AL: Most Valuable Player

Go brew some coffee, this is a long one!

The American League MVP -- David Ortiz or Alex Rodriguez -- will be announced on Monday. Either player would be a worthy selection, but who do I think should win?

I should say right off that these awards are pretty much worthless since they have so often gone to the wrong player. The members of the BBWAA have next-to-zero credibility.

Second, the actual definition of the award is highly subjective; it seems that everyone had his or her own definition. The letter each voter receives states that they should consider:

1. Actual value of a player to his team, that is, strength of offense and defense.
2. Number of games played.
3. General character, disposition, loyalty and effort.
4. Former winners are eligible.
5. Members of the committee may vote for more than one member of a team.

My definition: "If I was putting together a team, what player would I choose first, based on his performance for the season in question?"

Ortiz:

Everybody's been talking to me about the MVP situation. As soon as they bring my name up, they always talk about, 'Oh, he's a DH. I don't think he deserves it because he's a DH.' You win the MVP because you help your ballclub, you win games whenever the team needs it and because you put up some numbers.

Tony Massarotti (Herald) calls it "discrimination ... Since the creation of the position in 1973, the closest any DH has come to copping the MVP was in 1979, when California Angels slugger Don Baylor won the honor. Baylor played 97 games in the outfield and 65 as the DH ..."

I agree with Mazz. Should a DH be considered? Yes. I don't like the DH and wish baseball would get rid of it, but it is a valid position, just like shortstop, catcher, relief pitcher. And players at all of those positions have won MVP awards. ... Reliever Dennis Eckersley won the 1992 AL MVP when he pitched 80 innings. Rollie Fingers pitched only 78 innings when he won the AL MVP in 1981.

Okay, some numbers:

The Hardball Times has the final Win Share totals. The AL's top 5 (with batting and fielding splits):

          TOT   BAT  FLD
A-Rod      37  33.3  3.3
Manny      34  30.9  2.9
Sheffield  33  30.5  2.2
Teixeira   32  29.1  3.3
Ortiz      31  31.4  0.2

Ortiz comes in fifth because he gains only .2 Win Shares from his fielding. However, the idea that he should be automatically removed from MVP consideration because he doesn't play defense is wrong.

There must be a point at which his hitting compensates for his lack of defense. (Though some might say that not playing first benefits the Red Sox, allowing them to put a better-fielding player at the position.) The question is: Where it that point?

Here are some other stats:

        AVG   OBP   SLG   OPS    RC   RC/27
A-Rod  .321  .421  .610  1.031  151.1  9.53
Ortiz  .300  .397  .604  1.001  139.6  8.51

        EQA  BABIP  ISO  SecA  VORP
A-Rod  .350  .347  .288  .474  99.7
Ortiz  .336  .303  .303  .476  85.8

Explanation of some stats:

RC (Runs Created): The number of runs a hitter contributes to his team (not adjusted for ballpark).

RC/27: Runs Created per 27 outs; also the amount of runs that nine Ortizes (for example) would score against league-average pitching.

ISO (Isolated Power): (2B + 3B + HR*3) / AB

BABIP: Batting Average on Balls put Into Play.

EQA (Equivalent Average): A measure of total offensive value per out, adjusted for league offensive level, home park, and team pitching. EQA considers batting and baserunning, but not defense.

SecA (Secondary Average): A ratio of bases gained from other sources than total number of hits (extra base hits, walks and net bases gained through stolen bases). (TB − H + BB + SB − CS) / AB

VORP (Value Over Replacement Player): The number of runs contributed beyond what a replacement-level player at the same position would contribute if given the same percentage of team plate appearances. VORP does not consider the quality of a player's defense.

Rodriguez is the clear winner here. ... And at that point, Yankees blogger Larry Mahnken writes that if Rodriguez is the superior offensive player, then the defense question is moot.

Ortiz drove in 148 runs; Rodriguez brought in 130. But RBI is a highly-misleading stat, because a player's total depends on opportunity. If he hits a ton, but has no one on base, he won't get many RBI. So Baseball Prospectus presents RBI Opportunities (the number of runs a batter has driven in per runner on base. RBIs resulting from a batter driving himself in on home runs are not counted).

