WAR and JAWS – Exaggerating Fielding Value
February 25, 2019 by Michael Hoban · 1 Comment
In an earlier post, I pointed out that Matt Chapman’s WAR value for 2018 (8.2) seemed to indicate that he had the third best season of any major leaguer – behind only Mookie Betts and Mike Trout. Of course, any serious fan looking at the numbers would know that this assertion is not believable. There were at least five or six (or more) players who had better seasons than Carpenter.
Then how did it happen? Unfortunately, WAR has a flaw – which is a tendency to exaggerate the value of a player’s fielding if he is perceived to be a very good fielder or a very poor one. That is, if (as with Matt Chapman in 2018) a player is found to be a very good fielder, then his defensive WAR is illogically increased out of proportion to his total accomplishments. And if a player is found to be a poor fielder, the opposite happens.
This “intrinsic fielding exaggeration” of WAR becomes most apparent when JAWS calculates the “career value” of a player using WAR.
JAWS calculates the career value of a player in a manner somewhat similar to the CAWS Career Gauge – but with two big differences. Here are the formulas:
CAWS = CV + .25(CWS – CV) where
CWS = career win shares and CV = core value (10 best WS seasons)
JAWS = 7WAR + .5(WAR – 7WAR) where
WAR = career WAR and 7 WAR = peak value (7 best WAR seasons)
You can see that by taking the 10 best seasons CAWS is looking for a longer “quality period” in a career – indicative of a “Hall of Famer.” By taking only 7 seasons, JAWS is far less demanding of a player’s career.
CAWS then adds only 25% of the remaining win shares to give less importance to “longevity” – whereas JAWS adds 50% of the remaining WAR.
The principal problem remains, however, that JAWS relies on WAR – which is a somewhat flawed metric if the player is perceived to be on either end of the fielding spectrum – that is, either a very good fielder or a poor fielder.
This is best illustrated by examining some JAWS results. Here are the “best shortstops” (since 1920) according to JAWS.
Alex Rodriquez
Cal Ripken Jr.
Robin Yount
Arky Vaughan
Ernie Banks
Ozzie Smith
Luke Appling
Alan Trammell
Derek Jeter
Barry Larkin
Lou Boudreau
Joe Cronin
Pee Wee Reese
And here are the best shortstops according to CAWS.
Alex Rodriquez
Arky Vaughan
Cal Ripken Jr.
Robin Yount
Derek Jeter
Luke Appling
Joe Cronin
Barry Larkin
Ernie Banks
Pee Wee Reese
Lou Boudreau
Alan Trammell
Ozzie Smith
Note a couple of interesting points:
1. The same 13 players are on both lists of best shortstops.
2. The same 4 players are the top 4.
3. CAWS suggests all 13 have HOF numbers.
4. JAWS suggests all but Reese have HOF numbers.
First of all, does anyone really believe that Alan Trammell had better career numbers than Derek Jeter – as JAWS suggests? Of course not! This is the first clue that something is wrong here – and it is “credit for fielding.”
It is the placement of Derek Jeter and Ozzie Smith on the JAWS list that best illustrates the “fielding exaggeration” of WAR. Derek Jeter clearly had better career numbers than Ozzie Smith but is ranked below him.
1. Ozzie is arguably the best defensive shortstop of all time – and, therefore, is given an exaggerated credit for fielding by WAR and winds up as #6 on the JAWS list. His appropriate ranking is #13 as on the CAWS list.
2. Jeter is considered by WAR to have been a poor fielder (which is accurate) but is then penalized in an inappropriately harsh manner by WAR and winds up ranked #9 by JAWS—when he is more accurately #5 as on the CAWS list.
What to conclude from all this?
1. Win shares is a more mathematically sound metric than WAR – which leads to CAWS being somewhat more accurate than JAWS at assessing a player’s career.
2. JAWS is fairly accurate except when dealing with a player who was a very good fielder or a poor fielder (because of the flaw in WAR).
3. When assessing “HOF credentials,” consult both systems!
Thank you for your time.
Mike Hoban, Ph.D.
Professor Emeritus (mathematics) – City U of NY
Author of
DEFINING GREATNESS: A Hall of Fame Handbook (2012)
profhoban@gmail.com
It is unclear what you are arguing, other than the known differences between what JAWS and CAWS are designed to measure.
You claim that Jeter is superior to both Trammell and Smith, which is a reasonable claim, but your support for that claim is essentially that everyone knows this to be the case. Your conclusion is implicit in your premise. Trammell is offensively similar to Jeter during his prime. Over a 10 year stretch of their primes for each, their respective offensive WAR (which does not include fielding ability) is within half a WAR per 650 PAs.
If JAWS only takes the best 7 of those seasons and CAWS takes 10, then that will benefit Trammell in this case. But the fact remains that they are relatively offensively similar. Jeter played longer and was generally more durable, but he was not the defensive SS that Trammell was. If they are offensively similar and Trammell is defensively superior, why is it such a stretch to believe that Trammell is a more valuable SS than Jeter?
The sleight-of-hand in your argument is that you ask the rhetorical question “Does anybody really believe that Trammell had better career numbers than Jeter?” But you do not define “career numbers.” You then use that lack of definition to support the claim that WAR is flawed. Neither JAWS nor CAWS are measuring “career numbers” as commonly defined- BA, OBP, SLG, R, RBI, so forth. Obviously, no one would claim that Trammell has better “stats” than Jeter. But that was never the topic.