My beef with WAR is how it's become the only stat anyone seems to use to evaluate current players. We'll probably never see another MVP or Cy Young winner with less than 6.5 WAR no matter their other stats or perceived impact because we only trust eye test that have WAR to back it up. If a player had a bunch of clutch hits in a successful pennant race with a .300 BA 100 RBI and 4 total WAR - a performance that would have made him an MVP most of baseball history - he might not even get any first place votes today.
We still have enough pre-WAR HOF candidates that it's helpful for finding underrated candidates. It also hasn't been too much of a demerit in lower cases. That is, a player that passes by the old stat / eye test standards is getting in even with lower relative WAR (Vladdy, Trevor Hoffman, Molina eventually, etc.). And as for those players with lower WAR than their reputation suggested, the only one I can think of that's been iced out is Jeff Kent (even though his WAR is pretty high - and there might be other factors afoot with his exclusion).
So WARs mostly been a positive for the HOF so far. It helps extra people get in without being the main reason others are left out. But soon enough it's going to be the only thing talked about with the HOF because it's the only thing anyone uses to evaluate anything anymore, and that's gonna be boring.
But if a player hits .300 with 100 RBI and a bunch of clutch hits but only 4 total WAR, that means there's something beneath the surface that is missing. Maybe it's a guy who just never walks, and/or doesn't have much power. Maybe he's a particularly bad runner (e.g. consistently costing his team bases). Maybe he's a defensive butcher. The eye test is great, but if a player has a seemingly low WAR, at the very least it's worth asking why. It very well may be something that really does hurt the team but doesn't show up on the back of the baseball card.
Since Joe says "it’s pretty clear there’s little-to-no correlation at all", I thought I would calculate the correlation coefficient between WAR and vote percentage for this little group. And yes, exactly, it's only +0.22. This is not "statistically significant", and it really could just be luck. Seriously, you might well get a bigger correlation if you shuffled all the WAR numbers and dealt them out at random.
One other thing. I did the correlation for Win Shares to vote percentage, because I think that Win Shares are much better than WAR and it saddens me that they are fast fading into obscurity. Anyway, this time there's a correlation of +0.46 which is much bigger and almost certainly not luck.
So for this little group, Win Shares do a much better job of predicting voting than WAR did. Even though neither measure was known to the voters in 2001.
But in the future, it will be the other way round because voters will be using WAR and not Win Shares.
The good news is that yes, there is, you can find Win Shares by subscribing to Bill James' website. The bad news is that he is closing it down in September. But the other good news is that he is considering using his new free time to write the NNHBA, which many of us are eagerly anticipating.
This is beside the point, but there was one very unusual feature about that 2001 ballot. One of the players, Jose Rijo, returned to play in the majors in 2001 and 2002. He got just one vote in 2001, but came back on the ballot in 2008 (getting no votes second time, even with five more wins).
That WAR figure of 36.5 for him should perhaps be 36.4 to exclude those last two years, but I'm not sure and, as I say, it's beside the point...
By Fangraphs's version of WAR, Yadier Molina's 55.7 and Buster Posey's 57.5 are quite a bit higher, and more Hall-like, than the B-Ref version's. The Baseball Prospectus version has Molina 58.5, even higher, and Posey at 53.6 with four MVP-type seasons in the 2012-16 years.
Given that catchers play fewer games than regulars at other positions, these are WAR-based arguments *for* both men, supporting the popular perception of their value.If they do go in early, I think that'll be part of why: even for voters who want to choose them, it's nice to have that easily-recognizable backing, and I bet it makes a difference.
Fangraph's version of WAR gives Molina and Posey an enormous boost from framing, as does BP's, but because we don't have historic framing data, we don't actually know if that makes them all-time greats or if every generation has their version of Molina and Posey, only in the '70s it was Bob Boone and in the '80s it was Lance Parrish
Oh, I know. Fangraphs *does* in fact attempt to estimate historic framing numbers -- but since it has do so without direct observation, its estimates are of course conservative and heavily regressed.
Look at the MVP voting for 2001 and the question is: Which of these players wasn't juiced? Good thing there are players with plausible deniability or else we would have no heros from that era.