In 2005, there were 223 players with at least 400 plate appearances (the last column is RBI per baserunner):

                        PA  R1  R2 R3 Tot RBI
  1. Jorge Cantu       631 209 128 85 422  89 .2109
  2. Mark Teixeira     730 265 138 78 481 101 .2100
  3. Garret Anderson   603 190 135 65 390  79 .2026
  4. Vladimir Guerrero 594 187 115 75 377  76 .2016
  5. Manny Ramirez     650 248 151 94 493  99 .2008
  6. David Ortiz       713 262 175 69 506 101 .1996 
  7. Travis Hafner     578 189 135 62 386  75 .1943
  8. Mike Sweeney      514 177  90 53 320  62 .1938
  9. Carlos Delgado    616 212 131 82 425  82 .1929
 10. Gary Sheffield    675 234 146 85 465  89 .1914
 11. Carl Everett      547 181  97 57 335  64 .1910
 12. Reed Johnson      439 126  85 53 264  50 .1894
 13. Richie Sexson     656 217 148 74 439  82 .1868
 14. Chipper Jones     432 126 104 45 275  51 .1855
 15. Frank Catalanotto 475 148  80 47 275  51 .1855
 16. Garrett Atkins    573 192 137 81 410  76 .1854
 17. Matt Holliday     526 186 130 52 368  68 .1848
 18. David DeJesus     523 129  79 49 257  47 .1829
 19. Aramis Ramirez    506 168 115 56 339  61 .1799
 20. Hideki Matsui     704 269 171 79 519  93 .1792
 21. Juan Uribe        540 155 103 50 308  55 .1786
 22. Rich Aurilia      468 148 104 52 304  54 .1776
 23. Jason Larue       422 122  91 49 262  46 .1756
 24. Carl Crawford     687 192 120 65 377  66 .1751
 25. Ben Molina        449 143  96 70 309  54 .1748
 26. Jeff Kent         637 229 126 80 435  76 .1747
 27. Mark Kotsay       629 211 116 58 385  67 .1740
 28. So Taguchi        424 132  76 51 259  45 .1737
 29. Grady Sizemore    706 172 117 52 341  59 .1730
 30. Aubrey Huff       636 195 126 84 405  70 .1728
 31. David Wright      657 223 129 82 434  75 .1728
 32. Felipe Lopez      648 165 135 60 360  62 .1722
 33. Craig Monroe      623 192 140 70 402  69 .1716
 34. Carlos Beltran    650 161 123 80 364  62 .1703
 35. Tadahito Iguchi   582 155 119 55 329  56 .1702
 36. Chase Utley       628 224 152 77 453  77 .1700
 37. Carlos Lee        688 231 162 90 483  82 .1698
 38. Matt Stairs       466 163  93 58 314  53 .1688
 39. Mike Young        732 215 115 67 397  67 .1688
 40. Ken Griffey Jr.   555 178 105 56 339  57 .1681
 41. Miguel Cabrera    685 254 152 88 494  83 .1680
 42. David Eckstein    713 140 119 57 316  53 .1677
 43. Albert Pujols     700 234 144 77 455  76 .1670
 44. Emil Brown        609 207 141 66 414  69 .1667
 45. Jay Gibbons       518 181  98 40 319  53 .1661
 46. Pat Burrell       669 248 177 89 514  85 .1654
 47. Raul Ibanez       690 206 145 67 418  69 .1651
 48. Nick Johnson      547 177 124 58 359  59 .1643
 49. Placido Polanco   551 141  84 61 286  47 .1643
 50. Brian Roberts     640 165 107 63 335  55 .1642
 51. Brad Wilkerson    661 120 107 54 281  46 .1637
 52. Vernon Wells      678 216 129 77 422  69 .1635
 53. Adam Laroche      502 173 112 70 355  58 .1634
 54. Cliff Floyd       626 183 141 68 392  64 .1633
 55. Khalil Greene     476 162 115 61 338  55 .1627
 56. Chris Shelton     431 127  82 43 252  41 .1627
 57. Bobby Abreu       719 237 165 78 480  78 .1625
 58. Edgardo Alfonzo   402 119  91 43 253  41 .1621
 59. Trot Nixon        470 175  97 62 334  54 .1617
 60. Morgan Ensberg    624 197 137 69 403  65 .1613
 61. Rod Barajas       450 121  78 43 242  39 .1612
 62. Orlando Hudson    501 159 104 66 329  53 .1611
 63. Derrek Lee        691 187 124 68 379  61 .1610
 64. Russ Adams        545 165 109 68 342  55 .1608
 65. Rondell White     400 127  79 49 255  41 .1608
 66. Alfonso Soriano   682 219 148 57 424  68 .1604
 67. Jose Reyes        733 143 105 70 318  51 .1604
 68. Miguel Tejada     704 227 134 89 450  72 .1600
 69. Alex Rodriguez    715 252 180 84 516  82 .1589
...
 91. Johnny Damon      688 207 138 80 425  65 .1529
124. Jason Giambi      545 201 116 59 376  55 .1463
166. Robinson Cano     551 192 108 62 362  48 .1326
170. Edgar Renteria    692 238 147 87 472  62 .1314
171. Jorge Posada      546 205 117 74 396  52 .1313
179. Bernie Williams   546 224 115 65 404  52 .1287
182. Jason Varitek     539 198 124 56 378  48 .1270
183. Derek Jeter       752 190 134 80 404  51 .1262
189. Bill Mueller      590 209 141 64 414  52 .1256
198. Kevin Millar      519 170 117 56 343  41 .1195
200. Tony Graffanino   417 139  85 37 261  31 .1188

