22 Responses to “Teams Who Have Gone To WAR This Season”

1. Jamee Says:
   August 27th, 2010 at 1:55 pm

   Good teams are good.

2. Steve Says:
   August 27th, 2010 at 1:56 pm

   Ludwick doesn't play for the Cardinals anymore.

3. Steve Lombardi Says:
   August 27th, 2010 at 2:03 pm

   Fair point Steve - but, his stats for them, when he was there, still count towards their record, etc.

4. hooplah Says:
   August 27th, 2010 at 2:05 pm

   The Padres have Ludwick now, although his numbers for the Padres haven't been too stellar. His line so far is .224/.306/.376 in the 24 games he's played with them.

5. John Q Says:
   August 27th, 2010 at 2:13 pm

   The Padres have two players, Gonzalez & Headley with about 10 WAR so they get a lot of production from two players.

   Headley seems to be one of the most underrated/least appreciated players in baseball this year.

6. Johnny Twisto Says:
   August 27th, 2010 at 2:20 pm

   I would be interested in seeing which teams are over- or underperforming their total WAR thus far. In theory, a team's WAR should equal its Wins minus ~40 (however many games a replacement team would win).

7. Jeff Says:
   August 27th, 2010 at 2:24 pm

   Interesting point Johnny. But shouldn't all the teams be about the same if you subtracted WAR from Wins? If they're not the same, that would seem to be a flaw in the stat (IMHO).

8. Dave Says:
   August 27th, 2010 at 2:37 pm

   If there's one thing good about the Pirates it's the fact that you don't have to look all over for their records and things. I literally got confused earlier this year when the Pirates were not in last place for their division. I looked in the last place spot but they were not there!
   It's good to know things are back to normal.

9. Michael E Sullivan Says:
   August 27th, 2010 at 2:41 pm

   WAR measures the expected runs produced by each action. If one team is lucky, or produces more runs in situations where runs are more valuable, they will outperform their WAR. If every teams Wins was equal to the total team WAR + some constant, then WAR would be measuring something different -- something a lot more like what WPA already measures. WPA should do this. The total team WPA + 81 should (within a game or two due to rounding errors) equal how many games that team wins.

   WAR is something else, it is trying to measure the expected runs produced by each action in a neutral context to get a better idea of a player's actual value.

10. birtelcom Says:
    August 27th, 2010 at 2:57 pm

    I did a quick check on the 2009 (most recent full season) NL East: according to my calculations Florida and Washington each won about 50 more games than their WAR totals, and the Mets 52. That seems about right -- the low .300s in winning percentage as a "replacement level" team. The Braves and Phils were the outliers on each side -- the Braves won only 46 more games than their WAR total while the Phils won 58. The average for the five teams still comes out to about 52 or so. The 1962 Mets, famously winners of only 40 games, show a minus 3.4 WAR for their hiting/defense, but a +3.2 in pitching WAR, giving them 0 WAR overall and 40 more wins than their WAR. I think most sabermetrics folk would estimate that a replacement level team should be able to exceed the '62 Mets' .250 win percentage.

11. DavidRF Says:
    August 27th, 2010 at 3:15 pm

    @6
    Here's this years team WAR leaders (bat/pitch/tot, actualW-L)

    NYY...25.3/13.8/39.1....78-49
    MIN...21.0/17.3/38.3....73-55
    TBR...20.1/17.3/37.4....78-49
    BOS...23.8/11.8/35.6....73-55
    SDP...22.4/11.4/33.8....76-50
    ATL...21.2/12.5/33.7....73-54
    STL...19.6/12.6/32.6....68-57
    TEX...14.7/17.2/31.9....72-55
    TOR...21.7/09.7/31.4....66-61
    CHW...13.6/17.6/31.2....69-58
    SFG...13.5/16.8/30.3....71-57
    CIN...20.4/07.0/27.4....73-54
    PHI...14.0/13.1/27.1....70-57
    ....
    BAL...04.4/07.3/11.7....45-83
    ARI...11.0/00.6/11.6....50-78
    HOU...-1.1/10.1/09.0....58-69
    PIT...-2.8/02.7/-0.1....43-84

    bb-ref should have this chart somewhere instead of split between the batter and pitching pages. It would save me a lot of typing. 🙂

12. jiffy Says:
    August 27th, 2010 at 3:29 pm

    These charts do a good job of showing why Cubs fans hated Lou Piniella this year for batting Lee/Ramirez #3 & 4 all season, as well as Theriot 1st, Byrd 5th, Soto 8th, on a team with really good pitching. Always makes sense to have a .400 OBP player hitting 8th while your cleanup guy isn't even slugging 400 into June. Good riddance to fat and tired old men.

