Front Office Fan

A Take on the Games We Watch and the Industry Behind Them

Archive for the ‘Statistics’ Category

Aggregate Return from Efficiency Averages

Posted by GM/VP of Fan Operations on Thursday, February 18, 2010

UPDATE: The formula for ROP has changed and AREA has thus changed with it. The changes are reflected below. UPDATE 2: It isn’t “ROP” anymore. Too many people complained about steals being involved as part of power and I agree. The stat is now called SPOT (Scoring Position Obtained Technique).

In baseball the prevailing wisdom says that getting on base and hitting for power are the most important factors to scoring runs. I disagreed with this notion and even toyed with the stats to try to disprove it; well, nothing doing. Apparently it’s more about power than I thought.

For a long time I’ve argued that batting average is just as important and it is essential to helping drive in runs. Though true that batting average is important to driving in runs, getting on base in scoring position is more important because of the opportunities it provides your teammates to bring you home.

Hitting for power, though, is only as relevant at the clip that a batter gets hits.A player that doesn’t hit for power but maintains a high batting average is useful as is the player that hits for a lot of power with a low batting average. They just belong at different spots in the lineup.

When looking at a batter’s hitting stats you want to know what kind of hits he’s getting.

The more often a batter hits for power the better. Obviously, a homer is more important than a triple and a triple more important than a double, but power in general is important.

If a batter often hits for power it doesn’t matter if he generally only hits doubles; what matters is that he’s always in scoring position. If he doesn’t hit for much power, though, you need that player to generally get more triples or, better yet, home runs.

Regardless, we need to know what type of hits these players are getting and Slugging Percentage just isn’t doing it.

"Every time I see you hit one in the air you owe me 20 push-ups."

You hear about players like Ichiro Suzuki that put on hitting displays in batting practice (not you, Willie Mays Hayes), but turn into singles machines too spite their power once the game starts. The fact is that the best-of the-best, the guys that make the “Greatest of all Time” conversations, are the ones that can hit for average, get power hits at a high percentage, and have the majority of their power hits be home runs.

What we have here is a case of three statistics pulling on one another.

Introducing two statistics that better explain this issue: Scoring Position Obtained Technique and Production of Power.

Scoring Position Obtained Technique (SPOT) is the ratio of extra-base hits and second base steals to total times on base. This shows how often a player gets on base in scoring position. Players that lack power but make up for it in speed are often overlooked by traditional sabermetricians and this formula properly accounts for their importance. The formula is very simple:

SPOT = (2B + 3B + HR + SB2) / (H+BB+HBP)

Production of Power (POP) is the ratio of power bases acquired to potential power bases. A power base is any base after first base. To explain, second base is the first power base, third base is the second, and home plate is the third power base. When a batter hits a home run they achieved 3-of-3 potential power bases. This is a similar theory to what developed slugging percentage but removes singles from the equation as singles have nothing to do with hitting for power.

POP = ((2B) + (3B*2) + (HR*3)) / ((2B + 3B + HR)*3)

So instead of the current, understood, slash-stats of BA/OBP/SLUG with the all-in-one wonder OPS, I propose the elimination of SLUG in favor of SPOT and POP and, with this revelation, the development of a better all-in-one stat: AREA.

Aggregate Return from Efficiency Averages (AREA) is more about geometry than advanced statistics. The acronym stems from nothing more than this rating looking to determine the area a batter’s four slash-stats cover. The more AREA a batter covers, the more they produce at the plate. AREAs will, for the most part, look like a statistic similar to batting average.

AREA = (((SPOT+POP) / 2) * (OBP)) + (((SPOT+POP) / 2) * (BA))

To properly map this it requires an X-Y plot graph. The Y-Axis is the on-base axis and the X-Axis is the power axis. The maximum X & Y values are ‘1’ and the minimum X & Y values are ‘-1.’ The four points placed on the graph are determined by their relative average at the given place on the graph. For example, the points (0, .400) would refer to a player’s OBP and (0, -.300)  would be the plotted point of the player’s BA. Similarly, (.667, 0) and (-.400, 0) would be the respective POPs and SPOTs of the same given player. On this plot you can then connect the dots into quadrilaterals. The bigger the given shape, the better a player is at bat. (NOTE: Negative numbers are only used for plotting simplicity. Take the absolute value of these numbers to determine the player’s given value for that statistic.)