In bringing home the bacon when there is bacon out there to be brought, Ortiz is much better than A-Rod.

Blue Jay pitcher Josh Towers: "As good a hitter as he is, and he's one of the best, no question, I could never vote for a guy who doesn't play defense."

Many people would agree. I would have given Ortiz more time at 1B this season (and less to Kevin Millar), but that's another story. The fact is: Ortiz did not play much in the field -- and Rodriguez is thought of as one of the best fielding third baseman in baseball. But is he?

A SoSHer looked at Slappy's defense and reported:

He is well below the average 3rd basemen in the league ... He's WORST in Range Factor for qualified AL starters, SECOND WORST in Zone Rating ... That is downright horrid. He is 9 runs worse then the average 3rd baseman.

and offered a random sampling of three other third basemen:

         A-Rod  Chavez  Beltre  Inge
RF:       2.59   2.80    2.78   3.25
ZR:       .730   .814    .798   .802
FRAA:       -9      9      10     14
FRAR:       13     30      30     35
EqR:       265     89      74     83
Rate:       94    106     107    109

RF (Range Factor): The number of plays made per 9 innings.

ZR (Zone Rating): The percentage of balls hit into the player's area that he turns into outs.

FRAR (Fielding Runs Above Replacement): The difference between an average player and a replacement player is determined by the number of plays that position is called on to make. While the value at each position changes over time, the all-time adjustment for third base is 22.

FRAA: Fielding Runs Above Average. Same idea as FRAR, but the fielder is compared to his peers.

Rate: A way to look at the fielder's rate of production, equal to 100 plus the number of runs above or below average this fielder is per 100 games. (This is similar to ERA+ or OPS+.) A player with a rate of 110 is 10 runs above average per 100 games, a player with an 87 is 13 runs below average per 100 games, etc.

Judging by these stats, Rodriguez was not a good third baseman in 2005 (regardless of the comparisons). In September, the media talked about how few errors he had made in the second half of the season, but with a poor Range Factor, he wasn't getting to tough balls on which he might have made errors. ... But was his hitting so much more productive than Ortiz's to overcome this deficiency (assuming Ortiz's defensive value was 0)?

ESPN's Buster Olney sees things differently:

A-Rod has made play after play after play at third base down the stretch, and he's had an incredible offensive season. Ortiz has become the baseball equivalent of Joe Montana in big spots -- you know he's going to get a big hit. Ortiz is hitting about 60 points higher than his overall average with runners in scoring position, while A-Rod is hitting about 20 points lower than his overall average; he's not the most feared hitter in his own lineup. The Yankees win the division, so A-Rod gets the edge.

New York was given the East title only because of necessary seeding for the playoffs. Both teams finished with identical records, but the Yankees went 10-9 against the Red Sox this year. For this discussion, looking at who won the East is meaningless.

Say a tie-breaker had been played and the Red Sox won 1-0 on a non-Ortiz HR. Should Ortiz get the MVP because the Sox won the tie-breaker (and the East)? No.

Scott Miller (Sportsline):

Ortiz, by the way, did everything in his power this weekend to wrap up what should be his rightful AL most valuable player trophy. No disrespect intended to the masterful season constructed by Alex Rodriguez, but given Ortiz's stunning array of clutch hits and his invaluable role to the psyche of this team, this is one of those very, very rare instances in which a designated hitter should win the award.

This is an argument put forth by Ortiz's many supporters -- that he hits much better than Rodriguez when the game is on the line, when his production would be more valuable.

This goes hand-in-hand with the perception (even among Yankees fans) that Rodriguez tends to perform better when the outcome of the game is not in doubt (such as hitting a two-run homer with his team ahead 10-1).

