Matt Kemp’s huge team margin in WAR
Posted by Andy on October 12, 2011
Matt Kemp registered a WAR of 10.0 this season. Among his teammates, the next highest WAR for non-pitchers was 1.8 from Jamey Carroll. In fact, the sum of all other Dodger WAR (ignoring pitching) was 6.4, meaning Kemp outscored all the rest of his teammates combined.
His margin of 8.2 above the next highest player is massive. Here are the 200 most recent times that a single player (non-pitcher) achieved a seasonal WAR of 8.2.
Working chronologically backwards, here's what I found:
- In 2007, Albert Pujols had a WAR of 8.3 and his next-closest teammate was Scott Rolen at 2.3, a difference of 6 wins. All Cardinals besides Pujols totaled 9.1, so Pujols nearly outscored them.
- The last time before Kemp that a player had double-digit WAR (non-pitcher) was Barry Bonds in 2004. Indeed, he outscored runner-up Ray Durham by 9.1, besting Kemp's 2011 mark. But the Giants other than Bonds did manage 16.5 WAR, so Bonds didn't outscore them single-handedly.
- Adrian Beltre also had some big margins in 2004. He finished at 10.1 over teammate Milton Bradley at 3.6, a difference of 6.5 wins. The team minus Beltre totaled 14.2.
- In 2003 and 2001, Bonds came somewhat close to outscoring his teammates in WAR.
- In 2001, Sammy Sosa registered an 11.4 WAR, a whopping 9.2 wins above runner-up Ricky Gutierrez at 2.2. Sosa's teammates totaled 11.5, just a pinch more than Sosa himself.
- In 1997, Larry Walker had 9.0 WAR and was trailed by Andres Galarraga at 2.9 WAR, for a margin of 6.1. The Rockies minus Walker had 9.6.
- Got it finally: in 1997, Mike Piazza racked up 9.3 WAR. Next closest was Raul Mondesi at 5.8, but all the Dodgers except for Piazza had just 7.6.
This means that Kemp is the first player since Piazza in 1997 to outscore all of his teammates (non-pitchers) in WAR.
I wonder when the last time a non-Dodgers player did it? Feel free to keep looking at the list.
Thanks to reader Bip for emailing in about this.
October 12th, 2011 at 2:26 pm
Yeah, I'm that good.
October 12th, 2011 at 2:29 pm
WAR is an approximate stat and nobody cares about statistical anomalies and thresholds in a stat that's very definition will likely change in a week...
...and if it doesn't change in a week, it should. It's garbage.
October 12th, 2011 at 2:33 pm
I'm sure everyone else appreciates you speaking for them.
October 12th, 2011 at 3:19 pm
Going through the rest of the Dodgers lineup, the funny thing is that almost all of them (maybe not Ethier) did as well or slightly better than could be expected. How could a big money team trot out a lineup that weak?
Maybe they thought they had the original Tony Gwynn & Ivan DeJesus.
October 12th, 2011 at 3:28 pm
Doh! McCourt.
http://www.bizjournals.com/boston/real_estate/2011/04/mlb-takes-over-los-angeles-dodgers.html
October 12th, 2011 at 3:30 pm
That report has not been picked up by ESPN or MLB.com so I doubt it is true (or official yet.)
October 12th, 2011 at 3:36 pm
Andy, did you just go through the list of WAR > 8.2? It seems likely someone on a crappy team could have outscored the rest of his teammates despite a WAR less than 8.2.
October 12th, 2011 at 3:41 pm
JT, yeah, I was looking for the last person to outscore all of his teammates by at least 8.2 each, and all in total. As you can see above, I found a few examples of guys who outscored the second-place guy by at least 8.2 but not the whole team. I'm sure there are other guys to outscore their entire team but not the second-place guy by at least 8.2.
October 12th, 2011 at 3:44 pm
For example, 2003 Tigers...Dmitri Young had 2.7 WAR and the rest of the team was -6.6. Heh. That also means that even Warren Morris (0.7 WAR) Shane Halter (0.4) and Cody Ross (0.4) outscored the rest of the team.
October 12th, 2011 at 3:46 pm
@2... I'm not a huge fan of WAR but this is a stats blog and a stats blog would deal alomst exclusively in "statistical anomalies and thresholds" and other random notes and happenings. If I wanted the obvious stuff I could very easily figure that out on my own just looking at the box score. These guys are here to show us random things that have happened recently, and even if it's a stat I don't like/agree with/understand completely it's still a stats blog. Go read the box scores if you don't want weird anomalies...
October 12th, 2011 at 3:47 pm
Although very unlikely, in principle a person could also have a WAR 8.2 better than the second best on his team even though his WAR was less than 8.2 -- if all others on the team had negative WAR.
