Worst of the Worst

The Atlanta Braves released Melky Cabrera two days ago, after he had finished the season as the worst player in baseball, with -1.2 WAR. Among players who played enough to qualify, he was one of six with negative WAR this year, along with Carlos Lee (-0.8), Adam Lind (-0.3), Cesar Izturis (-0.3), Skip Schumaker (-0.2), and Johnny Gomes (-0.1). That’s about standard: there were five players with a negative WAR in 2009, seven in 2008, seven in 2007, and seven in 2006. Only one player had a negative WAR in more than one year, Jermaine Dye, whose lead glove produced -0.4 WAR in 2009 and -0.8 WAR in 2007.

Otherwise, though, appearances on the negative WAR list tend to be brief rather than protracted — though you still don’t hear the phrase “replacement player” escape Jon Miller’s lips very often, very few players can hang on in the majors when they’re performing below replacement level. Most of the players at the bottom of the barrel are fringy guys who tend to be around zero, give or take a win, like Mark Teahen, Jeff Francoeur and pre-2010 Delmon Young, but some of the players on the list are surprising: baseball’s second-worst position player in 2009 was its 17th-best position player in 2010, Aubrey Huff. Baseball’s second-worst position player in 2007 (above only Dye) was 23rd-best in 2009, Jason Bay.

Though there are usually a handful of position players with a negative WAR in any given year, that isn’t always the case with pitchers. This year, there wasn’t a single one, and from 2006-2009, there were only six total. (There weren’t any Jason Bays in the bunch, either. The “best” pitcher in the bunch is Jason Marquis, who posted a -0.7 WAR in 2006, then received a $21 million contract two months after the season ended, but that may have been more a reflection of Jim Hendry’s spending priorities than Marquis’s star power.)

Having a negative WAR, of course, can be a leading indicator of the end to come. Of the fourteen position players with a negative WAR in 2006 and 2007, six were out of the majors in 2010: Jermaine Dye, Ray Durham, Craig Biggio, Angel Berroa, Preston Wilson, and Shawn Green. And Jason Kendall’s career surely isn’t long for the world. We can do a similar analysis of the 18 players with a negative WAR from 2008-2010, and predict that within the next four years we’ll see the disappearances of Garret Anderson, Jose Guillen, and Carlos Lee, and very possibly Yuniesky Betancourt, Emilio Bonifacio, and Melky Cabrera himself.

Replacement level is an awfully low bar to clear, and being the worst player in baseball is not something that is easy to recover from. Whether or not WAR is a mainstream stat, the underlying truth that it measures is plainly reflected by the careers of the players who post a negative WAR. If Melky can’t clean up his act in a hurry, he’ll soon be out of baseball.




Print This Post



Alex is a writer for FanGraphs and The Hardball Times, and is a product manager for The Washington Post. Follow him on Twitter @alexremington.


35 Responses to “Worst of the Worst”

You can follow any responses to this entry through the RSS 2.0 feed.
  1. Erik says:

    Brandon Wood had less WAR (-1.8).

    Vote -1 Vote +1

    • That’s why I said “played enough to qualify.” It’s a slightly arbitrary cutoff, but there are a lot more players who suck in a relatively short amount of time. It’s much more interesting when they are allowed to suck for a season’s worth of at-bats.

      Vote -1 Vote +1

  2. Erik says:

    So, I don’t think it’s fair to say he was the WORST player of the season…though he may have been the worst to play full-time all year. According to WAR, though, which includes playing time, Wood hurt his team more than Melky.

    Vote -1 Vote +1

  3. Thomas says:

    I think that Wood should qualify for worst WAR. Abysmal production >150AB.

    Vote -1 Vote +1

  4. rmtx97 says:

    I hope the Cards dump the Skip experiment. It’s tough to watch.

    Vote -1 Vote +1

  5. Tom Zig says:

    Actually Melky was only the 8th worst player. Since WAR is how many wins you added or took away from your team and is essentially a counting stat, does it matter if the person is “qualified” or not? I think not.

