Clemens or Maddux? Mays or Mantle?

Editor's note: Rob Wood is a guest columnist who is filling in for Rob Neyer, who is on vacation.

Debating who is the greatest probably goes back to the time of Cain and Abel. I don't know which of those two guys was the better player, but spirited debates along these lines date back to the dawn of professional baseball.

Ranking baseball's all-time greats has always been a favorite pastime. Rather than settling the issue, the rise of statistical analysis in the 1970s poured fuel on the fire. Indeed, sabermetrics has spawned a host of its own all-time great rankings using its different alphabet-soup performance measures (TPR, OPS, EQA, OWP, WAA, etc.).

Pioneers in the field like Pete Palmer and Bill James have developed several methods to measure a player's performance. However, every stat (and probably every analyst) has a peculiar perspective or bias. Bias is largely inescapable, insofar as no single stat can capture everything that happens on a ballfield. To mix metaphors, the Holy Grail of sabermetrics is the Grand Unified Stat.

Fans and analysts alike should be glad that Gus is unattainable. After all, it is fun to debate whether Roger Clemens, Greg Maddux, Tom Seaver, Warren Spahn, Sandy Koufax, Randy Johnson, Pedro Martinez or anyone else is the greatest living pitcher and whether Willie Mays or Mickey Mantle was the greater all-around player.

The baseball survivor island
The Society for American Baseball Research (SABR) is chock full of members who love to debate these issues. Partly in response to a few well-publicized yet dubious lists of all-time greats announced at the turn of the century, in early 2001 Tom Hanrahan invited fellow SABR members to form an internet community that would discuss, debate, and vote on baseball's all-time best players. Modeled after the television show Survivor, the group would send the 100 greatest players of all-time to a mythical island, and then vote them off one-by-one until the ultimate survivor is crowned the game's all-time greatest player.

A group of 20 devoted fans (a good mix of ages, hometowns, opinions, and statistical devotion) joined Tom and had the time of our lives. Over the next year, lively discussions on every topic imaginable raged across the Internet (in total there were several thousand messages). Co-leader Justin Kubatko developed a website that housed a great deal of statistical data on each of the players who were initially voted on to the island, as well as a complete record of the weekly voting.

Tom and Justin's idea was that an informed consensus of expert opinion should yield a highly credible ranking of baseball's all-time greats. The biases and idiosyncrasies inherent in each statistical measure and each individual voter would be washed away by the overall consensus view.

At the outset of the year-long exercise, each voter submitted his own personal list of the top 100 players of all time. The 100 players who were mentioned on the most lists were sent to the mythical Survivor Island. The only criteria that the group imposed to govern initial selections was that only MLB performances could be considered (which meant no Negro Leaguers made the island).

The positional balance of the top 100 players mirrors Hall of Fame selections, with 30 starting pitchers, 11 first basemen and right fielders, and only six third basemen.

The era balance among the top 100 is tilted toward the present day, reflecting improved quality of play as well as a greater number of major league teams in the modern game.

After the initial group of players was shipped to the island, each week one or more players were voted off, starting with the worst (least great) among those remaining islanders. Every voter utilized his own criteria to evaluate players. Some voters put a lot of weight on peak value while others put a great deal of weight on career value. Some voters only looked at what a player did, while other voters considered what a player might have done (e.g., in the absence of injuries or military service, in other eras or parks, etc.).

All-time top 10
Let's start by removing any suspense regarding the Ultimate Survivor. Babe Ruth was unanimously voted the game's greatest player. Everyone knows about the Babe, so I don't need to document his awesome career. If any voter had any doubt as to Ruth's pre-eminence, he only had to consider that Ruth was also one of the game's greatest pitchers with the Red Sox prior to joining the Yankees. While it is possible to exaggerate how great a pitcher Ruth was, his four-plus years of pitching more than tip the balance in his favor.

Baseball Survivor's Top 10
Babe Ruth
Honus Wagner
Barry Bonds
Willie Mays
Ted Williams
Ty Cobb
Walter Johnson
Mickey Mantle
Hank Aaron
Stan Musial

While our top 10 did not coincide with any of the individual voters' top 10s, this consensus ranking is a good representation of our collective views. Indeed, it is actually a very credible top 10 (personally, I would swap Mays and Wagner).

Active players
When we began in early 2001, no voter had Barry Bonds in his personal top 10. However, Barry's unbelievable 2001 and start to 2002 was compelling evidence that Bonds really belongs among the game's elite. Ruth and Williams are still the two greatest offensive forces the game has ever known, but Bonds may well be the best of the rest.