But from the Batting Game Finder, all teams have had at least 6 position players with non-negative WAR, so in practice this hasn't happened -- yet.
October 12th, 2011 at 3:48 pm
It was from April. I didn't follow the twists, but the Dodgers are still in court and it will be at least another month before it's settled.
In any event, it explains why they weren't able to sign anybody that can hit the ball. A normally funded & sanely run Dodgers team with Kemp & Kershaw would have won 90 games sleepwalking.
October 12th, 2011 at 4:05 pm
Vladimir Guerrero beat his Expos teammates 4 years in a row and almost did it 5 times.
1998 Expos - Guerrero 7,1 WAR, the rest of the team 3.3
1999 Expos - Guerrero 3.5 WAR, the rest of the team -0.2
2000 Expos - Guerrero 5.3 WAR, the rest of the team -5.6
2001 Expos - Guerrero 4.8 WAR, the rest of the team -5.7
Then in 2002 Vladdy was at 6.5 and the rest of the team combined to beat him by a whopping 0.2 with 6.7
October 12th, 2011 at 4:13 pm
This post had me thinking about Ernie Banks on the 1950's Cubs.
In 1959, he had 10 WAR while the next closest Cub, Al Dark had a 2 WAR.
Banks in 1959 hit .306/.374/.596 with 45 HR & 143 RBI's. That was the second highest HR total and 3rd highest RBI total for a SS up until that point. No other Cub even had more than 14 HR that year.
In 1991, Cal Ripken had a WAR of 11 and the next closest Oriole (Mike Devereaux) had a WAR of 4.1.
Ripken put up a .323/.374/.566 with 34 HR & 114 RBi, 210 hits with 368 total bases. Not to mention Ripken played excellent defense at SS. The 368 total bases was the second highest total bases for a SS up until that point.
October 12th, 2011 at 4:16 pm
I am not sure if historically it is as hard to find such a gap with pitchers.
In 1972 Steve Carlton's WAR was 12.2 compared to the 1.8 of next highest Phillie pitcher Barry Lersch. The highest batter on that team the 1.2 of Don Money.
October 12th, 2011 at 4:21 pm
T. Williams had a 11.3 WAR in his .406/.553/.735 season in 1941. The next closest Red Sox was Joe Cronin with a 4.9 WAR.
T. Williams had a 9.9 WAR in 1957 with a .388/.526/.731 season. The next closest Red Sox was Frank Malzone with a 3.2 WAR.
October 12th, 2011 at 4:24 pm
Barry Bonds had a 10.8 WAR on the 1996 Giants with his .308/.461/.615 season. The next closest Giant was Matt Williams with a 3.3 WAR.
October 12th, 2011 at 4:29 pm
Thomas Court beat me to the '72 Carlton punch, but I can add that Carlton had an edge of 12.3 WAR over the rest of the team combined
-- Carlton: 12.4 WAR (12.2 pitching, 0.2 hitting & defense)
-- Phillies' WAR minus Carlton: 0.1 (0.7 pitching, -0.6 hitting & defense)
October 12th, 2011 at 4:29 pm
Barry Bonds had a 12.4 WAR in 2004 with a .362/.609/.812 season. The next closest Giant was Ray Durham with a 3.3 WAR.
October 12th, 2011 at 4:34 pm
George Brett in 1985 had an 8 WAR, the next closest team-mate, Frank White had a 1.9 WAR.
October 12th, 2011 at 5:15 pm
Rogers Hornsby was 9.4 WAR ahead of the #2 guy on the 1924 Cardinals: 13.0 WAR for him, and 3.6 for CF Ray Blades. Unless I missed someone else for #2, which I don't think I did.
October 12th, 2011 at 5:17 pm
@18
Wow... I beat John Autin on the BR blog. That's probably a first and last for me.
Kinda makes me feel how Bill James felt when he was working Tim Raines' arbitration case and during a break arrived at the men's room first. As they were relieving themselves, James remarked, "One day I can tell my grandkids that I beat Tim Raines in a footrace." Raines laughed hysterically.
October 12th, 2011 at 5:32 pm
Babe Ruth WAR in 1920 12.8...Del Pratt 3.8
1921 13.7...Roger Peckinpaugh 4.0
1924 11.9...Wally Pipp 2.7
**1923 14.7...Aaron Ward 4.1**
Babe Ruth led the Yankees in WAR every year from 1920
until 1932 except for 1925.
October 12th, 2011 at 5:37 pm
@2 and 10, regarding WAR...
