pythagorean difference between a 10-1 game and a 15-1 game is negligible.

PythagenMatt is PythagenPat, but applied to one game at a time and then summed and averaged (per game).

For example, if you win 14-3, the PythagenPat equation gives us an exponent of 17^0.285 or 2.24 and a winning percentage of 0.969.  Do this for every game and you get something that correlates much more strongly to actual winning percentage than seasonal pythag (I demonstrated a 4% improvement in R^2 in the article I linked in comment #4).  Doing just that gives you numbers that bias toward .500 from both sides (a center-pull) because you're by definition taking away some of the extremes, but the center-pull is easily remedied by applying an adjustment based on the linear correlation I ran to prove that PythagenMatt was indeed a step in the right direction.

That equation is:  Projected W% = (PythagenMattRaw - 0.1531) / 0.6938

The cool thing is that I got that linear best fit without doing anything other than asking Excel for it, and it works out that there is no directional bias (a .500 Raw PythagenMatt is a .500 projection…which proves that there is no systematic error in PythagenMatt).

To boil it down…PythagenMatt is game-averaged PythagenPat with an adjustment to fit a theoretical model.  I believe it works quite well.  The data I have thus far seems to suggest it does.

Updated through Games of: 08/24/07

Rk

Team

GP

W

RS

RA

Matt

Pat

Act

CLOSER

GAP

M-P

L10

L30

1

BOS

129

78

660

501

100.5

101.4

98.0

MATT

0.9

-0.9

0.749

0.662

2

NYY

128

71

764

611

96.9

98.5

89.9

MATT

1.6

-1.6

0.576

0.629

3

NYM

127

72

618

554

94.0

89.3

91.8

TIE

0.3

4.7

0.689

0.587

4

ANA

128

75

653

573

93.6

91.0

94.9

MATT

2.6

2.6

0.502

0.601

5

SDP

127

69

570

498

92.8

91.0

88.0

PAT

1.8

1.8

0.494

0.507

6

SEA

126

73

637

609

90.7

84.5

93.9

MATT

6.2

6.2

0.695

0.629

7

CHC

127

66

591

538

88.8

88.1

84.2

PAT

0.7

0.7

0.697

0.501

8

CLE

127

70

629

577

87.2

87.6

89.3

TIE

0.4

-0.4

0.677

0.465

9

PHI

127

66

683

657

86.8

84.1

84.2

PAT

2.4

2.7

0.244

0.518

10

DET

128

69

712

658

86.0

87.3

87.3

PAT

1.3

-1.3

0.279

0.304

11

TOR

128

64

576

544

84.9

85.3

81.0

TIE

0.3

-0.3

0.385

0.500

12

ARI

129

72

548

583

83.8

76.3

90.4

MATT

7.4

7.4

0.465

0.590

13

ATL

129

67

649

594

83.7

87.8

84.1

MATT

3.3

-4.1

0.475

0.540

14

BAL

127

58

587

611

82.9

77.9

74.0

PAT

5.0

5.0

0.368

0.455

15

COL

128

65

647

615

82.7

84.9

82.3

MATT

2.2

-2.2

0.457

0.571

16

MIL

128

65

612

626

80.7

79.3

82.3

MATT

1.5

1.5

0.384

0.352

17

OAK

130

65

586

566

80.2

83.6

81.0

MATT

1.8

-3.4

0.619

0.561

18

MIN

128

65

574

561

79.4

82.7

82.3

PAT

2.4

-3.4

0.630

0.497

19

LAD

128

66

573

553

79.2

83.7

83.5

PAT

4.1

-4.4

0.675

0.326

20

SFG

129

57

557

567

75.6

79.7

71.6

MATT

4.0

-4.0

0.630

0.532

21

CIN

128

58

617

685

74.1

72.8

73.4

TIE

0.1

1.3

0.512

0.471

22

STL

125

61

559

631

73.9

71.7

79.1

MATT

2.2

2.2

0.546

0.566

23

KCR

127

57

577

610

72.9

76.7

72.7

MATT

3.8

-3.8

0.569

0.558

24

HOU

129

57

575

659

72.6

70.5

71.6

TIE

0.0

2.1

0.286

0.480

25

PIT

127

56

579

645

71.4

72.7

71.4

MATT

1.2

-1.3

0.728

0.550

26

FLO

129

57

624

675

71.2

74.9

71.6

MATT

2.9

-3.7

0.300

0.379

27

TEX

128

56

629

668

69.5

76.3

70.9

MATT

4.1

-6.8

0.381

0.361

28

CHW

128

56

546

674

68.2

65.0

70.9

MATT

3.2

3.2

0.276

0.399

29

WAS

129

58

511

604

67.9

68.6

72.8

PAT

0.6

-0.6

0.437

0.522

30

TBD

128

49

580

767

57.9

59.4

62.0

PAT

1.5

-1.5

0.353

0.381

PythagenMatt leads PythagenPat 16-9-5 (W-L-T)

Matt = PythagenMatt estimated wins in 162 games
Pat = PythagenPat Seasonal win estimate.
Act = Projected Wins on current W%
GAP = The number of wins (per 162 games) by which the projection method that is closer to the team's actual prorated wins is closer.  Found by subtracting (Matt - Act) from (Pat - Act) and taking the absolute value.
M-P = PythagenMatt minus PythagenPat projected wins
L10 = The team's PythagenMatt winning percentage in the last ten calendar days (NOT games…the implementation of that would have been nearly impossible to do in Excel the way my spreadsheet is arranged).
L30 = The same as L10 only over the last 30 calendar days.  A longer term streak meter.

NOTE: I consider it a tie between PythagenMatt and PythagenPat when the closer of the two is less than half a win closer to real win rates than the further of the two (GAP < 0.5).