An Analysis of Two Statistical Measures of Value

I found this on my desktop today. It was never finished, but my goal was to analyze adjusted statistics that are supposed to give a better picture of a player's merit than the stats that most people look at on a day to day or season to season basis. Further explanation will be included. I didn't really include any analysis at the end, but I think the teams are very interesting in themselves. Here it is:

My goal here is to assemble the all time greatest roster based solely on career OPS+ for position players and ERA+ for pitchers. While these two statistical measures are considered as comprehensive as a single number can be in evaluating a player’s performance and, if I dare use such a controversial word, value, I am skeptical about their ability to paint a complete picture. Feel free to comment about any player that you do not think deserves a spot on the roster and offer a replacement. Any general thoughts about the topic would be appreciated as well.

Before we proceed, the stats must be defined. OPS+ is a hitter’s on base percentage plus his slugging percentage (total bases per official at bat) adjusted to the league as a whole and to the player’s home field. ERA+ is similar. It measures the number of earned runs, which are runs scored due to “normal pitching,” that a pitcher allows per nine innings. This basically includes anything that a pitcher has control over. Errors are the major way in which unearned runs are scored. ERA+, like OPS+, is adjusted to league averages as well as the pitcher’s home park.

There are a few things to consider before putting any faith in these statistics. OPS+ falls short by failing to consider a player’s defensive value. While this cannot be ignored in judging a player’s overall worth (I’d take Brooks Robinson over Eddie Mathews, Mike Schmidt, and Alex Rodriguez any day), for our purposes there is simply no way to quantify a player’s complete skill set. OPS+ also ignores base running ability, so it does not really even give a reliable picture of complete offensive prowess. It is important not to downplay this omission. The value of a player capable of stealing fifty bases in a season is incredible.

ERA+ does not, of course, consider a pitcher’s performance at the plate. This is very rarely a significant problem, but it is a problem nonetheless. The more legitimate complaint is that relief pitchers tend to have particularly high scores. Mariano Rivera, for instance, has a career ERA+ fifty points higher than any starting pitcher. At first glance, this ridiculous margin seems to indicate that Mariano Rivera is by far the greatest pitcher of all time. It is unreasonable, however, to conclude that the closer is that much better than the likes of Walter Johnson, Bob Gibson, and Sandy Koufax. It is more likely that the statistic tends to favor relievers, partially because it fails to consider innings pitched.

The simple fact that this roster will be chosen using career statistics introduces a problem. Too many great players have lingered in the Majors long after their greatness has faded. In these career percentage stats, one or two weak years can significantly hurt a player. Because of this, current players like Albert Pujols and Alex Rodriguez who have not yet passed their peaks, may be favored over past legends like Willie Mays, a painful example of a player who delayed retirement.

It is clear that these statistics are not perfect. But they seem to be the best and most complete ones available. So let’s see where they take us.

First of all, this is my pre-statistics roster. No research went into this. These are simply my picks for the best players at each position, based only on my memory and preconceptions. They are not my guesses regarding who has the highest OPS+. Also, I’m refusing to do anything other than one player per position out of fear of making terrible decisions. With only eight spots, I don’t have to worry much about whom I omit.
Starting Position Players
Catcher Johnny Bench
First Base Lou Gehrig
Second Base Rogers Hornsby
Third Base Brooks Robinson
Short Stop Alex Rodriguez
Outfield Babe Ruth
Outfield Ted Williams
Outfield Ty Cobb


All OPS+ Starting Position Players
Note that all players are considered only eligible at the position at which they played the most games.
Edgar Martinez had a higher OPS+ than any other third baseman but played more games at DH.

Position Player Career OPS+
Catcher Mike Piazza 142 (61st)
First Base Lou Gehrig 179 (4th)
Second Base Rogers Hornsby 175 (5th)
Third Base Mike Schmidt 147 (39th)
Shortstop Honus Wagner 150 (31st)
Outfield Babe Ruth 207 (1st)
Outfield Ted Williams 191 (2nd)
Outfield Barry Bonds 181 (3rd)

Bench
The bench is simply the six players with the highest OPS+ that are not the highest at their positions, kind of like a wild card.
Player OPS+
Mickey Mantle 172 (6th)
Albert Pujols 172 (6th)
Dan Brouthers 170 (8th)
Joe Jackson 170 (8th)
Ty Cobb 168 (10th)
Jimmie Foxx 163 (11th)

Pitching Staff
For the staff, I simply took the eleven pitchers with the best career ERA+ regardless of starting v. relief roles. I originally planned to take ten, but ties complicated the situation and rather than introduce any subjectivity and choose one somehow, I took both Clemens and Santana.

Pitcher ERA+
Mariano Rivera 202
Pedro Martinez 154
Lefty Grove 148
Trevor Hoffman 147
Walter Johnson 147
Dan Quisenberry 146
Ed Walsh 146
Hoyt Wilhelm 146
Joe Wood 146
Roger Clemens 143
Johan Santana 143
I believe that the bias toward current players was really prevalent here. Santana, for instance, simply has not yet played at any time other than his prime.