    So the real list including pitchers from worst to less worse (I guess that’s how you can phrase it)

    Pedro Feliz (-2.1)
    Brandon Wood (-1.8)
    Akinori Iwamura (-1.6)
    Ryan Rowland-Smith (-1.6)
    Luis Valbuena (-1.5)
    Brian Tallet (-1.5)
    Nate McClouth (-1.3)

    and finally

    The Melkman (-1.2)

    +16 Vote -1 Vote +1

    • Max says:

      I agree completely. Why does a player have to “qualify” with a certain number of innings or AB’s? Doesn’t the nature of the statistic make that moot? A player can’t achieve an extraordinarily low WAR without having a reasonable amount of playing time anyway. Ryan Rowland-Smith had a 6.75 ERA and 6.55 FIP over 109 innings. Why shouldn’t that qualify?

      Vote -1 Vote +1

      • Mark says:

        Because in a sense, you’re creating a greater degree of arbitrary endpoint use by not giving it a playing-time baseline. At least if there’s a required minimum, you have a relatively consistent denominator in comaring sample sizes. Rowland-Smith’s “season” really only contains about 60-75% as much data as the pitchers tracked in this study. It may not be incredibly likely, but its probably less unlikely than you think that he would have thrown another 50-70 innings and had an FIP low enough over that span to get his mark over the total 160-180 inning data set into more less-than-awful territory. The same goes for a guy like Brandon Wood. If you just run a BABIP regression, even leaving all his other numbers the same, his wOBA certainly gets at least a bit better than it was. He still may have wound up worse than Melky, but his positive UZR would have also carried more relative weight and it may have been fairly close.

        Vote -1 Vote +1

    • brendan says:

      right. in fact, I think the opposite approach makes sense: who had the most negative WAR per plate appearance or per inning pitched? Surely racking up -2WAR in a month or two is worse performance than a full season of -3, correct?

      Vote -1 Vote +1

    • nemo1 says:

      I think the reason that there is a qualifying number of plate appearances is to show how consistently bad the players are. Someone that racks up more negative WAR in half a season is usually going to see some positive regression, and probably would have if they had played full time. That’s the only explanation I can really think of… although I’m not sure if I really agree with it. I see the point in both sides of the argument though.

      Vote -1 Vote +1

    • I may be mistaken, but, since a lot of the negative value is because of defense (and UZR is quite volatile over short periods of time), qualification is important as it means the player is being judged by a more stabilized statistic. That’s my understanding of how UZR must be handled, anyway. If, however, it is used in a way that clears up that issue (or I’m totally wrong), then I agree with you absolutely.

      +5 Vote -1 Vote +1

    • schmenkman says:

      Agreed, and the difference between Cabrera’s PAs (509) and Feliz’s (429) is not that great. Unfortunately for Pete Happy, he’s the “winner”. The -28.9 batting runs alone is a noteworthy achievement. .218 BA with .076 ISO and a 3.0% BB%.

      Vote -1 Vote +1

  6. delv says:

    You’ve gotta figure that, given Melky’s age (26), someone would eventually sign him and put him in their minor league system… and that he’d eventually come back up at some point closer to his prime when he’d likely put up better numbers. By contrast, most of the players named above are at least 35 yrs old.

    Vote -1 Vote +1

  7. SteveS says:

    I’m not sure whether it says more about me or them, but I’m a little shocked not to see any Mariners on this list – is it the qualifying requirements?

    Vote -1 Vote +1

  8. Zach says:

    Come on, Jonny Gomes isn’t sub-replacement, he has 86 RBI! That only 14 less than 100! ;-)

    Vote -1 Vote +1

  9. philosofool says:

    I’m not sure that using the normal “qualified” cut off on PA makes sense here. A platoon player won’t qualify because he’ll get just half a season or so worth of work in. Nevertheless, his negative WAR counts against the team’s success. Is -1 WAR in a platoon. Casey Kotchman, really better than the same as full time player, Melky?