I think they are probably differing correlations for different aspects of war. I bet that offinsive war is more highly correlated to Hall of Fame voting percentage than defensive war. And for defensive war, I bet outfield assists is one of the factors which correlates more with Hall of Fame voting. I wonder if there’s a way to compare the different aspects of war. I bet there’s virtually no correlation for things like baserunning, range, factor, etc.
As always, the Hall of Fame consideration needs to be holistic. What were the counting numbers? How did the Player compare to other players in his era, especially those who played the same position? How great was the player at his peak? How long was that peak? Did the Player excel in the postseason? Did he seem to help his team win A lot? I bet we could compare some of these with certain aspects of war and get some correlations.
Recently I did a fun little project based on HOF players who wore uniform number 11. Too long and amateurish to post here, but it fits into todays main topic. The pitchers were Carl Hubbell 68.5 WAR 253-154, 2.98 ERA, Early Wynn 61.0 WAR 300-244 3.54 ERA, Waite Hoyte 52.5 WAR 237-182 3.59 ERA, Herb Pennock 45.5 WAR 241-162 3.60 ERA and Lefty Gomez 37. WAR 189-102 3.34 ERA. How does this fit it? One of my all time favorite Mariners was Jamie Moyer 49.8 WAR 269-209 4.25 ERA, one time All-Star, finished 4th, 5th and 6th in the three seasons he received Cy Young votes. Jamie was truly a "crafty lefty." He might occasionally have hit 80 MPH with his fastball, but not in most games. Compared to the 5 HOFs he would have been 4th of 6 in WAR and 2nd of 6 in wins.
Was he Hall of Fame worthy? I would have to say no. He was never truly dominant at any time but had incredible longevity. His closest connection to the Hall of Fame comes from the fact that he is the son-in-law of Digger Phelps, nominated but not yet selected for the Naismith H0 all of Fame. Digger's Notre Dame teams upset a record 7 #1 teams during his tenure. The most notable was 1974 71-70 win over UCLA that snapped the Bruins 88 game winning streak. The other 6 came against 6 other teams, 1977 San Francisco, 1978 Marquette, 1980 DePaul, 1980 Kentucky, 1981 Virginia and 1987 North Carolina.
Joe, I think I've got a question that will interest you.
It's easy to track down career WAR. Easy to track down single season WAR.
What's not easy to track down is WAR/season as a comparitive.
Like, where's the career leader list for WAR per season? Or WAR per season by position?
Not WAR per 162 games … but WAR/season.
I feel like that could shed much light on a player's overall worth over his career AND even more over his prime.
I came to this question considering Will Clark's Hall of Fame case:
Career .303/.384/.497 … 56.5 WAR over 15 seasons (4.6 average).
Isn't that a fairly spectacular WAR average? If so, is it valued spectacularly? Or not because he "only" played 15 seasons and only hit 284 HRs and drove in 1205?
I'm not saying we should devalue career numbers, many achieved beyond a player's prime.
But I am suggesting a player's effectiveness per season should be elevated.
This doesn't solve the career leader list problem, which I agree would be great, but it might be helpful.
It's only slightly easier than doing the math, but if you click on one row on BRef, then click on another, a window will pop up with averages per season and per 650 PA. So for Clark's Player Value - Batting section, click any field in the row for 1986, then do the same for 2000. The result will show you an average of 3.8 WAR per season and 4.4 per 650 PA.
This can be useful if a particular stretch jumps out at you, like the five best years of a player's prime, stats before and after a major injury, or something other than the seven-year peak listed toward the bottom of a page.
WAR is awful at measuring catchers & a bit subpar when measuring pitchers. I say this as someone who absolutely feel it's maybe the best metric to give you an idea on how good a baseball player is overall. It's more telling than OPS+ & runs saved on defense combined. But it just does an awful job measuring the impact of catchers on the game, by far the hardest position on the diamond to fill with the shortest peak & is off the mark measuring pitching performance. I mean Dave Winfield & Kershaw have similar WAR. But if you could pick a side in a trade Kershaw or Dave Winfield+ Frank White, what would you choose? Every GM & saber metrically minded fan & MLB exec would take Kershaw.
