clock menu more-arrow no yes

Filed under:

Projection System Faceoff

New, 32 comments

Three major projection systems have data for the Mariners now. Lets see how they all compare.

Joe Nicholson-USA TODAY Sports

Yesterday, FanGraphs released their ZiPS projections for the Seattle Mariners. Overall, Dan Szymborski's system is pretty conservative towards the Mariners. The dimensions of Safeco Field certainly don't do any favors to our batters and the relative youth of the back end of our pitching staff does some harm to their projections. The good news is that the quartet of stars -- Felix, Cano, Seager, and Iwakuma -- are all projected to continue their dominance, providing a solid foundation to the team.

Now that we have these projections publically available, there are three leading projection systems available to compare (Steamer and PECOTA are the other two). First, I'll compare the individual projections each projection system spits out. Then, I'll compare the win projections that are calculated via each projection system.

Note: ZiPS doesn't spit out projections based on actual projected playing time. Since both Steamer and PECOTA have a win projection calculated based on their respective projections, projected playing time is accounted for in these systems. To best compare the three projection systems, I scaled the ZiPS projections based on the average projected playing time that Steamer and PECOTA used. That also allowed me to build a win projection based on ZiPS, something that isn't available from FanGraphs or Dan.

To the data!

Battters

Player

swOBA

zwOBA

sWAR

zWAR

pWAR

Robinson Cano

0.355

0.346

5.0

5.0

6.0

Kyle Seager

0.337

0.333

4.1

4.1

3.9

Nelson Cruz

0.334

0.332

1.5

2.2

2.4

Logan Morrison

0.333

0.314

1.8

0.7

1.3

Seth Smith

0.327

0.317

1.5

1.1

1.2

Jesus Montero

0.322

0.294

0.1

-0.1

0.1

Dustin Ackley

0.312

0.305

1.9

1.8

2.0

Austin Jackson

0.310

0.299

2.2

1.9

2.4

Brad Miller

0.310

0.304

2.0

1.8

1.7

Justin Ruggiano

0.306

0.307

0.7

0.6

0.8

Mike Zunino

0.300

0.290

2.8

1.8

1.7

Chris Taylor

0.297

0.300

1.5

1.6

1.4

Willie Bloomquist

0.279

0.285

0.2

0.2

0.0

James Jones

0.269

0.272

0.0

0.0

-0.2

John Hicks

0.260

0.246

0.0

0.0

-

Jesus Sucre

0.247

0.240

0.1

0.3

-0.1

Total

-

-

25.5

23.0

24.6

Note: I don't have access to the necessary data to calculate a wOBA for the PECOTA projections so we'll only be comparing projected wOBA for Steamer and ZiPS.

Overall, ZiPS is much more conservative in its offensive projections for the Mariner batters than both Steamer and PECOTA. A few observations:

  • Despite a projected wOBA that is essentially the same, Steamer and ZiPS wildly disagree on Nelson Cruz's overall value. I'm not sure if that has to do with his positional adjustment because both ZiPS and Steamer have his fielding value at or near 0.
  • Both ZiPS and PECOTA are bearish on Mike Zunino's value but Steamer sees him as the third most valuable player on the team. This almost certainly due to his defensive projection. Steamer sees him as a +4 defender but ZiPS sees him as a -2 defender. Based on his reputation and the defensive numbers he put up last year, I'd think Steamer is a closer bet to what we can expect from our backstop in 2015.
  • The projection systems really love Chris Taylor. Over a full season, Taylor is projected to accumulate 2.3 zWAR and 2.7 sWAR. Brad Miller is also projected favorably. This position battle during Spring Training is going to be fun to watch.

