What Have we Seen at the Projected Worst Positions in Baseball?

There are a few annual series we run here at FanGraphs. One of them is the annual trade-value series, where Dave gets to write glowingly about Mike Trout. Another one of them is the annual positional power rankings series, where some randomly assigned lucky author gets to write glowingly about Mike Trout. The positional power rankings are intended to highlight the strongest positions in baseball, but you can’t identify the strongest without also identifying the weakest. Granted, the rankings are only as strong as the projections, but the projections are solid, and so, here were the five weakest projected positions coming into the 2014 regular season:

  • Mets bullpen, -2.3 standard deviations from average
  • Astros rotation, -2.1
  • Marlins first basemen, -2.1
  • Marlins third basemen, -1.9
  • Blue Jays second basemen, -1.9

Nothing in there was shocking. And a lot of that stuff wasn’t particularly relevant. Only the Blue Jays, out of that group, were looking to contend. So there was a lot of conversation in spring training about what the Blue Jays were going to do at second base. Their solution back then: nothing! The situation, however, has changed. All of the situations have changed to some degree, because the future always introduces new information. The point of all this: Those were the five worst projected positions coming into the year. How have the players at those positions actually done? How have the teams actually maneuvered?

Mets bullpen

  • Projection: -2.3 standard deviations from average
  • Performance: -1.7

The Mets were projected to have the worst bullpen in baseball, with the most valuable arm being Bobby Parnell‘s. This year, Parnell made one appearance and then was told he needed Tommy John surgery. In a sense, then, the Mets’ bullpen has over-performed. Without Parnell, this unit was supposed to be particularly dreadful. Now, by WAR, it’s still sucked. It’s actually second-worst in baseball, and below replacement-level. Confusingly, the unit is middle-of-the-pack by RA9-WAR, because they have a 3.14 ERA to go with a 3.77 FIP. It’s been helpful to get so much out of Jenrry Mejia. Jeurys Familia is running another sub-2 ERA. But it shouldn’t be forgotten that, this season, the Mets’ bullpen has featured Daisuke Matsuzaka, Jose Valverde, and Kyle Farnsworth. Maybe that tells you enough about how things have gone, but while the bullpen hasn’t exactly helped the Mets to hang in the race, they were never supposed to hang in the race in this season, and there are bits and pieces there that could contribute for the next good team. You worry about the bullpen when the rest of the roster is good enough to make the bullpen matter.

Astros rotation

  • Projection: -2.1 standard deviations from average
  • Performance: -0.2

An excerpt from the positional power rankings:

But that future is not here, and it’s not even all that close yet. So you get Scott Feldman, staff ace. I like Scott Feldman, but when he’s not only your ace but projects for as much value as your #2-#4 starters put together, well, that’s not great.

One thing we know the Astros don’t have is a wonderful starting rotation, today. But it’s absolutely been better than expected, to the point where Jeff Luhnow has expressed a willingness to move a starter because the team thinks it might have too many. The Astros’ rotation has been basically average, sandwiched between the A’s and the Royals, and you already know the big reasons why. About Feldman projecting for as much value as the other guys behind him put together: Feldman’s got a WAR of 0.6. Four other guys are at 1.2 or better, with Dallas Keuchel leading the way and Collin McHugh coming completely out of nowhere. Keuchel, we maybe should’ve seen coming to some extent. McHugh, no way. That’s just a miracle. Maybe your version of a miracle doesn’t have a 94 FIP-, but considering what the Astros could’ve been dealing with, the rotation has kept them from being a total embarrassment.