To give you an idea of what happened in the world of AREA last season, I looked at every team and the all players in Major League Baseball with at least 500 plate appearances in 2009 (there were 143):

The Top 10:

NAME AREA OPS AVG OBP SLG SPOT POP
Albert Pujols 0.392 1.101 0.327 0.443 0.658 0.345 0.674
Joe Mauer 0.360 1.031 0.365 0.444 0.587 0.234 0.655
Alex Rodriguez 0.358 0.934 0.286 0.402 0.532 0.284 0.757
Prince Fielder 0.356 1.014 0.299 0.412 0.602 0.291 0.710
Troy Tulowitzki 0.353 0.929 0.297 0.377 0.552 0.346 0.702
Mark Reynolds 0.350 0.892 0.260 0.349 0.543 0.420 0.729
Hanley Ramirez 0.347 0.953 0.342 0.410 0.543 0.345 0.577
Ben Zobrist 0.345 0.948 0.297 0.405 0.543 0.322 0.661
Ryan Howard 0.340 0.931 0.279 0.360 0.571 0.368 0.698
Carl Crawford 0.340 0.816 0.305 0.364 0.452 0.434 0.582

The Worst 10:

NAME AREA OPS AVG OBP SLUG SPOT POP
Jason Kendall 0.168 0.636 0.241 0.331 0.305 0.169 0.420
David Eckstein 0.171 0.657 0.260 0.323 0.334 0.190 0.398
Edgar Renteria 0.191 0.635 0.250 0.307 0.328 0.206 0.480
Emilio Bonifacio 0.194 0.611 0.252 0.303 0.308 0.217 0.481
Randy Winn 0.200 0.671 0.262 0.318 0.353 0.280 0.408
Skip Schumaker 0.201 0.757 0.303 0.364 0.393 0.192 0.410
Luis Castillo 0.206 0.733 0.302 0.387 0.346 0.161 0.438
Russell Martin 0.207 0.681 0.250 0.352 0.329 0.175 0.513
Jhonny Peralta 0.207 0.691 0.254 0.316 0.375 0.232 0.496
Pedro Feliz 0.218 0.694 0.266 0.308 0.386 0.229 0.530

2009 Major League Baseball Teams:

TEAM AREA OPS AVG OBP SLG SPOT POP R
New York Yankees 0.296 0.840 0.283 0.362 0.478 0.297 0.621 915
Tampa Bay Rays 0.283 0.782 0.263 0.343 0.439 0.329 0.605 804
Boston Red Sox 0.280 0.806 0.270 0.352 0.454 0.305 0.595 873
Texas Rangers 0.280 0.765 0.260 0.320 0.445 0.341 0.623 784
Los Angeles Angels of Anaheim 0.276 0.791 0.285 0.350 0.441 0.284 0.587 883
Philadelphia Phillies 0.276 0.781 0.258 0.334 0.447 0.318 0.615 820
Colorado Rockies 0.271 0.784 0.261 0.343 0.441 0.298 0.599 803
Minnesota Twins 0.265 0.774 0.274 0.345 0.429 0.257 0.598 817
Toronto Blue Jays 0.264 0.773 0.266 0.333 0.440 0.292 0.589 798
Chicago White Sox 0.263 0.741 0.258 0.329 0.412 0.277 0.621 724
Detroit Tigers 0.260 0.747 0.260 0.331 0.416 0.257 0.622 743
Milwaukee Brewers 0.260 0.767 0.263 0.341 0.426 0.259 0.601 785
Los Angeles Dodgers 0.256 0.758 0.270 0.346 0.412 0.260 0.571 780
Arizona Diamondbacks 0.255 0.742 0.253 0.324 0.418 0.302 0.582 720
Florida Marlins 0.252 0.756 0.268 0.340 0.416 0.257 0.572 775
Cleveland Indians 0.252 0.756 0.264 0.339 0.417 0.269 0.565 773
St. Louis Cardinals 0.251 0.747 0.263 0.332 0.415 0.271 0.574 730
Washington Nationals 0.249 0.743 0.258 0.337 0.406 0.253 0.584 710
Baltimore Orioles 0.249 0.747 0.268 0.332 0.415 0.264 0.566 741
Houston Astros 0.248 0.719 0.260 0.319 0.400 0.287 0.571 643
Kansas City Royals 0.248 0.723 0.259 0.318 0.405 0.285 0.573 686
Seattle Mariners 0.245 0.716 0.258 0.314 0.402 0.278 0.580 640
Chicago Cubs 0.245 0.739 0.255 0.332 0.407 0.260 0.576 707
Oakland Athletics 0.244 0.725 0.262 0.328 0.397 0.284 0.543 759
Atlanta Braves 0.243 0.744 0.263 0.339 0.405 0.248 0.559 735
Cincinnati Reds 0.243 0.712 0.247 0.318 0.394 0.281 0.579 673
New York Mets 0.237 0.729 0.270 0.335 0.394 0.267 0.515 671
San Diego Padres 0.234 0.702 0.242 0.321 0.381 0.260 0.572 638
Pittsburgh Pirates 0.234 0.705 0.252 0.318 0.387 0.277 0.545 637
San Francisco Giants 0.234 0.698 0.257 0.309 0.389 0.275 0.551 657