Tom Verducci, in making his case for Rodriguez, tries to shoot that theory down, writing that A-Rod has been

good enough in the clutch. ... [T]he idea that Rodriguez doesn't come through often enough in big spots is a myth. ... Rodriguez is just as tough an out in a big spot as Ortiz -- actually, a little tougher if you read on-base percentage as the percentage of time the batter wins the war against the pitcher. The numbers do show that Ortiz is better at delivering the big blow -- the best in the game, in fact. But don't discount Rodriguez's work in key spots.

Years ago, we wouldn't have had any idea if Verducci was right or wrong. Now we do. So, what is the truth?

Thanks to SoSHer Eric Van, we have the evidence:

Boston and New York each played 65 close games (extra innings or decided in regulation by one or two runs):

        PA   AVG  OBP  SLG  HR RBI  RS
Ortiz: 288  .321 .417 .699  24  62  49
ARod:  282  .243 .340 .465  15  38  33

Each team played 20 games won by six or more runs:

        PA   AVG  OBP   SLG  HR RBI  RS
Ortiz: 100  .277 .360  .639   9  33  27
ARod:   98  .549 .622 1.171  15  46  39

If you remove the blowout-win stats from their season totals:

        PA   AVG  OBP   SLG  HR  RBI  RS
Ortiz: 613  .303 .403  .598  38  115  92
ARod:  617  .285 .389  .522  33   84  85

So Verducci, in citing Rodriguez's allegedly higher OBP in "big spots" (non-blowouts), is actually wrong. And in looking at "close and late" situations (below), we'll see that Verducci is wrong again.

The "close and late" numbers are easily found. I guess that's why Verducci merely made a claim, rather than backing it up with any evidence. Because the evidence supporting his claim does not exist. Indeed, the evidence shows the opposite of his claim.

A closer look (again, thanks to Van):

Games Won by 6+ Runs

        GM   PA   BA   HR  RBI   R   OBP    SA    OPS
Ortiz   20  100  .277   9   33  27  .360   .639   .999
ARod    20   98  .549  15   46  39  .622  1.171  1.793

Games Won by 3-5 Runs

        GM   PA   BA   HR  RBI   R   OBP    SA    OPS
Ortiz   36  161  .359   9   33  32  .478   .680  1.158
ARod    38  175  .382  12   32  40  .520   .721  1.241

Games Won by 2 Runs

        GM   PA   BA   HR  RBI   R   OBP    SA    OPS
Ortiz   10   41  .355   2   13   9  .512   .645  1.157
ARod     8   36  .233   2    5   4  .361   .467   .828

Games Won by 1 Run

        GM   PA   BA   HR  RBI   R   OBP    SA    OPS
Ortiz   22  101  .356   7   17  17  .446   .690  1.135
ARod    25  107  .250   6   16  16  .374   .477   .851

Extra Inning Wins

        GM   PA   BA   HR  RBI   R   OBP    SA    OPS
Ortiz    6   28  .261   5    8   7  .393   .913  1.306
ARod     4   21  .400   1    2   2  .429   .600  1.029

Extra-Inning Losses

        GM   PA   BA   HR  RBI   R   OBP    SA    OPS
Ortiz    2   11  .111   0    0   0  .273   .222   .495
ARod     4   17  .063   0    1   0  .059   .063   .121

Games Lost by 1 Run

        GM   PA   BA   HR  RBI   R   OBP    SA    OPS
Ortiz   13   55  .327   7   14  11  .382   .837  1.219
ARod    13   55  .260   6   11   9  .327   .680  1.007

Games Lost by 2 Runs

        GM   PA   BA   HR  RBI   R   OBP    SA    OPS
Ortiz   12   52  .298   3   10   5  .365   .596   .961
Arod    11   46  .205   0    3   2  .326   .256   .582

Games Lost by 3-5 Runs

        GM   PA   BA   HR  RBI   R   OBP    SA    OPS
Ortiz   21   94  .224   1   12   5  .277   .282   .559
ARod    23   97  .298   6   12  10  .381   .560   .941

Games Lost by 6+ Runs

        GM   PA   BA   HR  RBI   R   OBP    SA    OPS
Ortiz   17   70  .220   4    8   6  .343   .458   .800
ARod    16   63  .217   0    2   2  .254   .250   .504

The complete breakdown:

