Positional Marginal Value Through Standard Deviation
Not all ball players are created equal. But then again, a player positioned at one defensive area has a different offensive value if he is to switch positions.

Why is the above statement true? Because these position's offensive values are radically different depending on the offensive average of each position. Historically the positions with a higher offensive average include the First Base, Third Base, Left Field, and Right Field, while positions up the middle [Catcher, Second Base, Shortstop, and Center Field] tend to be documented as being weaker.

One method for documenting positional marginal value is to first find the data for league positional average, [which I found on the Baseball Prospectus website. With the on base percentage, slugging percentage, and total plate appearances, I determine runs created percentage [which I simply multiply the on base percentage by slugging percentage].

Then I dig until I locate the individual player's stats and multiply the on base percentage by the slugging percentage and multiply this figure by plate appearances to calculate a runs created figure. Adjust runs created with the ballpark factor and divide runs created by total plate appearances. Then subtract the player's runs created average from the league average and square this figure to give us a standard deviation version of an average.

So you now have a whole series of numbers in front of you, but do not know how to read all of this. Below is an explanation how to read this set of numbers.

1. Dominent [Bonds .109561]
2. Superstar [Pujols .010609]
3. Excellent [Beltre .009025]
4. Good [Soriano .000576]
5. Above Average [Burrell .000036]
6. Marginal [Jenkins .000004]
7. Exactly Average [.000000 Roberts]
8. Below Average [Cintron -.000256]

Lastly, remember, do not compare players from different positions against each other. Simply said, the average positional data pool differs amongst each position.

2004 Positional Marginal Value [Standard Deviation]

Catcher
1) I. Rodriguez .004356
2) J. Posada .004096
3) J. Lopez .002601
4) J. Varitek .002601
5) V. Martinez .002025

First Base
1) A. Pujols .010609
2) T. Helton .009801
3) J. Thome .002809
4) M. Teixeira .002025
5) Ca. Delgado.001444

Second Base
1) M. Loretta .003136
2) J. Kent .001849
3) J. Uribe .001156
4) R. Durham .001024
5) A. Soriano .000576

Shortstop
1) Ca. Guillen .005329
2) M. Tejada .002916
3) K. Greene .001369
4) J. Rollins .001296
5) D. Jeter .001156

Third Base
1) A. Beltre .009025
2) S. Rolen .009025
3) M. Mora .006241
4) A. Ramirez .003249
5) A. Rodriguez .001764

Outfield
1) B. Bonds .109561
2) J. Edmonds .01404
3) L. Berkman .008649
4) M. Ramirez .007921
5) P. Guerrero .007744
6) J.D. Drew .007569
7) B. Abbreau .005041
8) A. Dunn .004096
9) G. Sheffield .003844
10) H. Matsui .003025
11) C. Beltran .002601
12) Er. Chavez .002209
13) A. Rowand .001600
14) M. Alou .001444
15) I. Suzuki .001296

Runs Created Over League Average
Many moons ago,sabermetric magician Bill James introduced us w/the idea of a new statistic called runs created." Basically, this mess of numbers is suppose to represent the amount of estimated runs the individual creates, while other team members are not taken into account ...read team independent statistics.

To take things even further, a stat whizzing through my caffein soaked brain suddenly took form of the name Runs Created Over League Average . After wracking my brain w/my over-simplified version of runs created [ob% times Slg% times plate appearances and adjust the runs created by the park factor again before dividing by plate appearances], I toss the figure located exactly up from this spot into the statistical blender [figure out the league average and subtract you figure from the league average]. When an individual's RC% is higher than the league average, add the points above average to .500, while subtracting points below average from .500.

Most of you would wonder how this set of numbers should be examined. Simply, when the percentage reads above .500, the player is above average, while in reverse the player is below average. A player near the .500 mark either positively or negatively is marginal, while a player above .550 is great. Someone below .490 ought to think about hanging up their cleats or must play way above the margin defensively. To make it easier for the average fan to comprehend this systematic series of numbers, I have created a tier system to classify the quality of the player. Players in tier one are the elite players of the league, while tier two represents the slightly worse, but not terrible horrible players to marginal players. The third teir are players who risk being sent to the minors.

2004 Runs Created Above League Average
[Minimum 500 Plate Appearances]

Tier I [.550+]
.845 B. Bonds
.639 J. Edmonds
.639 A. Pujols
.619 T. Helton
.614 S. Rolen
.614 A. Beltre
.612 L. Berkman
.606 J.D. Drew
.605 T. Hafner
.599 V.Guerrero
.589 M. Ramirez
.589 B. Abbreau
.586 J. Thome
.584 M. Mora
.584 A. Dunn
.574 D. Ortiz
.573 G. Sheffield
.572 S. Casey
.569 C. Beltran
.568 C. Guillen
.567 A. Ramirez
.566 H. Matsui
.560 M. Loretta
.558 I. Rodreguez
.558 E. Durazo
.556 C. Lee
.554 M. Cabrerra
.554 M. Alou
.554 J. Posada
.554 A. Rodriguez
.553 B. Wilkerson
.551 M. Lowell
.551 L. Overbay