    Vote -1 Vote +1

  10. Steve says:

    I hope that the yankees can pick him up for really cheap now and get him motivated again – I would love to have him in pinstripes if he’s putting forth effort!

    Vote -1 Vote +1

  11. Xeifrank says:

    In Melky’s defense, it is his UZR that brings his fWAR down below replacement level.

    Vote -1 Vote +1

    • Basil Ganglia says:

      So it is Melky’s defense that brings his WAR down below replacement level.

      Vote -1 Vote +1

    • Bronnt says:

      The .294 wOBA from a corner outfielder certainly didn’t help his cause.

      Vote -1 Vote +1

      • Xeifrank says:

        A 0.294 wOBA from LF is 0.1 WAR. His offensive production in 2010 was basically replacement level. UZR knocked his fWAR way down below replacement level. Fan scouting has his defense at -0.1, which would bring his fWAR down to exactly replacement level for the year. Just something to keep in mind.
        vr, Xeifrank

        Vote -1 Vote +1

  12. Gio says:

    Fangraphs; where you have to qualify to be the Worst of the Worst. You’re better than this Alex.

    Vote -1 Vote +1

    • The reason I made sure they qualified is that it’s interesting that a player with negative Wins Above Replacement wasn’t replaced. Brandon Wood, obviously, lost his major league job. Melky never did.

      Vote -1 Vote +1

  13. Kyle says:

    Agree with you, Alex and Zach. This isn’t all telling ALWAYS… Gomes is a power hitting DH playing in the National League, and had to play more than was expected due to the Reds’ injuries. He’s paid less than a million a year and made for the American League. Obviously he’s an inept fielder, but the Reds have so many great fielding players around him, I guess they’ll take one weak link in the outfield with 5 of 6 great fielders around him.

    As for Lind, he’s a power hitting DH who plays for Toronto… I don’t see how he could have managed to be such a mediocre player by playing only 16 games in the outfield. A year ago he hit .305, 35 homeruns, and drove in 114 RBI. Since he just turned 27, and hit 40 points under his career average, I doubt he’ll have a repeat next year. He barely makes 550 grand a year, and I bet many teams would love to have him at that cost. Being a guy a year removed from owning a RAR of 35.3. I know the Red Sox would love to have him at that cost.

    Do you think the Giants wouldn’t have gambled on Lind if he had been available over a Burrell, Huff, or one of the other guys they picked up just hoping for some run production so their pitchers could actually have a chance to get some wins? The Blue Jays weren’t dumb enough to give up on Lind… while Delmon Young is the only reason the Twins didn’t collapse with the injury to Justin Morneau and power outage of Mauer. He was their only player to have at least 100 RBI…

    Question… how is Posada not on here? The guy hit .248, and in 83 games as catcher had 8 errors, 8 passed balls, while throwing out guys at a clip of 18 percent. He also struck out 99 times in 383 at-bats. This is where sometimes geek stats just don’t show EVERYTHING. How did his numbers not end up being below WAR? I just don’t get it I suppose, because in 36 games he didn’t catch. Are his offensive numbers even close to most DH’s in baseball? I think not…

    Vote -1 Vote +1

    • Jason B says:

      Not even close – his bat was above-average. .357 OBP, 122 wRC+, .357 wOBA, walks in 13% of plate appearances, 18 HR in 383 AB. Even with woeful fielding and K’s in a quarter of his plate appearances, Jorge scores 2.4 WAR (per FanGraphs methodology).

      Vote -1 Vote +1

      • Kyle says:

        Posada is also paid 13 million, and has had many pitchers complain about the way he calls a game. I wonder how many runs he’s given up by being the easiest catcher to steal a base off of… I know he’s not a horrible hitter, but his stats are similar to the guys I mentioned(except they have lower OBPs….