Agreed on catchers, although Fangraphs WAR values them a bit more. I think, as better and better metrics come out in regards to pitch framing, catchers will get more respect (although one of the most crucial elements, game calling, still goes neglected. Maybe we need like +/- data showing how much better or worse pitchers are based on their battery mate?)
Pitchers, I'm not sure if I agree. I mean, you say Winfield and Kershaw have 'similar' WAR, but 79 vs. 64 bWAR is a pretty big gap, and Kershaw is still an active player who, I fully suspect, will be at 90+ WAR when he hangs it up.
"WAR, huh, yeah; What is it good for; Absolutely nothing"
I was, and remain, an eyeball test guy.
I didn't need bWAR to tell me that Roberto Clemente was a great rightfielder. He unleashed missiles with his arm. (Forget cannon....he was throwing missiles!) The book on Clemente: Don't run on him. (Willie probably did. Sometimes.)
More random eyeball thoughts:
Sorry for Yankee haters: Everyone knew that Mariano Rivera was a HoFer. The reaction after Luis Gonzalez's Game 7 hit told the world. Mo simply didn't fail in that spot. The shock was that he did. Much more than a "one inning guy".
Lou Whitaker, Luis Tiant and Tommy John all are HoFers. A shame they may never get in.
Posey, Molina and Mauer all are HoFers. Maybe not first ballot. (An honor that I romantically think matters given HoF history. Although, I can't fix why Cy Young was not in the first ever class.) They all made their teams better...and were players that could have played at any time in history.
Munson and Posey had remarkable similar careers. Both were ROY and MVP winners. They both won multiple World Series. Posey winning 3, Munson winning 2. Both played in 3 WS with Munson having the more impressive postseason batting statistics. They had similar career WAR and similar games played. Both had their careers shortened by unfortunate circumstances.
If Posey goes into the HOF, what is the argument for keeping Munson out?
As much I didn't like Munson, (Royals fan. It was a real rivalry then. When we finally broke through, it was after Munson was gone) he was a great player. You would think he might get some sympathy for having died while he had some years left. I think Puckett, mentioned above, got a lot of sympathy vote when he had to retire early because of health issues - three years later in age than Munson lived.
While I don't like comparisons based on "If this guy, why not that guy?' because it would lead to a gigantic Hall (I am mildly a big hall guy, but not that big) it is still a legitimate question.
I am not sure if Posey makes the Hall, or Mauer. I would bet my house that Yadier Molina (perhaps the most overrated modern player) gets in on the first ballot, and all three guys are better than him.
Besides some antipathy for Munson when he played, if he came on one of the committees and I had a vote, he would get my vote.
I too used to despise Munson (Royal fan as well). But I read something about the 1977 playoff brawl between George Brett and Graig Nettles that forever changed my mind. Watch the video of the fight. Munson charges into the pile then climbs on top of Brett to protect him from the horde. That just blew my mind knowing what heated rivals the two teams were. Munson may not ever make the HOF but he’s got my respect.
Sure, let's give you the benefit of the doubt that your eyeball test is accurate.
It clearly ain't for a lot of voters who don't use WAR, because they consistently underrate defense, baserunning, plate discipline, eating lots of innings, and many other things that contribute to winning.
You can quibble with elements of the formula, but at least WAR accounts for all that stuff. It's pretty clear to look at what kind of players get historically underrated by the HOF and see where people's eyeballs lie, and it mostly comes down to getting lots of hits and home runs.
I was lucky enough to be in Boston on May 5, 1904. Was looking forward to a pitchers duel against Waddell. Only perfect game I've ever seen. Waddell had an uncharacteristic off day, surrendering 10 hits and somehow contained that to three runs.
My beef with WAR is how it's become the only stat anyone seems to use to evaluate current players. We'll probably never see another MVP or Cy Young winner with less than 6.5 WAR no matter their other stats or perceived impact because we only trust eye test that have WAR to back it up. If a player had a bunch of clutch hits in a successful pennant race with a .300 BA 100 RBI and 4 total WAR - a performance that would have made him an MVP most of baseball history - he might not even get any first place votes today.