You may wonder if we got carried away by Barry's contemporaneous exploits and voted Bonds too high. I wonder about this myself, but other active stars do not seem to be significantly over-valued. In fact, the group was conservative in evaluating active stars; many voters even declared that they give absolutely no credit for future performance. Here is where all the active players finished:

3. Barry Bonds
14. Roger Clemens
20. Greg Maddux
29. Rickey Henderson
34. Mike Piazza
37. Randy Johnson
38. Ken Griffey Jr.
53. Frank Thomas
58. Pedro Martinez
66. Jeff Bagwell
91. Kevin Brown
94. Barry Larkin

Mike Piazza was voted the best catcher ever, over Johnny Bench, Yogi Berra, and Mickey Cochrane. Piazza is clearly the greatest hitting catcher ever, though his defense leaves a lot to be desired. This was the result that I argued against the loudest. Personally, I think Johnny Bench is the greatest catcher ever. Offense is easier to measure than defense, especially for catchers, which works in Piazza's favor.

Roger Clemens was voted the third-best pitcher ever (see below). Recently-retired Mark McGwire was voted the third-best first basemen, behind Lou Gehrig and Jimmie Foxx. Cal Ripken was voted the second-greatest shortstop, albeit far behind Honus Wagner. Despite these rankings of modern players, we may have overlooked some other deserving active players, most notably Alex Rodriguez.

At the time of the initial voting early in 2001, Rodriguez's potential seemed unlimited, but his all-time great status was not clear. Today, with almost two more seasons in the bank, it is clear that Alex certainly does rank among the all-time greats.

Mays vs. Mantle
The Mays-Mantle debate is usually pitted as Mays' career value vs. Mantle's peak value. Mays contributed more to his teams over the course of his career than did Mantle by virtually every measure. However, Mantle had a higher peak.

The Baseball Survivor group came down on the side of Willie Mays, relegating Mantle to the eighth slot. There were several compelling factors that tipped the balance in Mays' favor. First, Mays is one of the greatest defensive players of all time; by our estimates, Mays contributed 15-20 more defensive wins in his career than did Mantle (a good center fielder in his own right). Call this about one additional win per season for Mays.

Second, research was presented that indicated that the average National League player of the 1950s and 1960s was about 0.5 wins better than the average American League player. Considering these first two factors, Mays' peak is now seen to be only slightly lower than Mantle's.

Third, it was believed that Candlestick Park had a more deleterious effect on Mays' home runs (and offensive performance in general) than traditional analysis reflects. Research suggests that Candlestick depressed home runs for right-handed hitters by about 10 percent from 1960 through 1971 (while increasing home runs for lefty hitters by about 25 percent). In addition, Mays modified his stroke when the Giants moved to Candlestick, due to the strong prevailing winds blowing in from left field (across to right field). Mays became more of an inside-out swinger and used this stroke on the road as well as at home for the rest of his career. I fervently believe that the combination of Candlestick's winds and Mays' stroke change cost Willie 30 or 40 home runs in his career (the Korean War cost him plenty too).

Finally, as mentioned above, some of the voters considered how the player might have fared in different eras throughout baseball history. Mays would have been a superstar in any era. Mantle would surely have been a star in all eras too, but Mantle's value was probably highest in the league and era in which he actually played.

Clemens vs. Maddux
Here are the Baseball Survivor top 10 pitchers of all time:

Walter Johnson
Lefty Grove
Roger Clemens
Cy Young
Pete Alexander
Greg Maddux
Christy Mathewson
Tom Seaver
Warren Spahn
Kid Nichols

Walter Johnson is universally acclaimed the game's best pitcher, and Lefty Grove is often considered the runner-up. Perhaps surprisingly, we voted Roger Clemens the third greatest pitcher ever.

The honor of greatest living pitcher comes down to Clemens vs. Maddux. Their overall stats are very close, with Clemens leading in most career totals and Maddux leading in most rate stats. Maddux's sustained peak was higher than Clemens'.

Three factors tipped the scales in Clemens' favor, though.

First, Clemens' winning percentage relative to his teams is significantly higher than Maddux's. I developed a stat to estimate what the pitcher's win percentage would be on a 500 club, based upon his own and his team's won-lost records. Clemens is estimated to have a .663 winning percentage for a 500 club, and Maddux a .631 winning percentage. In fact, Clemens ranks behind only Randy Johnson in this stat.

Second, Clemens strikes out significantly more batters than Maddux. Through 2001, Clemens averaged 8.6 strikeouts per nine innings and Maddux 6.4, despite Clemens pitching in the American League, where strikeouts are less common. And strikeout pitchers deserve slightly more credit for their run prevention than control pitchers.

Finally, Clemens has pitched in front of worse defenses than has Maddux over the course of their careers. Despite this fact, the Rocket has given up fewer unearned runs than has Maddux.

Maddux is about 3-½ years younger than Clemens (Maddux is 36, Clemens just turned 40), so Maddux may well eventually surpass Clemens as baseball's greatest living pitcher. But given what we know today, I think the Baseball Survivor group got it right. Roger Clemens and Greg Maddux are both all-time greats, with Clemens being a smidgen greater.

For the complete Top 100 and more information about the Baseball Survivor exercise, click here.