How we ranked the greatest players in baseball history
There are many ways to rank baseball players, we’re aware of this. Many books have been written using a variety of methods. We didn’t try to rank our players based on how others looked at them, we used a method that made sense to us.
We did not rely solely on statistics, though stats were the framework by which we started our ranking process. Our raw statistical formula helped us identify the best 150-200 players at each position, and since we were interested in ranking the top 50 players at each position (and the best 100 overall), we started there and whittled them down.
The basis for our statistical formula is a stat called WAR. WAR (Wins Above Replacement), is generally accepted among the baseball stat community (often to referred to as SABRmetrics) as the best single measurement of baseball players. Over at FanGraphs, a website dedicated to baseball statistical analysis, Steve Slowinski wrote that WAR:
Is an attempt by the sabermetric baseball community to summarize a players total contributions to their team in one statistic. WAR basically looks at a player and asks the question, If this player got injured and their team had to replace them with a minor leaguer or someone from their bench, how much value would the team be losing?
We understand that there are many baseball fans who hate SABRmetric analysis. We don’t care. In order for our rankings to be taken seriously by serious baseball historians and statisticians we needed to use a method that was credible. If you think WAR is ridiculous, that’s fine. Start your own website and make your own rankings.
We used a combination of career WAR and WAR7 (the total of a player’s WAR in his best seven seasons). We also added value for a more narrow peak performance using the player’s best three seasons based on WAR. Some of you may be familiar with JAWS, a stat devised by Jay Jaffe. It’s fairly simple: it divides career WAR by WAR7 to arrive at a marriage between a player’s career value and his peak value. Ok, fair enough, but we feel peak performance is very indicative of greatness. So we went a step further.
Our statistical formula:
Where WAR3 is the player’s top three seasons based on WAR.
Using those multipliers (2 for Career WAR and 1.75 for WAR7) we are able to arrive a slightly larger number while also giving more credit for a player’s peak.
JAWS rates Career WAR and WAR7 equally, in our method it comes out to about 55 percent for peak value and 45 for career value for most players. There are some players with very high peaks and lower Career WAR who see that balance shift a bit, but not as much as you think. The 2x multiplier for Career WAR gives quite a head start to players who were good for a long stretch of time.
Coincidentally, the system has a useful side effect: Babe Ruth ends up with a score of about 500. He’s the only player to top 500 (though one controversial player comes very, very close). We’re comfortable with The Babe being the greatest player of all-time, and a score of 500 seems fitting to “top off” the scale. Conveniently, anyone with a score of 400 is a legendary player and 300-399 is in the next tier of greats. Many players who score 200-299 are in the Hall of Fame, while those players who score between 100 and 199 were all-stars at least once and had decent careers. Of course within those ranges there are sub-ranges. A player at 150 is much better than one at 100. A player who scores 250 is almost certainly a Hall of Famer.
The only players who score 250+ who are not in the Hall of Fame are the steroid guys, Pete Rose, active superstar Albert Pujols, and candidates Mike Mussina and Curt Schilling.
We used something we call Championship WAR (CHWAR). This is how it works: we add up all of the WAR accumulated by a player in seasons in which he was on a pennant-winning team. The player gets 25 percent of their CHWAR added to their total score. This resulted in helping players who had several good seasons for pennant-winning teams, such as Mickey Mantle and Lou Gehrig. However, most of the players who got the biggest pennant adjustment, were already among the top players at their positions. It was players further down the list who were helped by the pennant adjustment, as it served to work as a tiebreaker with players who didn’t play for many pennant-winning teams. Among the notable players helped by the pennant adjustment are Roger Maris, Gil McDougald, Willie Randolph, Carl Furillo, Boog Powell, Jack Barry, Paul Blair, Chase Utley, and Sal Bando.
We used subjective opinions to discount players who were suspected of using PEDs, but we didn’t do much to change their positions. We admit that we could be wrong in some of our rankings in regards to PEDs, but we felt we had to penalize Barry Bonds, Roger Clemens, Manny Ramirez, Rafael Palmeiro, and Alex Rodriguez. We also lowered Mark McGwire and Sammy Sosa and a few others. Your opinion may vary.
For those players who merely had rumors swirling around them, such as Ivan Rodriguez, Jeff Bagwell and Mike Piazza, we did not pretend to know if they “cheated,” we simply rated them as if they did not use steroids.
Early Baseball Competitive Balance Adjustment
We made an adjustment for players who played most of their careers in the late 19th Century and early part of the 20th Century, when the level of competition was lower and rules were also changing that made quite an impact. This impacted quite a few starting pitchers. We simply do not think it makes sense to have 10-15 pitchers who toiled during that era rank in the Top 100. It would be mostly due to the fact that they were able to start 45-55 games a year in leagues where the competition was not well balances, which increased their impact on the field.
You’ll notice the absence of players who played their entire careers prior to 1900. In our opinion, baseball was such a different game in the 19th century that it’s impossible to compare the stars of that era with those who played later. At a later date we intend to compile a list of the top 50 players of the 19th century.
We also made an adjustment for players who started their careers prior to integration, and thus only faced white players (for the most part). The adjustment was significant, because we believe that prior to the 1950s, Major League Baseball was not nearly as good or competitive as it would become when black and latino players were welcomed to the game. It was easier for Nap Lajoie and Walter Johnson and Babe Ruth to dominate their eras when the competitive balance of their leagues was lower.
The Era Adjustment was significant, we did it this way: we subtracted the year the player debuted from 1951 and multiplied that number by .01. So, for a player who debuted in 1931, the number would be 20 or .20, which we multiplied by their career WAR and subtracted from their career WAR. In essence, a player lost 1% of their career WAR for every year they debuted before 1951.