And for Fun: My 2009 Lineup Card featuring the best players at each position:

NAME POS Bats AREA AVG OBP SLG OPS SPOT POP
Ben Zobrist 2B Switch 0.345 0.297 0.405 0.543 0.948 0.322 0.661
Alex Rodriguez 3B Right 0.358 0.286 0.402 0.532 0.934 0.284 0.757
Albert Pujols 1B Right 0.392 0.327 0.443 0.658 1.101 0.345 0.674
Joe Mauer C Left 0.360 0.365 0.444 0.587 1.031 0.234 0.655
Troy Tulowitzki SS Right 0.353 0.297 0.377 0.552 0.929 0.346 0.702
Adam Lind DH Left 0.324 0.305 0.370 0.562 0.932 0.339 0.621
Justin Upton RF Right 0.337 0.300 0.366 0.532 0.898 0.367 0.646
Carl Crawford LF Left 0.340 0.305 0.364 0.452 0.816 0.434 0.582
Matt Kemp CF Right 0.340 0.297 0.352 0.490 0.842 0.374 0.672

*Data was compiled using information provided by Baseball Prospectus and Baseball-Reference.com.*

Advertisements

Posted in Baseball, Ideas, Statistics | Tagged: , , | Leave a Comment »

GRIP: Given Runners per Inning Pitched

Posted by GM/VP of Fan Operations on Monday, July 6, 2009

UPDATE: I’m changing the title of this formula from “Given Runs per Inning Pitched” to “Given Runners per Inning Pitched. I just think it makes more sense.

The three most commonly used statistics in analyzing a pitcher are ERA, K/BB and WHIP. Each has its pros and cons but generally they each get the job done. Earned Run Average tells you how many runs you can expect a pitcher to give up through 9 innings. Strikeout to Walk Ratio gives you the pitcher’s ability to control the strike zone and how batters react to him. And Walks plus Hits per Innings Pitched generalizes how many batters a pitcher allows to reach base each inning. Problems include defensive factors, the inability to measure if a pitcher is a fly ball/groundball guy, and that where each accounts for one part of the pitcher’s production you must take each into account and assume that each is more valuable than the last.

Well in my amateur sabermetric studies I created a new stat and solved this problem (wow, aren’t I awesome?).

Allow me to introduce you to GRIP, Given Runners per Innings Pitched:

((TB + BB) – (K + DP)) / IP

This bastard child of WHIP and K/BB solves problems that the two of them create and was then adopted by ERA and raised as its own.

First its “problem.” Any sabermetrician will tell you in a few ways that the twin pillars of pitcher analysis are Park Factors and Defensive Efficiency. Namely, that Park A allows more home runs on Sundays, in April, in prime numbered years than Park B which always allows for more ground-rule doubles on Saturdays except for during a full moon. Defensive Efficiency: the infielders have no range and the outfielders have no arms so obviously a baserunner will have an advantage.

And there ends the problem. This statistic doesn’t factor the park or a pitcher’s batting average on balls in play. Well, here’s something to think about: do park factors and defense matter if a pitcher does his job? No.

Now if you’ll excuse me I’m going to start using my copy of Baseball Between the Numbers as a coaster.

A good pitcher will be a good pitcher in any ballpark regardless of outside factors. I agree that they will affect the numbers and could make a significant difference on the game, but the box score gives the onus of a win and a loss to a pitcher, not his defense and certainly not the field he’s playing on.

GRIP Explained:

GRIP measures total bases achieved rather than hits. WHIP would tell you that a home run an inning and a single per inning are equal, it would even tell you that two walks per inning are worse than one home run in an inning, and I’m telling WHIP to take a nap. That’s asinine.

K/BB works great and is entirely independent of defensive and park factors but it doesn’t tell you anything about hits.