        -------------------Ortiz-------------------
        GM   PA   BA  HR RBI   R   OBP   SLG    OPS
6+ W    20  100  .277  9  33  27  .360  .639   .999
3-5 W   36  161  .359  9  33  32  .478  .680  1.158
2 W     10   41  .355  2  13   9  .512  .645  1.157
1 W     22  101  .356  7  17  17  .446  .690  1.135
EE W     6   28  .261  5   8   7  .393  .913  1.306
EE L     2   11  .111  0   0   0  .273  .222   .495
1 L     13   55  .327  7  14  11  .382  .837  1.219
2 L     12   52  .298  3  10   5  .365  .596   .961
3-5 L   21   94  .224  1  12   5  .277  .282   .559
6+ L    17   70  .220  4   8   6  .343  .458   .800

       ---------------------ARod-------------------
        GM   PA   BA  HR RBI   R   OBP   SLG    OPS
6+ W    20   98  .549 15  46  39  .622 1.171  1.793
3-5 W   38  175  .382 12  32  40  .520  .721  1.241
2 W      8   36  .233  2   5   4  .361  .467   .828
1 W     25  107  .250  6  16  16  .374  .477   .851
EE W     4   21  .400  1   2   2  .429  .600  1.029
EE L     4   17  .063  0   1   0  .059  .063   .121
1 L     13   55  .260  6  11   9  .327  .680  1.007
2 L     11   46  .205  0   3   2  .326  .256   .582
3-5 L   23   97  .298  6  12  10  .381  .560   .941
6+ L    16   63  .217  0   2   2  .254  .250   .504

[Extra inning games are not double-counted, i.e., 2-W is 2-run wins in regulation]

Van:

Is there any question who the actual MVP is? How many of the Yankees' close losses would they have won if ARod had hit like Ortiz in them? How many of the Red Sox' close wins would they have lost if Ortiz had hit like ARod? Is the difference in defensive value that large? ... I'm amazed that people can still say "ARod had the better year at the plate" with a straight face, as if all situations were equal leverage and/or all differences in performance across leverage were random.

An analysis by Win Expectancies shows Ortiz being massively more valuable. The splits by game result are a handy proxy for that, one that requires no explanation of an abstruse methodology. You could simulate the season in Diamond Mind from now until the death of the universe and not get that split.

Still, the methodology isn't perfect. Using the final score is defining "clutch" retrospectively.

Say A-Rod hits three-run home run breaking a 5-5 tie in the 8th; then the Yankees score three more in the 9th, winning 11-5. Although his hit was crucial, the game will be listed as a "non-clutch" game. Likewise, if Ortiz does nothing much as his teammates build a big lead, only to have it nearly blown by the bullpen, then that game suddenly becomes "clutch".

Even if you calculated what the score/situation was in every single plate appearance, that still wouldn't clear things up. Because batting in a tie game is much different if it's 0-0 in the first inning or 4-4 in the eighth. A batter knows that in the first inning, he (and his team) will have plenty of time to score runs, and although it is a "tie game", there is less pressure.

What a batter does in "close and late" situations can be known at the time they are occurring. Close & Late is defined as "in the 7th inning or later with the batting team either ahead by one run, tied or with the potential tying run at least on deck."

Looking at the batting splits for Ortiz and Rodriguez, we see:

       PA  HR RBI  AVG  OBP  SLG   OPS
Ortiz  93  11  33 .346 .447 .846 1.293
A-Rod  91   4  12 .293 .418 .520  .938

Random stat in Ortiz's favor -- with runners at 2nd-and-3rd (at any time):

Ortiz: 7-for-12  1.476 OPS
A-Rod: 2-for 10   .729 OPS

There is also the issue of base-running. Rodriguez runs on the bases, Ortiz does not.

Rodriguez stole 21 bases in 27 attempts (77.8%). A 75% success rate is considered the break-even point; below that, in general, you should not attempt to steal (game situations will vary, of course). ... Ortiz attempted one stolen base -- against Baltimore on July 10 -- and he was safe. 100%!

If stats on taking an extra base (1st-to-3rd, 2nd-to-home) were readily available, Rodriguez would likely get an edge, but again, you would also have to know how many times he was thrown out in those attempts.

So ... after all that, who is my AL MVP? Who is better when it comes to the "actual value of a player to his team ... the strength of offense and defense [as well as] general character, disposition, loyalty and effort"? ... Whose production would I take overall?

I think it should be obvious there is no clear-cut answer (and I'm very curious how close the voting will be) (Over at the Courant, the Yankees writer picks Ortiz and the Red Sox writer tabs Rodriguez).

I compiled a lot of these stats right after the season ended, and when I thought about refining this post, I kind of expected I'd pick Rodriguez in the end. But, assuming baserunning and fielding give neither player an edge (not provable, though), can Rodriguez's slightly-better numbers with the bat offset Ortiz's better numbers when his team needed his production the most?

No.