This adjustment proved critical for players who debuted early in the 20th century or late in the 19th century. But we strongly believe the game has evolved and gotten much better since it integrated. This adjustment helped remove a bias toward “old time” players. Without the Era Adjustment, our lists would have had many, many more players who played from 1895 to 1925. That doesn’t make sense.
Loss of playing time due to war or the color barrier
We made major adjustments for those players who missed time for reasons beyond their control, such as World War II or the color barrier. This especially helped Bob Feller, Johnny Mize, Roy Campanella, Hank Greenberg, Joe DiMaggio, Minnie Minoso, Charlie Keller, Joe Gordon, Enos Slaughter, Pee Wee Reese, Joe Gordon, and a few others. The player who was probably most helped by the “war time” adjustment was Ted Williams, who missed significant time in both WWII and the Korean War.
Obviously, we felt we had to make adjustments for those players who were denied entry to the major leagues due to the color of their skin. This helped Jackie Robinson, Roy Campanella, Larry Doby, Minnie Minoso, and a few others.
We did not feel we could fairly judge the talent of players who played all or most of their careers in the negro leagues. We have no doubt that Oscar Chrleston, Josh Gibson, Cool Papa Bell, Satchel Paige, and several other black players would have performed very well in the major leagues, but we cannot speculate on how well because their isn’t enough data to help us compare them to the talent level in MLB prior to 1947. If someone has a method for doing this and wants to share it with us, we’d love to rectify this omission.
Injury and short careers
We did not make adjustments for players who retired early due to injury. Injuries are part of the game, and if a player was unable to perform because his body broke down, that’s part of how we rated them. As a result of this approach we probably are going to take a lot of heat for how we rank Sandy Koufax. We acknowledge that at his very peak Koufax ranked among the greatest pitchers ever. But his peak was really only four seasons, and he had a fifth season where he was pretty good, but not at the level of his zenith. We valued peak over career, especially with starting pitchers, but 4-5 years alone is not enough to put Koufax in the upper echelon of pitchers. Sure, had he been able to pitch after the age of 30 he would have added to his resume, but he wasn’t able to. And it wouldn’t be fair to give him credit for doing something he didn’t do because of injury.
Postseason Performance Adjustment
There’s a lot of debate about whether it’s appropriate to consider postseason performance when rating players over their careers. It’s a dicey subject. On the one hand, postseason play is a result of opportunity, and some great players, like Ernie Banks, Rod Carew, and Ken Griffey Jr., had little opportunity to show what they could do in October baseball. Others, like Reggie Jackson, Catfish Hunter, and Derek Jeter, were in the playoffs a hell of a lot.
Ultimately, we considered postseason performance as a small tiebreaker when players were clustered. Starting pitchers, who can really impact games, of course, were given a bit more consideration for postseason play. Curt Schilling and Bob Gibson are two players who got some love in our ratings for their performance in postseason baseball.
We also felt we had to reward players who served as player/managers and had success. That’s a unique role to fill and an important one. Tris Speaker, Frank Robinson, Rogers Hornsby, Mickey Cochrane, and Lou Boudreau got a slight bump for being great at it, among a few others. It wasn’t a major ranking factor, though.
This adjustment was minor. For example, most players received a +1 WAR adjustment to their score. Only 27 players received +2 adjustments, and the highest adjustment was for Reggie Jackson, who got +4. Generally, this adjustment did not make a difference in overall rankings, but it pushed a few players up one slot.
Off the field impact
We did not consider any special off-field contributions, which didn’t help Jackie Robinson, Curt Flood, or others who made an impact that way. We didn’t consider the “good guy factor” either. Nor did we penalize anyone for being a jerk. Dale Murphy rates where he does in center field because of what he did between the lines, not because he was one of the nicest men to ever play professional baseball. Kevin Brown, Jonathan Papelbon, and Gary Sheffield could each be described as a “royal pain in the ass,” but they still made the top 100 for their positions.
How we chose the primary position for players who played multiple positions
There are a few players who could have been ranked at multiple positions. How did we choose where to rate them? We used common sense and also looked at games played.
But games played wasn’t the defining criteria. For example, Rod Carew played more games at first base than second, but we ranked him at second. Similarly, Robin Yount played more in center field, but he ranks here at shortstop. Those two players were better suited for the position in which they started their careers, as was Ernie Banks, who actually played more games at first base than short. Joe Torre was the same way. For Pete Rose we settled on left field, though he played hundreds of games at second, third, first, and right field. Left field was where he ranks best.
Outfielders are special too: some of the players may have had a few more games in right or left, etc., but we placed players in the outfield spot that they were most associated with, for the most part. Dave Winfield and Andre Dawson, however, are players who put in a lot of time at two outfield spots, but we slotted them where they played the most games.
On relief pitchers
We’ve seen two “Top 100” lists that include relief pitchers. We simply can’t see how a pitcher who pitches only 600 to 1,500 innings (the range for the best relievers in history) can be one of the Top 100 players in baseball history. That’s not enough playing time to make it. Even Mariano Rivera, the greatest relief pitcher of all-time, only faced about 5,000 batters in his career, and Hoyt Wilhelm, who pitched when relievers often hurled multiple innings, only had about 7,000 batters faced as a reliever.
Relief pitching has become more important in baseball, but a close examination of performance shows that most starting pitchers could go to the bullpen and do at least a passable job as a closer.
Three pitchers who spent all or part of their careers as relievers rated in our Top 100 Pitcher rankings (Mariano Rivera, Dennis Eckersley, John Smoltz, and Hoyt Wilhelm). But only Rivera spent all of his time as a reliever, and he still rated in the middle of our all-time list. Relievers simply do not play enough to make an impact on our lists, but if the usage trends change in the future (or if another Mo comes along), they may appear.