The biggest problem stat-geeks have with ERA is its dependency on defense and park factors, which is why ERA was so quick to adopt GRIP as its own.

Strikeouts aren’t requisite of good pitching but they certainly do make the job easier. Indeed there are situations (bases loaded no outs, runner at third one out, etc.) where a strikeout is the best possible option and as such should be factored when discussing a pitcher’s numbers. Furthermore, ground balls are more valuable than fly balls and pitchers need to work in situations with runners on-base so it should be important to note how many double-plays result in a pitcher’s performance.

Therefore, taking strikeouts and double plays from total bases and walks gives you the total picture as to all things valuable that a pitcher does:

  • You get the idea of his command of the strike zone.
  • You better know if he is a ground ball or fly ball pitcher.
  • You know how he works in important moments.

Using GRIP and the other three traditional metrics let’s find out who has had the best pitching this season.

The 30 MLB teams ranks 1-30 with the best GRIPs.

TEAM

LG

RA

K/BB

WHIP

GRIP

Los Angeles Dodgers

NL

3.77

2.02

1.264

0.789

San Francisco Giants

NL

3.75

2.14

1.316

0.858

Atlanta Braves

NL

4.38

2.12

1.342

0.902

Chicago Cubs

NL

4.16

2.03

1.346

0.954

Boston Red Sox

AL

4.37

2.3

1.373

0.983

St. Louis Cardinals

NL

4.25

2.19

1.289

0.991

Kansas City Royals

AL

4.82

2.1

1.361

1.010

Seattle Mariners

AL

4.19

1.93

1.333

1.021

Chicago White Sox

AL

4.5

2.18

1.342

1.023

Florida Marlins

NL

4.83

1.97

1.414

1.048

New York Yankees

AL

4.82

2.08

1.366

1.076

Colorado Rockies

NL

4.73

2.11

1.391

1.087

Arizona Diamondbacks

NL

4.84

2.15

1.385

1.088

Toronto Blue Jays

AL

4.52

2.16

1.354

1.095

Oakland Athletics

AL

4.75

2.05

1.392

1.096

Detroit Tigers

AL

4.65

1.83

1.418

1.103

Cincinnati Reds

NL

4.39

1.72

1.360

1.106

Tampa Bay Rays

AL

4.63

2.01

1.361

1.106

San Diego Padres

NL

4.98

1.91

1.392

1.113

Minnesota Twins

AL

4.43

2.39

1.309

1.126

Houston Astros

NL

4.57

2.1

1.399

1.137

New York Mets

NL

4.68

1.67

1.426

1.150

Milwaukee Brewers

NL

4.73

1.96

1.385

1.182

Los Angeles Angels of Anaheim

AL

4.97

1.91

1.444

1.198

Pittsburgh Pirates

NL

4.47

1.5

1.394

1.247

Texas Rangers

AL

4.59

1.66

1.388

1.261

Philadelphia Phillies

NL

4.89

2.01

1.454

1.297

Cleveland Indians

AL

5.58

1.58

1.541

1.367

Baltimore Orioles

AL

5.57

1.83

1.486

1.442

Washington Nationals

NL

5.75

1.35

1.560

1.484

Top-10 Starters (at least 10 starts):

NAME

TEAM

LG

ERA

SO/BB

WHIP

GRIP

Tim Lincecum

SFG

NL

2.23

4.7

1.050

0.132

Dan Haren

ARI

NL

2.16

7.93

0.826

0.231

Javier Vazquez

ATL

NL

3.05

5.65

1.071

0.295

Zack Greinke

KCR

AL

2

6.32

1.055

0.354

Chris Carpenter

STL

NL

2.32

5.27

0.824

0.373

Felix Hernandez

SEA

AL

2.62

3.35

1.174

0.429

Justin Verlander

DET

AL

3.54

3.71

1.198

0.439

Jake Peavy

SDP

NL

3.97

3.29

1.188

0.490

Chad Billingsley

LAD

NL

3.14

2.25

1.205

0.490

Josh Johnson

FLO

NL

2.76

3.03

1.128

0.507

Bottom-10 Starters (at least 10 starts):