Marlins first basemen

  • Projection: -2.1 standard deviations from average
  • Performance: -0.8

Fun fact: Marlins first basemen, combined, projected for -0.4 WAR. Garrett Jones didn’t project well, but at least the Marlins signed him for two guaranteed years. Now, Marlins first basemen are on pace to be worth -0.3 WAR. Jones, it turns out, still isn’t good. But while the team is hitting its positional projection, it’s still better, relatively speaking, than the expectation, because other teams have had disasters. The Astros are already at -1.9. The Rangers are at -1.7. Seven teams are in the red, with the Marlins being the best of them. Marlins first basemen haven’t over-performed. Other first basemen have under-performed. The internal suck level has been consistent, but the external suck context has made for a different landscape.

Marlins third basemen

  • Projection: -1.9 standard deviations from average
  • Performance: 0.2

So the Marlins didn’t project very well as a team. This, right here, was totally understandable. The Marlins were in position to give the playing time at third base to Casey McGehee and Ty Wigginton. Over McGehee’s two previous years in the bigs, he’d been below replacement. Wigginton had been a catastrophe since 2008. McGehee was coming off a year in Japan where he did well, but he had a track record of not being very good. McGehee, at this writing, has two home runs. He also has a 114 wRC+ and a BABIP closer to .400 than .300. The Marlins have expressed reluctance to trade him for some reason. This is just one of those weird things. McGehee and Jose Fernandez, together, were projected to be worth 4.7 WAR. They’re on pace to be worth, together, 4.7 WAR. They’re just getting there in a different way than we thought. McGehee and the Marlins at third base have been as productive as David Wright and the Mets at third base, who ranked No. 2 in the positional power rankings.

Blue Jays second basemen

  • Projection: -1.9 standard deviations from average
  • Performance: -0.1

Maybe the oddest thing about last offseason was the way we were all forced to learn the name Ryan Goins. He was the guy the Jays were willing to say they were happy with. They tried to do some things — they tried to get Ian Kinsler — but they didn’t try hard enough, which is why Goins looked like the starter at second out of spring training. Our projections split time between Goins, Maicer Izturis, and Munenori Kawasaki, and in all, Jays second basemen projected to be almost exactly replacement-level.

So far they’ve been average, as a unit. They’ve out-performed the Rays at second, and the Rays ranked No. 2 in the positional power rankings. They’ve been even with Jason Kipnis and the Indians. Goins has been lousy, but he hasn’t even batted 100 times, because second base has also included Brett Lawrie and Steve Tolleson. Allowing for that flexibility has been the semi-emergence of Juan Francisco, and now the Jays have added Danny Valencia, who makes for an intriguing Francisco platoon partner. So when the Jays get healthy again, they’ll still have the option of having Lawrie at second, and that’s just more and more time we can all spend not watching Ryan Goins play baseball.

Coming into the year, Toronto’s second-base situation looked like one of the very worst situations among potential contenders. Through four months, they’ve done more than just manage — they’ve done pretty well for themselves, to the point now where they might not need to address the position in the next few days. What Toronto hasn’t done is identify a long-term solution at the spot, but you can’t have long-term solutions everywhere all the time, and mixing and matching has served them well. For as much criticism as the Jays’ front office received after a winter of little activity, they never panicked and they never had to. That inactivity currently has them in a playoff position.




Print This Post



Jeff made Lookout Landing a thing, but he does not still write there about the Mariners. He does write here, sometimes about the Mariners, but usually not.


35 Responses to “What Have we Seen at the Projected Worst Positions in Baseball?”

You can follow any responses to this entry through the RSS 2.0 feed.
  1. Angel dust in the outfield says:

    Giants 2B wins all the bad prizes.

    Vote -1 Vote +1

  2. Ben says:

    i dont know why i keep caring about projection numbers like this. every single one of them are wrong and, in most cases, wrong by a lot.

    -10 Vote -1 Vote +1

    • Jason B says:

      Can we see all of your preseason projections? I bet they were all spot-on!

      Vote -1 Vote +1

      • Ben says:

        my numbers were just as wrong. im just realizing how useless numbers like this are. i wish we’d run some tests on non-trivial projections from the last 5-10 years to evaluate their accuracy. i’d be surprised if it was more than 50% (aka useless).