Tier II [.500-.549]
.549 M. Teixeira
.547 P. Nevin
.546 C. Delgado
.545 P. Konerko
.545 B. Giles
.543 J. Kent
.542 C. Wilson
.541 M. Tejada
.541 I. Suzuki
.541 A. Rowand
.539 A. Huff
.538 V. Martinez
.537 Javey Lopez
.537 J. Burnitz
.536 Jo Guillen
.536 J. Varitek
.534 J. Blake
.534 J. Bagwell
.533 C. Jones
.532 R. Durham
.531 D. Lee
.530 K. Millar
.529 J. Damon
.528 J. Estrada
.528 O. Infante
.528 D. Jeter
.527 T. Martinez
.527 S. Green
.526 M. Piazza
.526 A. Jones
.525 P. Burrell
.525 M. Kotsay
.524 S. Sosa
.523 S. Finley
.522 J. Kendell
.522 D. Bell
.521 L. Ford
.521 K. Greene
.520 M. Bradley
.520 J. Wilson
.520 H. Blalock
.519 V. Castilla
.519 S. Hillenbrand
.519 G. Jenkins
.519 C. Pena
.518 R. Ibanez
.517 J. Pierre
.517 M. Cameron
.517 B. Williams
.516 M. Biggio
.516 J. Rollins
.514 M. Young
.514 M. Lawton
.514 M. Bellhorn
.514 E. Byrnes
.514 C. Crisp
.513 J. Uribe
.512 J. Conine
.510 C. Crawford
.509 R. Belliard
.509 P. Polonco
.508 T. Hunter
.508 J. Jiminez
.507 M. Lieberthall
.506 T. Wiggington
.506 S. Hatteberg
.506 R. Palmeiro
.505 J. Dye
.504 R. Winn
.504 R. Freel
.504 P. Feliz
.504 J. Cruz, Jr
.502 M. Tucker
.502 A. Kennedy
.501 P. Lo Duca
.501 M. Grissom
.501 R. Furcal
.500 T. Womack
.500 J. Randa
.500 A. Soriano

Tier III [.499-.000]
.498 O. Hudson
.498 C. Patterson
.497 Mackowiak
.497 J. Vizquel
.496 B. Higginson
.495 R. Baldelli
.495 K. Matsui
.495 J. Gerut
.495 E. Rentaria
.495 C. Tracy
.494 L. Castilla
.494 L. Bigbie
.494 J. Lugo
.493 R. Hildago
.492 C. Figgons
.491 S. Burroughs
.489 T. Batista
.489 E. Alfonzo
.489 D. Bautista
.489 Bo Crosby
.487 J. Pierzenski
.485 J. Jones
.485 Br. Boone
.483 J. Payton
.483 E. Martinez
.482 O. Cabrera
.482 Ed. Chavez
.479 T. Redman
.479 R. Clayton
.479 B. Roberts
.477 A. Berroa
.476 A. Gonzalez
.475 J. Crede
.471 R. Johnson
.471 D. Eckstein
.470 C. Guzman
.468 C. Counsell
.466 E. Hinske
.466 A. Miles
.465 A. Cintron
.461 S. Podsednik

The Importance of Independent Stats
In the world of insider baseball, known to most as Major League Baseball, most of the statistics posted by this group of individuals end up resembling a mess of illogical mathematics.

So how do you make a set of numbers resemble a set of logical calculations? The answer is simply using independent statistics in which other batters or baserunners are not to be included in the calculations. Some examples of these incomplete, team orientated stats include what most purists consider essential components, runs, and runs batted in. The problem is that these are a set of numbers which only tell half the picture of what an individual is doing on the field.

A quick, yet well thought out stat I have assembled is what I call, "Independent Bases Percentage." I originally used this set of numbers to be included in a stat very close to Bill Jame's Secondary percentage. The big difference from James is that I include base on balls as part of the primary stat, while James beleives that bases on ball should be part of the secondary stat. Another difference is that I have added on an extra bases : stolen bases ratio to document how the player achieves the secondary bases.

In the end, this stat ends up resembling a revised version of OPS, but instead of counting bases twice, I only count the extra bases and stolen bases in the secondary portion of the stat. Also, stolen bases are included in this part of the stat while not being included in the ops.

The calculation reads the following:

Primary Independent Base Percentage: On Base Percentage adjusted to the ballpark factor.

Secondary Independent Base Percentage: All extra bases [which means not counting first base with the extra base calculations] plus stolen bases divided by at bats plus stolen bases plus caught stealing. Also, adjust this with the ballpark factor. Additionally, I document the extra bases to stolen bases by showing a ratio between these two ways of base advancement.

Total Independent Base Percentage: Simply add up the primary and secondary category to display the final calculation.

