The Baseball Guru Extrapolation Method by Eric Gartman (original) (raw)

Also Read:2006 update

[Updated Through the 2003 Season.]

RATING THE TOP BASEBALL PLAYERS OF ALL TIME:

THE EXTRAPOLATION METHOD

by Eric Gartman

The New Bill James Historical Baseball Abstract introduces a novel new way to evaluate players: The Win Shares System. What is revolutionary about the system is that it does not use statistics that are subject to variation as the game changes. It is based on how many games a player helps his team win, based on his performance. Since the ratio of win shares to wins is always 3-1, the system treats good players on bad teams as equally as bad players on good teams. But most of all, it allows us to fairly look at players across eras, in one single integer. The system is brilliant. James application of it, however, is not. Its biggest flaw, as many people have noted, is that it does not evenly rate pitchers' contributions. But the system does allow for fair assessment of position players.

In evaluating players, James creates a six-step system to include all relevant data. Simply adding up raw win shares and rating the players against each other does not work, according to James. As proof of this, he cites the fact that Rusty Staub has almost as many Win Shares as Joe Dimaggio. So to even it out, he weighs a players' three best seasons, five best consecutive seasons, and win shares per 162 games. It is these three values that skew his system, but more on this later. His fifth element is a time line adjustment, and the sixth a subjective element. The time line adjustment is a good factor, since players do tend to get better over time.

The problem with the 2, 3, and 4 factors is that they overate a player's prime, while vastly underrating his career stats. This leads to some skewed ratings of players. James himself admits to one: His system rates Mickey Mantle higher than Ty Cobb, mostly since Mantle's prime numbers were so good. Similarly, the system underrates Hank Aaron, while overrating Honus Wagner, and overrates Mark McGwire while undervaluing Eddie Murray. The reason these three variables were introduced was to compensate for a situation that rarely occurs: Players like Dimaggio missing significant time due to wartime service. A better system would look at total win shares while extrapolating what a player would have done in the years he missed, based on performance before and after the time missed. This would give us a rough total win shares, while not overcompensating for peak values. There is no inherent reason why peak values are more important than total values. Longevity is a key factor in rating a players total worth. There are players, however, who hang around late in their careers to pad their stats, like Pete Rose and Rickey Henderson. And some greats opt to quit early, like Hank Greenberg and Joe Dimaggio. Others, like Mickey Mantle and Lou Gehrig, were forced out to due other factors. In comparing these players to those with long careers we can examine performance until the age when they retired, in these cases, 36. One may argue that guessing what a player would have done is unscientific. Surely it is. But it is far better than skewing an entire system to accommodate for a couple of players who missed time.

My system proposes that we weigh total win shares as the most important variable. Players who missed time will have a projected total win shares, which will count just as much. There will be a slight time line adjustment, and some players who hung around too long will be penalized as well. There will only be tiny subjective element that will come into play in close cases. I have chosen not to include Negro league players due to lack of stats. I also did not include pitchers. I don't think its fair to rate pitchers vs. hitters. They should be rated by against their own kind of player (many try to attempt to replicate these feats by training withHomerun Monkey baseball bats).

Here then is the new rating system, along with each players total win shares:

1. Babe Ruth- With 758 Win Shares, Ruth comes out as the best ever.

2. Ted Williams- Williams had 558 actual win shares. Extrapolating for the years he missed he ends up with 728 projected win shares. Of all the players on the list, Williams is by far the most affected by the extrapolation technique, ending up with an additional 170 win shares. How I arrived at this figure must therefore be explained. Williams missed three years due to World War Two, 1943-45. His win shares for the two years before were 42 and 46. For the two years after, 50 and 44. I assigned Williams 40 for each year for a total of 120. This is clearly a low figure, based on his performance both before and after. Williams also missed most of 1952 and 53 due to the Korean War. He had 34 win shares in 1951 and 29 in 1954. I assigned him 29 win shares on top of the one he had in 1952, and 20 on top of the 10 in 1953. None of these estimates seems improbable to me. In fact, they seem quite likely. Adding it all up, he is second only to Ruth.

3. Ty Cobb/Willie Mays- Cobb had 726 total win shares and Mays had 642 total win shares, but missed most of 1952 and all of 1953 in the service. In 1951, his first year he had 19 win shares, and upon returning had 40 in both 1954 and '55. Assigning Mays shares of 30 and 35 for the years he missed seems reasonable, bringing his total to 701. Mays finished 25 shares behind Cobb, and adjusting for the time line, these two seem about even, leaving open the question about who was the greatest all-around player ever

5. Hank Aaron- Aaron had 643 win shares, and did not miss any time. He made the most of his career, and is the all-time leader in Total Bases, Home Runs, and Runs Batted In.

6. Honus Wagner- The Dutchman finished with 655 total win shares, but the time line favors Aaron. James rates him higher than Cobb because his peak years were better, but 70 career win shares is a big difference.

