clock menu more-arrow no yes mobile

Filed under:

*St. Louis Cardinals 2014 VEB Community Projection Results: Allen Craig

USA TODAY Sports

I'll give you one guess as to what stat VEB most outguessed the projections systems.

It wasn't PAs, where VEB was below two of them (though Oliver projects 600 for all players). It wasn't runs or home runs, where the VEB collective tied Steamer for the high (78 and 20). And it definitely wasn't stolen bases or caught stealing, where there was almost no spread in the data.

It was, of course, RBIs, where the Hivemind projected 99 RBIs, a full 14 more than Steamer, and 32 more than PECOTA.


PA

R

HR

RBI

SB

CS

AVG

OBP

SLG

OPS

wOBA

Oliver

600

71

13

72

3

1

0.295

0.357

0.432

0.789

0.346

Steamer

622

78

20

85

2

1

0.298

0.357

0.480

0.837

0.364

PECOTA

500

59

17

67

2

1

0.288

0.344

0.467

0.811

0.351

ZiPS

503

61

16

84

3

0

0.284

0.338

0.454

0.792

0.346

VEB

555

78

20

99

2

1

0.306

0.364

0.498

0.862

0.364*

*Using ((1.7*OBP)+SLG)/3 to estimate wOBA: .372

Image_medium

via s3.amazonaws.com


Otherwise, the normal rose-colored glasses are there: Higher AVG, OBP, and SLG than all of the projection systems, and tied Steamer, who has consistently been the most optimistic system, with the highest wOBA.

Let's look at the quartiles and high/low:



PA

R

HR

RBI

SB

CS

AVG

OBP

SLG

OPS

wOBA

Minimum Values

440

60

0

76

0

0

0.279

0.331

0.430

0.770

0.298

25th Percentile

520

72

18

91

1

1

0.300

0.355

0.478

0.836

0.358

50th Percentile

557

77

19

100

2

1

0.307

0.365

0.495

0.857

0.367

75h Percentile

595

82

22

104

3

1

0.311

0.371

0.510

0.880

0.372

Maximum Values

650

105

34

125

10

6

0.333

0.422

0.877

1.299

0.400




PA

R

HR

RBI

SB

CS

AVG

OBP

SLG

OPS

wOBA

VEB Low OPS (mikeishere)

450

80

14

95

1

0

0.285

0.34

0.43

0.77

0.345

VEB High OPS (mike)

544

99

34

113

5

4

0.321

0.422

0.877

1.299

0.298


One of the issues that I'm having with running these projections is dealing with outlier data. Some of them are blatantly obvious, and I discount those, but others are in a gray area. I probably should have discounted this high projection and weirdly low wOBA, but here we are. If Ben's graphics differ from my numbers, exclusion is likely the culprit

CELEBRITY DOPPELGÄNGER: This one was apparently tough for all of you, but we did get both President Obama and First Lady Michelle Obama as answers! I do kind of like Colin Kaepernick (suggested by Milt Thompson), though