2004 Stats

Primary Independent Base Percentage
Barry Bonds .600
Lance Berkman .450
Jim Edmonds .438
J.D. Drew .436
Albert Pujols .435
Travis Hafner .430
Scot Rolen .429
Bobby Abbreau .426
Tod Helton .422
Jorge Posada .419

Secondary Independent Base Percent [w/extra base : stolen base ratio]
Barry Bonds .451 [97:3]
Jim Edmonds .367 [96:4]
Albert Pujols .345 [98:2]
Carlos Beltron .342 [80:20]
Adam Dunn .324 [97:3]
Adrian Beltre .315 [96:4]
Jim Thome .304 [100:0]
Scot Rolen .302 [97:3]
Manny Ramirez .296 [99:1]
Vladimeur Guerrero .291 [92:8]

Total Independent Base Percentage
Bobby Bonds 1.051
Jim Edmonds .805
Albert Pujols .780
Scot Rolen .731
Adam Dunn .730
Carlos Beltron .728
Adrian Beltre .721
Travis Hafner .718
J.D. Drew .716
Bobby Abbreau .714

Total Adjusted Base Value
After reading a writeup by Bill James in his "Historical Baseball Abstract," I have come to realize that the statistic "secondary average" has some major flaws [bb+ extra bases + stolen bases-caught stealing divided by plate appearances].

My biggest gripe is including walks in the secondary average. I consider receiving base on balls a primary part of batting, not a secondary statistic. The player's primary goal is to get on base anyway possible, including a hit, a base on ball or score a run via home run. Even though hits are worth more than base on balls, batters have been trained and disciplined to take a substancial amount of base on balls because this increases the value of a player. Therefore, I advocate removing the base on balls statistic from the secondary average.

Another misconception used not only in secondary bases, but many other statistics is that "all bases are created equal." This is another of many numerous baseball myths. Hits are of more value than base on balls because they have a much higher potential to create a run for a team. The only way to create a run w/a walk while at home plate is w/the bases loaded [a walked in run]. Stolen bases have very little value unless you sucessfully steal 75% or more of these bases prolifically. Only a few elite players have ever reached this stature [ie; Rickey Henderson, Lou Brock, Ty Cobb, Billy Hamilton, etc]. Most frequent base stealers do so because they lack power or other batting qualities. On the other hand, the above statement is not always true, [see Ty Cobb's and Barry Bond's stats].

What I have done to correct this all too simple "napkin statistic" is to keep the original total bases count, yet reduce the value of the base on ball, stolen bases, and caught stealing. This reads as an approximate total bases per/27 outs. Another way of saying this is that this stat is calculated as if the player is batting one through nine on a batting card in a nine inning game [27 outs].

The statistical formula starts: [h + .75 bb] divided by at bats minus hits and multiply by twenty seven, which represents the amount of adjusted base on balls and regular hits per twenty seven outs [a complete game]. As always, I adjust the base totals to the park factor. [This factor is necessary since every ballpark has different nooks and crannies, some fences are longer and/or higher than others, some climates are more humid than others, and in one case, the altitude has a major impact on the field of play.]

Next, extra bases are added up to the adjusted stolen base count. Each stolen base is worth .667 of a stolen base, while each time a player is caught stealing is worth minus 1.333 stolen base. Once again I adjust the extra bases and stolen bases with the ballpark factor. Then I divide this figure by at bats minus hits. This number represents the number of extra bases and adjusted stolen bases per 27 outs.

To sum things up, I add up the above two calculations, which gives you total bases plus the adjusted base on balls, stolen bases and caught stealing.

So what is this data suppossed to represent? It is meant to document the exact amount of bases off of a hit, while attempting to adjust the base on balls, stolen base and caught stealing calculations. Since walks, stolen bases, and caught stealing are worth considerably a lot less than a hit, it deserves a lesser value than a total base.

Here are the 2004 MLB Adjusted Base Value top ten in the following categories.

Primary Base Value [h x .75 bb divided by ab-hits x 27]per 27/outs

Bonds 34.036
Berkman 19.379
Pujols 19.364
Edmonds 19.241
Drew 18.525
Rollen 18.184
Loretto 18.022
Hafner 17.963
Suzuki 17.898
Guererro 17.889

Secondary Base Value [extra bases + .667 stolen bases - 1.333 caught stealing divided by at bats minus hits x 27] p/27 outs

Bonds 18.832
Edmonds 14.586
Pujols 14.114
Beltre 13.296
Beltron 12.997
Dunn 12.367
Rollen 12.201
Guererro 12.103
Hafner 11.630
Thome 11.049

Total Adjusted Base Value [Primary Base Value + Secondary Base Value]

Bonds 52.886
Edmonds 33.827
Beltre 31.138
Rollen 30.385
Guererro 29.992
Hafner 29.603
Drew 29.100
Berkman 29.032
Dunn 28.748
Beltron 28.414