7. Stan Musial- Stan the Man finished with 604 win shares. He missed one year to World War Two, but his stats the previous two years were helped by weak competition, with many of the best players off to war. I'd adjust Musials overall totals slightly giving him 16 extra career win shares, finishing with 620.

8. Barry Bonds-Bonds has 611 career win shares. His 2001 campaign is now recognized as the greatest offensive season of all time, and garnered him 54 total win shares. He followed up with a stellar 2002 season, gaining 49 win shares, and 39 in '03. The real question is where this man will end up. He could finish his career even higher than Hank Aaron, and challenge his godfather, Willie Mays, for the title of best all-around player. But during the 2003 off season, allegations of steroid use have dogged Bonds, and could affect our perception of him.

9. Tris Speaker- Perhaps the most underrated player of all time, he collected 633 win shares but was overshadowed by Cobb and Ruth. A whopping 118 of his total win shares comes from fielding, further hiding his true value. Speaker played center field very shallow so that no balls would fall in front of him, but was able to catch those behind him through his excellent speed, much like Andruw Jones would many years later. The time line favors Musial and Bonds over him.

10. Mickey Mantle- Mantle finished with 565 win shares. James rates him higher than me since his prime years were so good. Mantle was the best player in the league every year from 1954-64 except for 1963. When he retired at age 36 he was still posting good numbers.

11. Lou Gehrig-Lou Gehrig finished with 489 career win shares. That figure would have been much higher had his career not been cut short due to a tragic illness at age 35. We will never know how much higher, and we cannot extrapolate based on performance afterwards. But in the two years prior to 1939, Gehrig had 36 and 25 win shares. Rating Gehrig is a very difficult task. I believe rating him higher than his total win shares is in order.

12. Eddie Collins-Eddie Collins finished with 572 win shares, the most of any second basemen, and enough to rate ahead of Joe Morgan as the greatest second basemen ever. Collins is generally not thought of as one of the all-time greats, but his defense was outstanding, and he had a long, productive, career.

13. Joe Dimaggio-Joltin Joe is the other player who, along with Williams, was most affected by the extrapolation method. Joe D. finished with only 387 win shares. Factoring in for the three seasons he missed, I'd add about 100 win shares for a total of 487. But Dimaggio also retired early at 36, and could have racked up more win shares. Putting all this together, Id rate him somewhat higher than his totals suggest.

14. Joe Morgan-Morgan finished with 512 win shares, not quite enough to overcome the time line factor which favors him over Collins. He had more career win shares than Dimaggio, but not through age 36.

15. Frank Robinson-Robinson finished with 519 win shares, a few more than Dimaggio, but unlike Joe, Frank stuck around a few extra years.

16. Pete Rose-Of all the players on this list, Rose stayed around perhaps the longest to pad his stats. He finished with 547 win shares, but should have left the game long before he did.

17. Rickey Henderson-A short stint with the Dodgers gained Rickey one share for a total of 535 career win shares. But a lot of those are in his later years when he really should have been retired but stuck around to set some records.

18. Mel Ott-Ott had 528 career win shares, aided a bit at the end by weak competition in World War Two

Top First Basemen

1. Lou Gehrig- With 489 Win Shares, Gehrig is easily first

2. Jimmie Foxx- Foxx weighs in with 435 Win Shares

3. Eddie Murray-Murray finished with an impressive 437 win shares. But since he did not have huge peak years, James rates him lower. Murray was consistent year in for the length of his career, but never brilliant. But consistency is important, as is longevity.

4. Johnny Mize- Mize finished with only 338 career win shares, since he served time in the Second World War. Before and after he averaged 30 win shares a season, so I extrapolate a total of 90 for the three years he missed, leaving him with 428 win shares

5. Hank Greenberg-Greenberg had a famously short career, missing more time for the Second World War than just about any other player. He missed nearly every game from 1941 until the end of the 1945 season and finished with a mere 267 win shares. Counting in for all the time he missed, I'd assign him an additional 140 win shares, leaving him with 407. He also left the game earlier than he could have, retiring at age 36.

6. Willie McCovey- Stretch finished with 408 win shares, but I'd give Greenberg the edge since McCovey played into his 40's.

7. Rafael Palmeiro-Years of consistency have gained Raffy 372 career win shares. He had yet another solid season in 2003, with 19 win shares. Now in his late 30's, he still swings a potent bat, but time may be running out.

8. Harmon Killebrew-The Killer finished with 371 career win shares

9. Jeff Bagwell-With 22 in 2003 Bagpipes now has 362 career win shares. He's still young enough to add to those totals

10. Mark McGwire-Big Mac finished with 350 career win shares. Injuries plagued him throughout his career, and he eventually was forced to retire at age 37 on account of his bad knees. James rates him third, mostly due to his monster peak years, but what good is a great player who can't play? McGwire spent so much time on the DL that he must be ranked lower

11. Frank Thomas-The Big Hurt had a solid year, and added 23 win shares in 2003 for a total of 347 career win shares..

Top Second Basemen

1.Eddie Collins-Collins is first with 572 win shares

2.Joe Morgan-Morgan finished with 512 win shares.