We still have enough pre-WAR HOF candidates that it's helpful for finding underrated candidates. It also hasn't been too much of a demerit in lower cases. That is, a player that passes by the old stat / eye test standards is getting in even with lower relative WAR (Vladdy, Trevor Hoffman, Molina eventually, etc.). And as for those players with lower WAR than their reputation suggested, the only one I can think of that's been iced out is Jeff Kent (even though his WAR is pretty high - and there might be other factors afoot with his exclusion).
So WARs mostly been a positive for the HOF so far. It helps extra people get in without being the main reason others are left out. But soon enough it's going to be the only thing talked about with the HOF because it's the only thing anyone uses to evaluate anything anymore, and that's gonna be boring.
But if a player hits .300 with 100 RBI and a bunch of clutch hits but only 4 total WAR, that means there's something beneath the surface that is missing. Maybe it's a guy who just never walks, and/or doesn't have much power. Maybe he's a particularly bad runner (e.g. consistently costing his team bases). Maybe he's a defensive butcher. The eye test is great, but if a player has a seemingly low WAR, at the very least it's worth asking why. It very well may be something that really does hurt the team but doesn't show up on the back of the baseball card.
It’s funny, Mauer has a higher WAR than Molina or posey and yet I would rate him a distant third when it comes to the HOF qualifications of those 3
This was a great article. Proper research.
Since Joe says "it’s pretty clear there’s little-to-no correlation at all", I thought I would calculate the correlation coefficient between WAR and vote percentage for this little group. And yes, exactly, it's only +0.22. This is not "statistically significant", and it really could just be luck. Seriously, you might well get a bigger correlation if you shuffled all the WAR numbers and dealt them out at random.
One other thing. I did the correlation for Win Shares to vote percentage, because I think that Win Shares are much better than WAR and it saddens me that they are fast fading into obscurity. Anyway, this time there's a correlation of +0.46 which is much bigger and almost certainly not luck.
So for this little group, Win Shares do a much better job of predicting voting than WAR did. Even though neither measure was known to the voters in 2001.
But in the future, it will be the other way round because voters will be using WAR and not Win Shares.
Is there a place to look up Win Shares the same way you can look up anything on Baseball Reference?
Waiting for Bill James to write the New New Baseball Historical Abstract.
The good news is that yes, there is, you can find Win Shares by subscribing to Bill James' website. The bad news is that he is closing it down in September. But the other good news is that he is considering using his new free time to write the NNHBA, which many of us are eagerly anticipating.
This is beside the point, but there was one very unusual feature about that 2001 ballot. One of the players, Jose Rijo, returned to play in the majors in 2001 and 2002. He got just one vote in 2001, but came back on the ballot in 2008 (getting no votes second time, even with five more wins).
That WAR figure of 36.5 for him should perhaps be 36.4 to exclude those last two years, but I'm not sure and, as I say, it's beside the point...
By Fangraphs's version of WAR, Yadier Molina's 55.7 and Buster Posey's 57.5 are quite a bit higher, and more Hall-like, than the B-Ref version's. The Baseball Prospectus version has Molina 58.5, even higher, and Posey at 53.6 with four MVP-type seasons in the 2012-16 years.
Given that catchers play fewer games than regulars at other positions, these are WAR-based arguments *for* both men, supporting the popular perception of their value.If they do go in early, I think that'll be part of why: even for voters who want to choose them, it's nice to have that easily-recognizable backing, and I bet it makes a difference.
Fangraph's version of WAR gives Molina and Posey an enormous boost from framing, as does BP's, but because we don't have historic framing data, we don't actually know if that makes them all-time greats or if every generation has their version of Molina and Posey, only in the '70s it was Bob Boone and in the '80s it was Lance Parrish
Oh, I know. Fangraphs *does* in fact attempt to estimate historic framing numbers -- but since it has do so without direct observation, its estimates are of course conservative and heavily regressed.
Look at the MVP voting for 2001 and the question is: Which of these players wasn't juiced? Good thing there are players with plausible deniability or else we would have no heros from that era.
A San Diego book tour stop would be greatly appreciated!
Spring Lake and Cherry Hill in New Jersey - what about us northern NJ fans? Is it too close to Yankee Stadium to merit a visit?