NAME

TEAM

LG

ERA

SO/BB

WHIP

GRIP

Matt Harrison

TEX

AL

6.11

1.48

1.642

1.705

Jamie Moyer

PHI

NL

5.72

2.13

1.472

1.706

Brian Moehler

HOU

NL

5.64

2.14

1.552

1.716

Trevor Cahill

OAK

AL

4.55

1.18

1.441

1.720

Manny Parra

MIL

NL

7.52

1.34

1.918

1.747

David Huff

CLE

AL

6.06

1.88

1.519

1.750

Bronson Arroyo

CIN

NL

5.85

1.32

1.573

1.767

Bartolo Colon

CHW

AL

4.23

1.8

1.500

1.789

Fausto Carmona

CLE

AL

7.42

0.88

1.813

1.846

Scott Olsen

WAS

NL

6.04

1.78

1.712

1.941

Baseball’s Closers (at least 10 save opportunities):

NAME

TEAM

LG

SV

BS

ERA

SO/BB

WHIP

GRIP

Jonathan Broxton

LAD

NL

20

2

2.72

4.06

0.857

-0.580

Joe Nathan

MIN

AL

22

2

1.35

6.14

0.750

-0.270

Joakim Soria

KCR

AL

12

2

1.74

3.86

1.016

-0.145

Heath Bell

SDP

NL

23

1

1.49

2.67

1.046

0.028

David Aardsma

SEA

AL

17

1

1.41

2.18

1.148

0.052

Andrew Bailey

OAK

AL

9

4

2.03

2.85

1.027

0.103

J.P. Howell

TBR

AL

6

5

1.63

2.88

1.086

0.103

Trevor Hoffman

MIL

NL

18

1

1.93

3.8

0.943

0.129

Mariano Rivera

NYY

AL

21

1

2.6

14

0.923

0.231

Francisco Cordero

CIN

NL

20

1

1.8

2

1.143

0.314

Frank Francisco

TEX

AL

14

2

2.1

4

0.935

0.351

Francisco Rodriguez

NYM

NL

21

3

1.59

1.82

1.185

0.353

Ryan Franklin

SLP

NL

20

1

0.84

3.43

0.844

0.375

Brian Fuentes

LAA

AL

24

3

3.38

3.56

1.227

0.375

Mike Gonzalez

ATL

NL

9

4

3.2

3.18

1.246

0.381

Brian Wilson

SFG

NL

21

4

3.41

3

1.216

0.405

Huston Street

COL

NL

19

1

2.8

4.1

1.075

0.425

George Sherrill

BAL

AL

18

3

2.43

2.67

1.140

0.570

Chad Qualls

ARI

NL

15

4

3.93

7.5

1.165

0.670

Bobby Jenks

CHW

AL

19

2

3.14

4.67

1.047

0.698

Fernando Rodney

DET

AL

17

0

4.11

1.75

1.314

0.771

Jose Valverde

HOU

NL

6

4

4.08

4

1.075

0.792

LaTroy Hawkins

HOU

NL

10

3

2.48

2.64

1.211

0.798

Kevin Gregg

CHC

NL

14

3

3.62

2.31

1.286

0.804

Jonathan Papelbon

BOS

AL

20

2

1.75

2.06

1.333

0.806

Kerry Wood

CLE

AL

10

4

5.08

1.94

1.447

0.988

Matt Capps

PIT

NL

18

2

4.88

1.82

1.482

1.120

Matt Lindstrom

FLO

NL

14

2

6.52

1.3

1.897

1.414

Joel Hanrahan

WAS

NL

5

5

7.71

2.5

1.959

1.653

Brad Lidge

PHI

NL

16

6

7.06

1.94

1.807

1.670

Interesting things to note:

  • Teams are evenly disbursed throughout the chart regardless of their league affiliation.
  • Natural Rivals Baltimore and Washington are second to last and last, respectively.
  • Despite being tied for first in the AL-West, the Angels and Rangers are 24th and 26th in GRIP.
  • Baseball’s Top-3 Closers have negative GRIPs. A perfect Grip is -3.00.
  • The Philadelphia Phillies are 1st in the NL-East and have the 27th overall GRIP. The Marlins are a game behind them and rank 10th in GRIP.
  • Most interesting AL All-Star selections: Mark Buehrle, Chicago White Sox and Tim Wakefield, Boston Red Sox (18th and 37th ranked AL Starters); Brian Fuentes, Los Angeles Angels of Anaheim (8th ranked closer).
  • Most interesting NL All-Star Selections: Jason Marquis, Colorado Rockies (40th ranked NL Starter).

So there you have it. It isn’t perfect, but no stat is.

Posted in Baseball, Statistics | Tagged: , , | Leave a Comment »