Just when you thought you had enough stats to last a lifetime someone had to come along and muck up the works with 10-year peak WAR.
As the most rabid of baseball stat junkies will tell you, wins above replacement (WAR) is a measure of performance that sets a player against the cumulative league averages to determine how much better or worse that player is compared to the “next best” option. Examining peak WAR as it is used currently, raises questions with about the validity of a player’s “peak” seasons as expressed through the 7-year peak WAR statistic.
My problem with 7-year peak WAR is that it does not give you a player’s peak production. It only tells you what his seven best statistical seasons were regarding wins above replacement. This is wrong for a couple of reasons. Allow me to explain my reasoning.
Peak is Prime
To me, peak is synonymous with a player’s physical prime. I would like to find common ground here because I hate to break it to the hardcore stat guys, peak does not mean seven best seasons. The definition of peak should be the same as talking about a player’s prime years, or when he is at his physical apex.
Settle down and let me finish before you go dusting off those torches. Don’t go thinking problems with peak WAR as it is currently considered is a challenge to WAR itself. Wins above replacement is very useful, especially when gauging a player’s Cooperstown credentials. My problem is with the way it is calculated with respects to a player’s peak.
I have spent countless hours poring over player data and calculating my own version of “peak WAR” and my application isn’t what might be usually expected. It’s hardly an attempt at reinventing the wheel though. Think of it as a minor tweak in how we view a player’s peak production. I must also add; the Cooperstown inductees have nothing to fear.
When looking at the peak of a pro-ballplayer, I don’t need to know what his seven best WAR seasons are, nor do I care. No, what I need to know is how well he performed through his physical peak. Here’s an example showing exactly what’s trying to be conveyed. Brett Favre in 2009 put up the greatest season of his entire career at 40-years old. Now tell me this, is this a guy in his peak? Or, is this an outlier of a season that happened outside of his physical peak? I’m going with the latter folks.
Let me get to the nuts and bolts. What I mean by physical peak is this: what is the player(s) production over his age 23-33 seasons when he is the strongest, fastest and fittest that he will ever be?
10-Year Peak WAR
Why pick 10 years as a sample? Firstly, this examination of peak WAR should only be used as a measure for Hall of Fame standards. The way I apply WAR should never be used on active players, unless you are comparing them with the career trajectory of a legend.
As I look at more and more data, those 10 years (23-33) look to be the general peak ages a player does his most damage. Granted there are players that don’t fit that criteria exactly, but these standards of peak envisioned here don’t care about that. If you enter the game at 24 years of age, like Kirby Puckett did for example, I take that as being a peak season. The reasoning is this, Hall of Fame players generally get to the bigs earlier and they stay longer.
Players should be rewarded for their production in their “non-peak” years as well. In my application of WAR, I generate two classes: 10-year peak WAR and Non-peak WAR. All 11 seasons that fall between a player’s age 23-33 seasons are his 10-year peak, and all other seasons up to age 22, and all seasons post-age 33 are calculated to be his non-peak WAR.
These calculations of 10-year peak WAR vs. Non-peak WAR speaks to one thing. Career Longevity. This is not to say that a player cannot be Hall of Fame worthy after playing a limited number of years, but generally, we all know that you need at least a decade of dominant play on your resume to get in to Cooperstown.
There are exceptions to every rule of course, but how many Sandy Koufax’s are there exactly? Koufax, by my system, had eight seasons of his 10-year prime only, and yet still managed a (50.2) WAR over that stretch.
It only becomes more impressive when you realize that in eight seasons from age 23-30, Koufax still put up better 10-year peak WAR than did Molitor, Stargell, Winfield and Puckett along with many more.
This is where examining peak WAR takes a twist. A player should be rewarded for his length of career. If a player makes it to the bigs at 21 for instance, those first two seasons while he’s developing are tacked on to whatever production he shows from age 34 until retirement. This is what I call Non-peak WAR.
Consider my application of WAR as I have outlined it so far. What I am essentially doing, is saying how good were these guys, and for how long? I am favoring career length as much as I am favoring the player’s overall production and worth to his team. Trust me, the Hall of Famers still stand out. Start doing some calculations if you don’t believe me.
If you are a purest like me, Cooperstown isn’t for those that burn out after five seasons (unless you’re ridiculous like Koufax), Cooperstown is for those that do it better and do it longer. In case you are wondering what Sandy’s Non-peak WAR was, it was (3) and that’s not a typo either. The fact that Koufax made the Hall is a testament to how great he actually was.
Consider Paul Molitor. From 1980 through 1990, Molitor posted a (41.3) WAR. That’s damn good. But it’s also off the pace of Hall of Fame standards for third basemen using this version of 10-year peak WAR by nearly 10-points. It’s what Molitor did in those other 10 of his 21 big league seasons that truly sets him apart. His Non-peak WAR (34.2) is over two-times higher than Hall standard at his position (15.9). Molitor’s Non-peak WAR is so good, it puts him as the best of all time at third base in Non-peak WAR by nearly 9-points over Mike Schmidt’s (25.6) Non-peak WAR.
What it Means
There really is no solid indicator for career longevity. Especially when you isolate a player’s seven best seasons irrespective of when they occurred in a player’s career chronologically. Those who play a shorter amount of time are going to have to be so good they won’t be denied. Like Koufax.
Falling short on one end of these WAR calculations isn’t scuttling a player’s shot at the Hall. But it is putting them to a higher standard to truly dominate for the brief moments they are playing.
What is harsh though, is Larry Walker only getting 21.9 percent of the vote in the most recent Hall of Fame voting. On his seventh ballot, mind you. Here’s a guy that finished with a 10-year peak WAR of (49.4) and a Non-peak WAR of (23.3). Not bad considering Hall average for RF is (52.6/20.6) by my system.
Walker is off the 10-year peak WAR of right fielders by 3-points, but he’s above Non-peak production by nearly 3-points. How is Walker not getting more than 1 in 5 Hall votes? And please, do not give me that, “He played in Colorado!” crap either. I’m not having it, where a player takes the field for their home games should not be looked upon as a sin. Furthermore, if that’s the standard we’re going by I feel bad for any great player that calls Coors Field home. Let’s not make Larry Walker another snub job that the Veterans Committee is going to have to fix.
Like the Alan Trammell debacle.
(feature photo courtesy of: Sports Illustrated)
“From our Haus to yours