        I was basically just being critical of this basically pointless article. Read what B N said below… I totally agree.

        Vote -1 Vote +1

  14. B N says:

    I don’t normally like to spend time criticizing these posts, because I enjoy reading them, but the following comment rather bothers me:

    “We can do a similar analysis of the 18 players with a negative WAR from 2008-2010, and predict that within the next FOUR YEARS we’ll see the disappearances of Garret Anderson, Jose Guillen, and Carlos Lee,”

    GASP! Someone predicting that Garret Anderson and Jose Guillen being out of baseball in 4 years time? Astonishing! But more seriously, that is such a long time frame as to make the statement almost useless. We’re talking about a bunch of aging players coming off some horrible seasons. Heck, Anderson is 38 and he should have been out of baseball a year or two ago.

    A very significant portion of the entire MLB will be out of the majors in 4 years, due to injury or ineffectiveness. Among veteran players in their 30′s, that is probably the majority of players. And not necessarily ineffective players either. Mike Lowell just retired this season after an injury plagued but useful performance, for example. So, not to knock on you too much but I just was not sure what that added to the piece, which I otherwise liked.

    Plus, I’m pretty sure the concept is wrong. Players don’t end up out of the league as the result of having strongly negative WAR. The very fact that they were able to accumulate that level of badness means that someone believed in them enough to trot them onto the field regularly. Hence, a WAR that bad indicates that there was enough upside to wait out the slump/decline. It’s very similar to what you see in fantasy baseball. The guys who kill your team aren’t scrubs- they’re stars like David Ortiz who have a 3 month stretch of hitting under 200 with no home runs. You gets killed because you’re waiting for them to break out. The same issue happens with real life clubs.

    So what you seem to be taking as a metric for being out of baseball, I’d take as a metric for busted upside. These are guys who have burnt one of their chances to be a productive player, but were stuck by due to upside and/or veteran reputation. As a result, most of them would latch on somewhere else because they’ve only burnt one of their shots. I think guys who have had a series of two low amplitude seasons would be more likely to be out of baseball the following year. That would indicate that they aren’t contributing much to the team at all, due to lack of playing time or lack of performance. With the exception of utility players, that’s not usually a recipe to be a regular player.

    Vote -1 Vote +1

    • B N says:

      On a side note, this issue also happens in poker. The hands that kill your performance aren’t your worst hands- they’re the hands that are good enough to stick with them due to their upside, which never materializes (see: chasing a flush, staying in with pocket aces when a pair of K shows up on the flop, etc). While the expected value of those hands may be neutral or positive, on a subsample of hands they can rack up the largest cumulative losses.

      Bad GMs and managers, much like bad poker players, have trouble gauging when to fold on their players. Or, alternatively- good ones might just be ones who were lucky enough to see their players rebound… but I doubt it.

      Vote -1 Vote +1

      • I think your analysis is right. (And yes, four years is a long time, and obviously I think it’ll be a lot sooner for Anderson. I basically wanted a time frame that would allow me to include a lot more of the players on the lists.) Obviously, I’m not saying that WAR is the reason that these players are out of baseball. I’m saying that, while it isn’t an entirely mainstream stat, what it’s measuring is entirely mainstream — not just wins, but the amount of time a player is likely to remain in baseball.

        Vote -1 Vote +1

  15. designated quitter says:

    Two things: BN is right- look at any group of randomly chosen players and half will be gone in 4 years- the average career is only 4 years and change, I think.

    Instead of wasting time with the worst players, come up with a statistical measure of broadcasters. The figure out why the ones with lowest TAR (Talk Above Replacement) aren’t gone in 4 years. (Stop cringing in the corner, Joe Morgan).

    Vote -1 Vote +1

  16. CFIC says:

    Feliz is the worst of all

    Vote -1 Vote +1

Leave a Reply

Your email address will not be published. Required fields are marked *

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>