        Vote -1 Vote +1

        • Jason B says:

          Fair enough. I would hazard a guess that projections fare considerably better than ordering teams randomly.

          Vote -1 Vote +1

        • Tim says:

          FWIW, always take negative WAR projections with a grain of salt. After all, the notion of WAR suggests that teams can do better with ~freely available waiver-wire talent. And that’s what’s happened in the above cases, mostly.

          Vote -1 Vote +1

    • Josh says:

      Its almost as if you are very wrong:

      Fun fact: Marlins first basemen, combined, projected for -0.4 WAR. Garrett Jones didn’t project well, but at least the Marlins signed him for two guaranteed years. Now, Marlins first basemen are on pace to be worth -0.3 WAR. Jones, it turns out, still isn’t good. But while the team is hitting its positional projection

      Vote -1 Vote +1

    • Joe says:

      Fangraphs etc does a poor job of explaining the point of projections really.

      They’re not a projection of what WILL happen, its basically an over/under mark. 50 percent of possible results are better, and 50 percent are worse.

      Really we should be demanding some measure of what the variance is though! We need to know whether a projected 2 WAR guy should be expected to be between 1.5-2.5 WAR, or 0-4 WAR, or whatever. It would help contextualize it much better.

      +21 Vote -1 Vote +1

    • My echo and bunnymen says:

      To some extent they are incorrect but the projections seem to be the most likely event to occur. Mike Trout was projected for 7-8 WAR and he is on pace to greatly stick it to that number, again. The projections are incredibly valuable because I would rather bet on Mike Trout and the high likelihood of his great play versus Casey McGehee. The projections are incredibly informative if you can understand. Perfection is not what you seek.”

      Vote -1 Vote +1

    • Josh 2 says:

      Everyone seems to complain when projections are wrong, but no one celebrates when projections are right.

      Vote -1 Vote +1

    • Catoblepas says:

      To add to the chorus of people explaining why you’re being dumb: the extreme projections (the ones this article is written specifically about) are also the ones most likely to be wrong. You’ll notice that not a single one of these teams was WORSE than their preseason distance from the mean; this is not unsurprising, since more extreme projections are generally going to be more uncertain. Drawing a conclusion based on a subset of the predictions, rather than a quantitative analysis of the whole, is dumb.

      Vote -1 Vote +1

    • John Stamos says:


      Vote -1 Vote +1

    • Joshua Northey says:

      Typically you would expect the worst and best projections to regress towards the middle, which is what happened here.

      Vote -1 Vote +1

  3. Jon L. says:

    “The internal suck level has been consistent, but the external suck context has made for a different landscape.”

    – deserves some kind of writing award

    +21 Vote -1 Vote +1

    • joser says:

      Indeed. In fact, I think Mr. Sullivan may have stumbled into an important concept, one that I would like to see Fangraphs employ at every opportunity. Imagine Mr. Cistulli nattering on in the podcast about some anonymous California League starter, only to be interrupted: “But what you’re forgetting,” interjects Mr. Cameron, “Is the external suck context.”

      Vote -1 Vote +1

  4. Ballfan says:

    fun fact : Goins had a career night last night with 4 hits and 4 RBI, albeit in a 14-1 thrashing

    he’ll not ever likely do it again, but the timing is impeccable!

    Vote -1 Vote +1

    • FanGraphs Supporting Member

      I saw that and tried to ignore it :(

      +7 Vote -1 Vote +1

      • Radivel says:

        I saw that and wanted to go remind Red Sox fans everywhere that their team is being destroyed by Ryan Goins. It would be nice to be able to say that twice, but I doubt it’ll happen.

        Vote -1 Vote +1

      • everdiso says:

        I’m angry at the jays for sticking with goins, but even i have to admit that the guy is a beautiful fielder. His glove is wow even at SS.