3.Rogers Hornsby-Most would rate Hornsby as the greatest second baseman of all-time, but he wasn't nearly as good defensively as the other two, and finished with only 502 win shares.

4.Napoleon Lajoie-Larry finished with 496 win shares. Some have overrated him due to misreading his fielding stats, however.

5.Jackie Robinson-Jackie Robinson finished with only 257 win shares, but he didn't start his career until age 28. His prime years are slightly better than Charlie Gehringer and Rod Carew. I would guess he would have finished slightly higher than them, maybe around 400, or more, win shares.

6/T. Charlie Gehringer/Rod Carew- Gehringer finished with 383 win shares, Carew with 384. I don't see how we can distinguish one from the other.

8. Craig Biggio- Biggio now has 377 with a strong 20 win share effort in '03. According to James, Biggio finished the 90's second only to Bonds in total win shares. His true value has been hidden by the cavernous Astrodome, which hurt his run production, and by the fact that his "little stats" are so good. Biggio is now ahead of Alomar, who has long been considered the better second baseman. Biggio is now a centerfielder, and was also a catcher for a few years, but we have to put him somewhere.

9. Roberto Alomar- Alomar followed a huge dropoff in 2002 by another poor season in '03, adding a mere 12 win shares. Robbie already has 373 win shares, and a plaque waiting for him at Cooperstown, but he seems to be fading quickly.

10. Frankie Frisch- The Fordham Flash finished with 366 win shares.

11. Ryne Sandberg- Sandberg finished with 346 career win shares

Top Third Basemen

1. Mike Schmidt-With 468 career win shares, Schmidt is the top third sacker

2. George Brett-Brett had 432 win shares but the time line favor him over Matthews

3. Eddie Matthews-Matthews had 447 career win shares, and a good case can be made for him over Brett

4. Paul Molitor-Molitor had the fourth highest total with 412 win shares, but James ranks him a mere eight. But Molitor was consistent, and continued to be productive late into his career. Of course his figures are aided by the advent of the DH, so it is reasonable to take some points off for him.

5. Wade Boggs- Boggs had 394 career win shares. After Boggs, there is a big dropoff.

Top Shortstops

1. Honus Wagner-With a whopping 656 win shares, Wagner is far ahead of the pack

2. Cal Ripken-Ripken finished with 427

3. Robin Yount-Yount finished with 423, but he played the second half of his career in the outfield

4. Arky Vaughan-Another overlooked star, Vaughan finished with 356 win shares but missed three seasons due to the Second World War. He was winding down his career, and I'd estimate about 60 missed win shares for a total of 416. After Vaughan, there is a big dropoff. And how about the bright young shortstops of today? A-Rod had 252 through 2003 at the tender age of 28. He could easily surpass Ripken, and might even challenge the Dutchman himself for the top spot.

Top Leftfielders

1. Ted Williams-Williams with his estimated 728 win shares is first

2. Stan Musial-Stan the Man's 620 estimated win shares rates second

3. Barry Bonds-Bonds had 611 win shares through the 2003 season, and is now chasing Musial.

4. Rickey Henderson-Henderson had 534 win shares through 2002

5. Carl Yastrzemski-Yaz finished with 488 total win shares

Top Centerfielders

1. Willie Mays/Ty Cobb-Too close to call in my opinion

3. Tris Speaker-Speaker had 633

4. Mickey Mantle-The Mick finished with 565, despite retiring at age 36 from injuries (and drinking)

5. Joe Dimaggio-Joe D. Would have had around 487

6. Duke Snider-The Duke of Flatbush finished with 352

7. Ken Griffey, Jr.-Junior already has 324 career win shares, but his best days seem long behind him.

Top Right fielders

1. Babe Ruth-The all-time leader, the Babe had 756 win shares

2. Hank Aaron-Hammerin' Hank finished with 643

3. Frank Robinson-Robinson had 519. Robinson, Rose and Ott are all very close to each other. I believe Robinson was the best of the three, but a good argument can made for all

4. Pete Rose-Rose had 547 win shares

5. Mel Ott-Ott had 528 win shares

6. Reggie Jackson-Jackson finished with 444 career win shares

Of the active players on this list Biggio, Bagwell, Palmeiro and Thomas all put up fairly good numbers, especially for aging players. Alomar fared poorly, while Bonds was injured once again. Bonds had another huge year, and continued to vault his way up the list. Gary Sheffield had a monster year, and it may not be long before he finds his way on this list.

Also Read:

Eric Gartman (baseball business) -Rating the Top Baseball Players of all Time: The Extrapolation Method (updated to include 2006 season)
Michael Hoban (Great player analysis) -Pitchers with Hall of Fame Numbers Position Players with Hall of Fame Numbers
Bruce Baskin (Latin Baseball) -Maestros of Mexico: Hector Espino and others
Craig Tomarkin (the Guru) -Baseball's Thrity Greatest Foreign Players (who never played in the MLB)