At the team level, WAR does reproduce the overall strength of a team fairly well, although not quite as well as Pythagorean Wins. Looking at 2011, the correlation between team WAR and team Wins was 87%, while the correlation between team Pythagorean Wins and team Wins was 94%.
The correlation between WAR and Pythagorean Wins was 91%. As WAR matures, hopefully that correlation would approach 100%. (The deviation between Pythagorean Wins and Wins is sometimes attributed to luck, or clutch play, so I'm not sure having WAR correlate with Wins is a reasonable goal.)
October 12th, 2011 at 7:17 pm
This one rather amazed me--I just picked a year and team at random--I have no idea if this is unusual or not.
1971, playing for Cleveland, Graig Nettle's WAR was 6.5. The rest of the non-pitchers on the team combined for a negative 5.5 WAR.
Nettle's WAR was therefore 12 points higher that of the total of the other position players on his team.
I do not see that anyone has found such a large margin for a position player up to when I posted this.
October 12th, 2011 at 8:20 pm
@22, Thomas Court -- I love that James/Raines story!
I guess I'm the one who pulled a Rock. 🙂
October 12th, 2011 at 8:46 pm
This interesting thread got me thinking about the flip side of the coin -- teams with a large WAR total but no MVP-caliber individual.
I searched for teams with the most batters at 0.5 WAR or better since 1893, then looked at the 10 teams who had at least 15 such players. Two had no hitter with 1/6 of the team total:
-- The 1949 Yankees had 17 such hitters and a team hitters' total of 30.6 WAR, led by Joe D. at 5.0. They went 97-57 and won the WS in 5 games, the first of 5 straight championships.
-- The 1986 Giants had 16 such hitters and a team total of 31.4, led by Chili Davis at 4.6. They went just 83-79 but with a pythag. record of 90-72.
-- The 1899 Chicago Orphans (future Cubs)
October 12th, 2011 at 8:49 pm
Whoops, I meant to delete the last line at #27, but I might as well flesh it out now:
-- The 1899 Chicago Orphans had 15 hitters with at least 0.5 WAR and a team hitters' total of 26.4, led by Bill Lange at 4.1.
(They met my threshold of no hitter at 1/6 of the team total, but I was also looking for closer to 30 total WAR.)
October 12th, 2011 at 8:56 pm
@26-
John-here is a another great Tim Raines story. I went to high school
in Palm Beach Gardens, FL from 1981-85. Tim Raines used to work
out with our high school baseball team.
One day a bunch of us were playing a pick-up basketball game and
Raines showed up, This must be 1984.
At one point I find myself guarding Raines. He shoots the ball right into
the palm of my hand. A blocked shot. I am six foot, taller than Raines.
maybe by three inches.
He was so mad that for the next several minutes he shadowed me on the
court. Paid no attention to what else was happening in this game, he wouldn't leave my face.
I knew the only way I could get rid of him, was to take a shot that he could
block.
Next time I got the ball, I gave him a lousy little head fake. Raines jumped
high and blocked the shot quite emphatically. Afterward, Raines left me
alone, he'd had his revenge on a high school kid who got lucky one moment..
October 12th, 2011 at 9:33 pm
1883 New York Metropolitans: Tim Keefe with 16.5 WAR, the rest of the team, 0.4 WAR
October 12th, 2011 at 11:36 pm
Interesting Whiz. Do any other systems of WAR or WARP correlate more with wins? You said WAr, I assume you mean the very specific formula used here, B-R WAR, not, say, Baseball Gauge.
October 13th, 2011 at 2:28 am
Good read over at Deadspin about Verlander and the MVP via a video game.
http://deadspin.com/5849307/the-playstation-3-says-barry-bonds-doesnt-come-close-to-this-years-true-mvp
October 13th, 2011 at 3:12 am
By any objective standard Verlander has had an excellent season, but it is debatable if it is even the most valuable this year, certainly not as valuable as Bonds in many of his best years. Or tons of other players & a bunch of pitchers. The video game rates W-L as very important, & uses no credible advanced metrics. From the the end of the above article:
The game credibly mimics the historic weight a layperson gives to a pitcher's wins and losses, which are glamour statistics largely dependent upon other factors, if not entirely discredited as a means of player evaluation.
October 13th, 2011 at 8:49 am
So, what do we learn from this experience? That great players can be on lousy teams? Is Kemp greater because his team was mediocre? And, since we bring Carlton into it, who was great amidst absurdly bad, how much of this is really team dependent (it's obviously player dependent, in the sense that you can't have another high performer the same team) and is the number more or less impressive for a pitcher?
October 13th, 2011 at 8:53 am
@13...no wonder Vlad's knees are shot...tough carrying 24 other guys 5 years running.