        Vote -1 Vote +1

  5. Nick says:

    Of all Mets relievers worth negative WAR, only Matsuzaka will actually pitch for them again this year, and a couple of those will be serviceable starts.
    OK, maybe some more Germen in September, but probably not. I mean, please no.

    Vote -1 Vote +1

  6. Za says:

    Of course, the blurb on the Mets’ bullpen is disingenuous. The relief corps was especially bad during April (-0.7 WAR, 27th) and May (-0.3 WAR, 27th) but is greatly improved. They put up 0.2 WAR in June (22nd) and 0.4 WAR in July (18th). Moreover, fWAR is inherently flawed at accurately gauging RP performance as it doesn’t, AFAIK, take leverage index into account. Moreover, RA9 WAR does a much better job indicating what actually happened – showing the Mets bullpen right now is an average pen and definitely not the worst or second worst in the game.

    Vote -1 Vote +1

    • FanGraphs Supporting Member

      What happened at the start counts, when you’re looking backwards. This wasn’t about projecting the Mets bullpen going forward. This is just comparing what they’ve done to what they were expected to do. And for the record, leverage is taken into account. If you split the middle between WAR and RA9-WAR — which I think is a fair compromise — the Mets come out at 0.8, fourth from the bottom.

      +5 Vote -1 Vote +1

    • Thor says:

      “Disingenuous”!?!?!?! How can the statement be disingenuous when it is true? Great, they have had a decent July. But just because they’ve had a decent July doesn’t mean you should ignore April and May — a full 1/3 of the season! Your analysis is akin to saying that the projections and critique of Goins are disingenuous because he had 4 hits and 4 RBI last night against the Red Sox. Additionally, you have to also consider who they played in July (it’s not like they faced all teams). They played the Braves (6 games), Rangers (3), Marlins (3), Padres (3), Mariners (3), Brewers (4), and Phillies (1). Not exactly the murderers’ row of offenses in July. So, settle down, turn Keith Hernandez back up on the radio, and enjoy Josh Edgin tearing up the NL and that amazing Mets bullpen. But watchout, the August schedule actually pits the Mets against top tier offenses, so we’ll see how they fare.

      Vote -1 Vote +1

  7. LongTimeFan1 says:

    There’s no sabermetric equivalent for using one’s eyes and common sense to make assessment. The conclusions drawn by Jeff Sullivan on the Mets pen are woefully devoid of current reality.

    After ridding this team of early season under-performers including Valverde, Farnsworth, the injured Scott Rice and the struggling John Lannnan, this pen is a major strength, power pitching, lights out set up and closer, backed by Edgin, Vic Black, Dana Eveland and others.

    Mejia: Closer – 15/17 save conversions, 1 ER in his last 10 appearances. 13 K’s in those 10.2 innings.

    Set-Up: Familia – Last 10 appearances, 0 earned in 9.2 innings.

    1.88 ERA, 1.14 WHIP in 50 appearances spanning 52.2 innings

    Vic Black: 28 appearances, 1.88 ERA, 1.25 WHIP.

    Dana Eveland – 17 Games, 2.00 ERA, 1.0 WHIP

    Josh Edgin: 29 Games, 1.59 ERA, .084 WHIP.

    ————————
    It’s so annoying when people who know not what they’re talking about pretend they do and offer current and future status summary on their own nonsense.

    What A Joke:

    “Maybe that tells you enough about how things have gone, but while the bullpen hasn’t exactly helped the Mets to hang in the race, they were never supposed to hang in the race in this season, and there are bits and pieces there that could contribute for the next good team. You worry about the bullpen when the rest of the roster is good enough to make the bullpen matter.”

    – Jeff Sullivan

    ———————
    This bullpen is an enormous strength right now and well into the future with young power arms, secondary pitches, and guys who love doing what they do. Sullivan is clueless.

    -13 Vote -1 Vote +1

    • AC of DC says:

      In cases like this, it’s the subtlety that makes the satire.