I think they are probably differing correlations for different aspects of war. I bet that offinsive war is more highly correlated to Hall of Fame voting percentage than defensive war. And for defensive war, I bet outfield assists is one of the factors which correlates more with Hall of Fame voting. I wonder if there’s a way to compare the different aspects of war. I bet there’s virtually no correlation for things like baserunning, range, factor, etc.
As always, the Hall of Fame consideration needs to be holistic. What were the counting numbers? How did the Player compare to other players in his era, especially those who played the same position? How great was the player at his peak? How long was that peak? Did the Player excel in the postseason? Did he seem to help his team win A lot? I bet we could compare some of these with certain aspects of war and get some correlations.
Fun piece. The fact that it took Gary Carter SIX elections to be voted in is and was completely ridiculous.
Yes, but Whitaker not being elected at all is even more ridiculous.
To riff off Joe's clever headline and for those who are familiar with Edwin Starr's song from over 50 years ago:
WAR, what is it good for?
Absolutely some things, but not everything.
(Say it again, y'all!)
Recently I did a fun little project based on HOF players who wore uniform number 11. Too long and amateurish to post here, but it fits into todays main topic. The pitchers were Carl Hubbell 68.5 WAR 253-154, 2.98 ERA, Early Wynn 61.0 WAR 300-244 3.54 ERA, Waite Hoyte 52.5 WAR 237-182 3.59 ERA, Herb Pennock 45.5 WAR 241-162 3.60 ERA and Lefty Gomez 37. WAR 189-102 3.34 ERA. How does this fit it? One of my all time favorite Mariners was Jamie Moyer 49.8 WAR 269-209 4.25 ERA, one time All-Star, finished 4th, 5th and 6th in the three seasons he received Cy Young votes. Jamie was truly a "crafty lefty." He might occasionally have hit 80 MPH with his fastball, but not in most games. Compared to the 5 HOFs he would have been 4th of 6 in WAR and 2nd of 6 in wins.
Was he Hall of Fame worthy? I would have to say no. He was never truly dominant at any time but had incredible longevity. His closest connection to the Hall of Fame comes from the fact that he is the son-in-law of Digger Phelps, nominated but not yet selected for the Naismith H0 all of Fame. Digger's Notre Dame teams upset a record 7 #1 teams during his tenure. The most notable was 1974 71-70 win over UCLA that snapped the Bruins 88 game winning streak. The other 6 came against 6 other teams, 1977 San Francisco, 1978 Marquette, 1980 DePaul, 1980 Kentucky, 1981 Virginia and 1987 North Carolina.
Joe, I think I've got a question that will interest you.
It's easy to track down career WAR. Easy to track down single season WAR.
What's not easy to track down is WAR/season as a comparitive.
Like, where's the career leader list for WAR per season? Or WAR per season by position?
Not WAR per 162 games … but WAR/season.
I feel like that could shed much light on a player's overall worth over his career AND even more over his prime.
I came to this question considering Will Clark's Hall of Fame case:
Career .303/.384/.497 … 56.5 WAR over 15 seasons (4.6 average).
Isn't that a fairly spectacular WAR average? If so, is it valued spectacularly? Or not because he "only" played 15 seasons and only hit 284 HRs and drove in 1205?
I'm not saying we should devalue career numbers, many achieved beyond a player's prime.
But I am suggesting a player's effectiveness per season should be elevated.
Your thoughts would be fabulous.
Thanks.
This doesn't solve the career leader list problem, which I agree would be great, but it might be helpful.
It's only slightly easier than doing the math, but if you click on one row on BRef, then click on another, a window will pop up with averages per season and per 650 PA. So for Clark's Player Value - Batting section, click any field in the row for 1986, then do the same for 2000. The result will show you an average of 3.8 WAR per season and 4.4 per 650 PA.
This can be useful if a particular stretch jumps out at you, like the five best years of a player's prime, stats before and after a major injury, or something other than the seven-year peak listed toward the bottom of a page.
JAWS sort of addresses this issue. https://www.baseball-reference.com/about/jaws.shtml
WAR is awful at measuring catchers & a bit subpar when measuring pitchers. I say this as someone who absolutely feel it's maybe the best metric to give you an idea on how good a baseball player is overall. It's more telling than OPS+ & runs saved on defense combined. But it just does an awful job measuring the impact of catchers on the game, by far the hardest position on the diamond to fill with the shortest peak & is off the mark measuring pitching performance. I mean Dave Winfield & Kershaw have similar WAR. But if you could pick a side in a trade Kershaw or Dave Winfield+ Frank White, what would you choose? Every GM & saber metrically minded fan & MLB exec would take Kershaw.
