Defining Excellence

One of the common debates in the Hall of Fame is the distinction between peak performance as compared to cumulative performance.  Put another way, we’re looking at a distinction between a player who was excellent versus one who accrued similar value by simply playing more.

 

The first thing to look at is a seasonal breakdown of WAR.

Let’s take a look at an example of two fictional 60 WAR players:

1. Player 1:    0, 1, 2, 2, 5, 8, 10, 10, 12, 10
2. Player 2:    2, 3, 3, 4, 4, 4, 5, 5, 4, 4, 3, 3, 3, 3, 3, 2, 2, 2, 1

The two players are equal in terms of value, but the shapes of their careers are different.  The first player is the Koufax model: a superior prime, but a short career with minimal value outside his great years.  The second is closer to the Ted Lyons model: a long career with a lot of solid years and a few very good ones, but nothing that comes close to the first player’s prime years.

 

Next, let’s look at two similarly valued seasons by a couple of contemporary first basemen:

1. Mark McGwire, 2000:  89 G,  .305/.483/.746,   32 HR,   4.5 WAR
2. Rafael Palmeiro, 2001:  160 G,  .273/.381/.563,   47 HR,   4.2 WAR

Palmeiro had the higher totals because he was healthier, but McGwire managed almost the same WAR because his performance on a per-game level was considerably higher.  On a value level it’s similar, but the shapes of the seasons were different.

 

Finally, let’s look at the offensive/defensive breakdown of some seasons:

1. Ozzie Smith, 1989:   3.2 oWAR,  3.1 dWAR
2. Dick Allen, 1967:  7.4 oWAR,  -1.5 dWAR
3. Mark Belanger, 1975:  0.2 oWAR, 3.3 dWAR
4. Joe Torre, 1971:  9.4 oWAR, -2.6 dWAR

When we break a player’s value into offense and defense, we accomplish two things.  The first is that we have a way of separating isolated excellence from total value.  It gives us a better understanding of a player’s strengths and weaknesses, rather than clumping everything together into one number.  The other thing it accomplishes, however, is that it’s a way to address the variance in defensive metrics.  Sometimes a player’s peak WAR seasons will appear higher or lower than expected because his best or worst defensive seasons will align unusually well or poorly with his best offensive years.

I want to leave a little bit of leeway in terms of how to define excellence, so I’ll be presenting three different approaches:

1. Compare a player’s seasonal WAR with the average player, and credit the player with gradually increasing value per point above the average
2. As above, but prorate the intervals of increasing value base on the playing time
3. Separate oWAR and dWAR, and credit the player with gradually increasing value in each as he separates himself from the average

The core element to all of these is that we’re looking only at performance that’s above the league average, and we’re favoring big seasons disproportionately.  With this method, there’s no difference between a player staying on as a fringe starter, hanging around as the worst player in the league, or choosing to retire.  The goal here is to evaluate the years in which a player made an impact, not to color his totals with years that are more representative of a choice in when to retire.