October 13th, 2011 at 9:22 am
@25 : I grew up in Cleveland, & can't believe you picked the 1971 Indians at random. They were epically bad, lucky to finish 60-102, a classic transition team of the worn out (Pinson, Harrelson), the unready (Chambliss, Lowenstein) & waves of AAA players in major league uniforms*. A few years later, we were still using their names to write mock-epics in high school. The attendance was 591,000, but if you take out the opener & the 4th of July it was 450,000 - under 6000/game in an 80,000 seat stadium.
*Not in appearance. They had some stretch poly crap that looked awful.
October 13th, 2011 at 9:41 am
@24 - WAR should not be expected to correlate 100% with wins because there is a timing element to wins (hence the discrepancy between Pythag and actual totals) that is noise around the run differential signal. It should be expected to correlate more strongly with runs scored (for offensive WAR) and runs allowed (for defensive WAR and pitching WAR).
The only stat which would be expected to correlate perfectly with actual wins at the team level is WPA, which is game-state dependent (and ties out by definition).
October 13th, 2011 at 10:39 am
Well, Michael Cuddyer had a WAR of 3.0, and the Twins' position players (including Cuddyer) combined for just 3.9, so his teammates were actually just 0.9. So Cuddyer actually had more than 3 times the WAR as the rest of his team combined.
The amazing thing is, Denard Span had a WAR of 2.6, so his teammates were combined for just 1.3. He doubled that.
Alexi Casilla was 3rd with 1.8 WAR, almost being more than his teammate's 2.1.
One more for the Twins, starting pitcher Scott Baker, in just 21 starts and 135 IP, posted a 3.8 WAR, almost as much as ALL of the position players combined.
Is 3.9 WAR the worst mark ever for a team's position players? Even Houston's position players posted a 15.6 WAR (though their pitching was at -0.5????)
October 13th, 2011 at 10:43 am
Well, I didn't read all of the previous posts. I guess there are plenty of other teams to post some horrid team WAR.
October 13th, 2011 at 11:41 am
@27/ John Autin: "This interesting thread got me thinking about the flip side of the coin -- teams with a large WAR total but no MVP-caliber individual.
-- The 1949 Yankees had 17 such hitters and a team hitters' total of 30.6 WAR, led by Joe D. at 5.0. They went 97-57 and won the WS in 5 games, the first of 5 straight championships..."
John A. - In 1949, Joe D. only played 76 games because he missed a big chunk of the season with a painful heel spur. If he had played a full season, he probably would've reached his usual level of 7-8 WAR (at least).
That 1949 Yankees team must've had a lot of injuries and/or platoooning (it was Casey Stengel's first year as Yankees manager), as all but one of the starters played between 103 and 116 games - Jerry Coleman (128 G) and Phil Rizzuto (153 G). Somehow they managed to win the World Series.
October 13th, 2011 at 1:35 pm
To go straight to the obvious - http://www.baseball-reference.com/teams/CLV/1899.shtml
-15.7 position, -33.9 pitching
Cleveland Spiders, of course
October 13th, 2011 at 4:09 pm
only read like the 1st 15,
ichiro?
October 13th, 2011 at 4:39 pm
@40, L.A. -- Considering the runaway success of Casey's platooning strategy with the Yankees, I think that "Somehow they managed to win the World Series" doesn't quite capture the mood.
I have not studied the matter, but it seems possible to me that one reason the Yankees of 1949-58 were so successful once they reached the WS (7-2 in series, 34-22 in games) is that they weren't worn out in the process of getting there.
October 13th, 2011 at 11:38 pm
@31 Mike, yes, it was the bbref version of WAR -- I haven't checked others.
@37, yes, I agree that WAR should not necessarily correlate with wins -- that was why I suggested it should correlate more with Pythagorean wins (which is after all related to the runs scored and allowed). I suspect that the offensive WAR does correlate very well with runs scored since that analysis is more mature -- pitching and defensive WAR are probably the dominant source of the discrepancy between WAR and Pythagorean wins.
October 14th, 2011 at 11:23 am
@43/ John Autin -
"@40, L.A. -- Considering the runaway success of Casey's platooning strategy with the Yankees, I think that "Somehow they managed to win the World Series" doesn't quite capture the mood..."
John A. - It was meant somewhat ironically, as at the time Stengel's platooning strategy was viewed with a lot of scepticism - see David Halberstam's book about that season,"The Summer of '49".
Remember, when Stengel took over the Yankees in 1949, he was seen by a lot of the press as an amusing clown and an unsuccessful manager, not any sort of strategic genius.
October 14th, 2011 at 11:00 pm
@45, Lawrence -- I must be the worst in the world at recognizing irony in the comments. Mea culpa.