Agreed on catchers, although Fangraphs WAR values them a bit more. I think, as better and better metrics come out in regards to pitch framing, catchers will get more respect (although one of the most crucial elements, game calling, still goes neglected. Maybe we need like +/- data showing how much better or worse pitchers are based on their battery mate?)
Pitchers, I'm not sure if I agree. I mean, you say Winfield and Kershaw have 'similar' WAR, but 79 vs. 64 bWAR is a pretty big gap, and Kershaw is still an active player who, I fully suspect, will be at 90+ WAR when he hangs it up.
"WAR, huh, yeah; What is it good for; Absolutely nothing"
I was, and remain, an eyeball test guy.
I didn't need bWAR to tell me that Roberto Clemente was a great rightfielder. He unleashed missiles with his arm. (Forget cannon....he was throwing missiles!) The book on Clemente: Don't run on him. (Willie probably did. Sometimes.)
More random eyeball thoughts:
Sorry for Yankee haters: Everyone knew that Mariano Rivera was a HoFer. The reaction after Luis Gonzalez's Game 7 hit told the world. Mo simply didn't fail in that spot. The shock was that he did. Much more than a "one inning guy".
Lou Whitaker, Luis Tiant and Tommy John all are HoFers. A shame they may never get in.
Posey, Molina and Mauer all are HoFers. Maybe not first ballot. (An honor that I romantically think matters given HoF history. Although, I can't fix why Cy Young was not in the first ever class.) They all made their teams better...and were players that could have played at any time in history.
Munson and Posey had remarkable similar careers. Both were ROY and MVP winners. They both won multiple World Series. Posey winning 3, Munson winning 2. Both played in 3 WS with Munson having the more impressive postseason batting statistics. They had similar career WAR and similar games played. Both had their careers shortened by unfortunate circumstances.
If Posey goes into the HOF, what is the argument for keeping Munson out?
As much I didn't like Munson, (Royals fan. It was a real rivalry then. When we finally broke through, it was after Munson was gone) he was a great player. You would think he might get some sympathy for having died while he had some years left. I think Puckett, mentioned above, got a lot of sympathy vote when he had to retire early because of health issues - three years later in age than Munson lived.
While I don't like comparisons based on "If this guy, why not that guy?' because it would lead to a gigantic Hall (I am mildly a big hall guy, but not that big) it is still a legitimate question.
I am not sure if Posey makes the Hall, or Mauer. I would bet my house that Yadier Molina (perhaps the most overrated modern player) gets in on the first ballot, and all three guys are better than him.
Besides some antipathy for Munson when he played, if he came on one of the committees and I had a vote, he would get my vote.
I too used to despise Munson (Royal fan as well). But I read something about the 1977 playoff brawl between George Brett and Graig Nettles that forever changed my mind. Watch the video of the fight. Munson charges into the pile then climbs on top of Brett to protect him from the horde. That just blew my mind knowing what heated rivals the two teams were. Munson may not ever make the HOF but he’s got my respect.
Sure, let's give you the benefit of the doubt that your eyeball test is accurate.
It clearly ain't for a lot of voters who don't use WAR, because they consistently underrate defense, baserunning, plate discipline, eating lots of innings, and many other things that contribute to winning.
You can quibble with elements of the formula, but at least WAR accounts for all that stuff. It's pretty clear to look at what kind of players get historically underrated by the HOF and see where people's eyeballs lie, and it mostly comes down to getting lots of hits and home runs.
In other words, there are voters that either don't watch enough, or truly understand, baseball. Agreed.
How did Cy Young do on your eyeball test?
I was lucky enough to be in Boston on May 5, 1904. Was looking forward to a pitchers duel against Waddell. Only perfect game I've ever seen. Waddell had an uncharacteristic off day, surrendering 10 hits and somehow contained that to three runs.
All the appearances should be recorded and posted as poscasts (suitably edited). And why do you hate Boston?