Pitchers

Player

sFIP

zFIP

pFIP

sWAR

zWAR

pWAR

Felix Hernandez

2.75

2.60

2.77

4.5

4.8

4.5

Hisashi Iwakuma

3.43

3.41

3.62

2.9

3.1

2.1

James Paxton

3.88

3.76

3.54

1.6

1.8

0.8

Roenis Elias

4.11

4.21

4.02

1.2

0.6

0.0

J.A. Happ

4.13

4.15

4.20

1.1

0.7

-0.6

Taijuan Walker

4.18

3.92

3.78

0.3

0.3

0.3

Fernando Rodney

3.11

3.18

3.30

1.0

0.4

0.5

Danny Farquhar

3.20

2.82

2.89

0.5

0.6

0.8

Charlie Furbush

3.28

3.02

3.42

0.5

0.5

0.3

Dominic Leone

3.31

3.54

3.19

0.2

0.2

0.6

Yoervis Medina

3.48

3.59

4.03

0.2

0.1

-0.2

Tom Wilhelmsen

3.59

3.74

3.91

0.4

0.4

0.3

Carson Smith

3.76

3.33

2.85

0.0

0.2

0.4

David Rollins

4.06

4.60

4.31

0.0

-0.1

-0.1

Lucas Luetge

4.04

4.30

4.02

0.0

-0.2

-0.1

Erasmo Ramirez

4.29

4.33

3.93

0.1

0.0

0.1

Danny Hultzen

4.00

4.21

3.35

0.1

0.0

0.1

Mayckol Guaipe

4.14

3.62

-

0.0

0.0

-

Mike Kickham

4.41

4.13

4.24

0.0

0.0

-0.1

Edgar Olmos

4.65

5.04

-

-0.1

-0.2

-

Total

14.5

13.6

9.7

The front end of the Mariners' rotation fairs well in each of the projection systems but the value of the back end varies widely. All three systems are in agreement about our bullpen, with Steamer giving Fernando Rodney a boost due to his high leverage innings. A few observations:

  • PECOTA generally hates the back end of our rotation. The four pitchers that will comprise the last three spots in the rotation are projected to accumulate a cumulative total of just 0.5 pWAR. J.A. Happ's projection is particularly ugly.
  • When ZiPS was released yesterday, there was some grumbling about James Paxton's projection. He was projected for just 1.2 zWAR, a total that is less than the amount of WAR he accumulated in 2014 in only 74 innings. But when we scale his projection to 170 innings, ZiPS projects him to accumulate the most WAR out of all three projection systems.
  • The projections don't really know what to do with Roenis Elias. Steamer sees him as the most valuable but PECOTA projects a lower FIP from him. PECOTA also thinks he'll only pitch 91 innings in 2015, a realistic scenario if Taijuan Walker makes the rotation out of Spring Training.
  • Out of our relief core, the projection systems love Danny Farquhar the most. If only Lloyd McClendon saw him as the best option out of the bullpen.

Win Projection

Team

W

L

W%

RS/G

RA/G

Steamer

89

73

0.547

4.20

3.79

PECOTA

87

75

0.537

4.13

3.80

ZiPS

89

73

0.547

4.30

3.87

FanGraph's projected standings are currently calculated using only Steamer -- once all of ZiPS have been released, the standings will use a combination of both systems. PECOTA also has a win projection built from its numbers. ZiPS is the only system that doesn't have a win projection calculated directly from the numbers it spits out. Using the BaseRuns formula, I took the ZiPS projections and built an expected runs scored and runs allowed for the Mariners. Then I threw those numbers into the PythagenPat formula to come up with a win projection. Despite being pretty conservative overall, ZiPS's win projection matches Steamer's win projection.

Since not all of the AL teams have been released by ZiPS, I wasn't able to build full standings for the AL or even the AL West from ZiPS. But I think we can assume that 89 wins will be towards the top of the standings and probably good enough to get us at least a Wild Card berth. PECOTA's projection is the most bearish as they project the Mariners to claim the first Wild Card spot with the second-best record in the AL. Steamer is the most bullish as it projects a division title for the Mariners with the best record in the AL.

These projection systems are great tools to get a feel for how the Mariners compare to the rest of the league. Each one has their quirks and players they like more than the other systems. By taking a broad sweep of all the data, we can begin to get a feeling for what we might expect from the Mariners in 2015.