13. Drew Cobb Says:
    August 27th, 2010 at 3:49 pm

    Can't believe Jeter doesn't have a 2 in WAR this year. Could be on the beginning of his descent????

14. Ed Says:
    August 27th, 2010 at 11:26 pm

    Pretty amazing that Carlos Santana made the list considering he only played in 46 games before his season ending injury.

15. Ed Says:
    August 27th, 2010 at 11:34 pm

    @13
    Just looked at Jeter's splits. In addition to his normally poor fielding, his been horrible against right-handers (651 OPS) and on the road (617 OPS). Interpret as you'd like.

16. Ghost of Horace Clarke Says:
    August 28th, 2010 at 12:19 am

    So the Detroit Tigers and Milwaukee Brewers are a better team than the Reds this year....and the Reds are no better than the Diamondbacks....based on this new ....almighty stat. But of course. stats don't lie.

17. Rich Says:
    August 28th, 2010 at 4:29 am

    @16
    I'm not a huge fan of WAR myself but that's not what this post is saying at all. The Brewers offense is pretty good. Their pitching is horrible.
    As far as the Reds, if this post had been made a little bit later, or if the cutoff of 2.0 had been different, they'd have several more players on this list. Drew Stubbs and Ramon Hernandez are at 1.9 and Jay Bruce is at 1.8

18. Basmati Says:
    August 28th, 2010 at 5:30 am

    This tells us how many players have contributed above a certain threshold. The Reds are obviously doing well because of good contributions from lots of players, as opposed to a few great players and nobody else doing anything.

    I posted a while back how the Angels won 100 games without anybody having >5 WAR (All star level). In other words the correlation between team performance and your best 1 or 2 players is quite low.

19. Jeff Says:
    August 28th, 2010 at 4:28 pm

    The problem with a stat such as WAR is that there are positional adjustments to try to offset that 1B hit more than 2B (i.e.). There's two problems with that general line of thinking.

    1) What about players who play 2, 3, 4 postitions? Which adjustment do you use? 2) and more importantly, nobody in the history of the game ever went to the plate as a 1B. Every batter goes to the plate as a batter.

    Assign them value for what they did at the plate. Then assign them value for what they did in the field (if that's really possible. Fielding stats generally suck compared to hitting and pitching stats).

20. DavidJ Says:
    August 28th, 2010 at 5:23 pm

    Jeff, that's not a problem with WAR; that's the whole point of WAR. Everything else being equal, the ability to play SS is more valuable than the ability to play 1B (because a replacement-level SS is going to be a worse hitter than a replacement-level first-baseman). WAR simply takes that into account, since the purpose is to measure a player's total value. If you just want to compare two players' offensive abilities independent of what positions they play, well, there are already countless offensive stats you can look at. The point of WAR is to give us the bigger picture.

    As for players who play multiple positions, the adjustment (as far as I understand it) simply depends on how many games they play at each position (e.g., a player who plays 100 games at 1B, 40 at 3B, and 20 in RF gets 100 games' worth of 1B adjustment, 40 games' worth of 3B adjustment, and 20 games' worth of RF adjustment).

21. paul Says:
    August 29th, 2010 at 11:06 am

    As suggested by post 17 (Rich), dichotomized cut-scores are almost always misleading. A regression model (GLM) with a team's summed WAR predicting winning percentage after accounting for confounds (not sure this is necessary) would be a robust test of WAR's predictive validity. I see two problems with my approach: Calculating a team's summed WAR is difficult in part because players change teams (post 2) and play multiple positions (19). The latter means that the WAR for player x as a shortstop (say, y) should be z and WAR for player x as a 3rd basemen (say, m) should be n. But the regression exercise might be helpful because useless statistics will continue to proliferate and excellent ones will continue to be regarded with skepticism (post 1, post 16) until statisticians demonstrate the predictive validity of each statistic, using appropriate criteria, like winning percentage.

22. David Says:
    August 30th, 2010 at 9:43 am

    Ah Brewers. If only you had more than 5 players above replacement level, maybe you'd be in the hunt for a division title.