      Vote -1 Vote +1

    • Josh 2 says:

      Jeff’s assessment: Mets bullpen used some retreads that didn’t work out(Farnsworth, Matsuzaka, Valverde); some bits and pieces (Mejia, Familia, maybe some others) could be a part of the next good Mets team.
      Your assessment: Because the Mets got rid of the retreads, now the pen is a great strength.
      This may be true but it doesn’t change the fact that Matsuzaka, Farnsworth, and Valverde threw ~110 of replacement level/below replacement level innings. Those still count in the seasonal assessment of how the Mets bullpen has done (even though some of those innings came in Matsuzaka starts). Just because the Mets are better now than they were before doesn’t mean the earlier part of the season never happened. This wasn’t a current assessment of how the bullpen is composed right now. It’s a seasonal assessment of how they did compared to their projections.
      On a side note, I also don’t think Jeff was factoring in how much the relievers liked doing what they do…

      Vote -1 Vote +1

    • bill says:

      Sullivan seems to agree with you. They have useful pieces going forward. However, this is not a projection piece. This article is discussing objective measures of reality. The Met’s bullpen this season has cost them 1.7 wins. I’m with you and Sullivan in your belief that the Mets have found useful pieces. Their bullpen will be better the rest of the way and probably next year, as well, but as a group, they have not been good this season.

      Vote -1 Vote +1

      • vivalajeter says:

        I don’t think it’s accurate to say the bullpen has cost them 1.7 wins though. For one thing, it can’t cost you part of a win. Secondly, WAR doesn’t really work that way. If you simulate the season a thousand times, then a relief crew with their K, BB and HR rate will lose, on average, 1.7 wins less than a relief crew filled with replacement players. That does not mean that this bullpen has cost them 1.7 wins though.

        I just checked baseball reference WAR, and they’re at +2.1 WAR for the bullpen (not including Mejia, since I’m not sure how to separate out his starting WAR and relief WAR).

        I think I’ve said it before, but I think Citi Field just has a weird impact on WAR. Whenever I look at a pitcher’s numbers and their WAR, it doesn’t add up. I’m not sure if the stadium factor isn’t taking account the new dimensions or if it’s something else, but it doesn’t pass the smell test for me.

        Vote -1 Vote +1

    • Michael says:

      He is accounting for the season as a whole and while I will agree that the Mets have found some good arms, it’s not like they have been doing this all year.

      Currently in their bullpen

      Vic Black: Good arm, not that many IP (24) but solid

      Buddy Carlyle: Career 5.46 ERA. Probably should be replaced.

      Josh Edgin: Basically a LOOGY, nice to have but not a big part of a projection.

      Dana Eveland: Not many IP and his career ERA (5.3) us nothing to be happy about.

      Familia: Very good, it’s basically his first year.

      Meijia: Good but not dominant stats. Just over 9 K/9 and under 2 BB/9.

      Carlos Torres: Seems to have improved, better K/9 and FIP.

      Basically the whole group is alright, but since only Torres and Familia have logged significant IP I’m going to argue SSS for most of these guys. Plus projections don’t work for people who weren’t in the org or were call-ups. Nobody strikes me as dominant.

      Vote -1 Vote +1

    • Luke says:

      Oh my gosh! The Mets relievers love doing what they do?! Wow. What a strength! Bet they win the World Series next year. I hear the relievers on them other teams absolutely HATE their jobs.

      Vote -1 Vote +1

  8. Thanks, Comcast says:

    I imagine the Astros’ rotation projection is the result of not putting enough emphasis on batted ball outcome for pitchers. Cosart and Keuchel are two of the best in all of baseball in that regard. FIP takes into account homerun suppression to some extent, but largely grades pitchers on strikeout and walk rates. Based on what Keuchel and Cosart did last season in that area, I thought the projections were awfully pessimistic.

    Vote -1 Vote +1

  9. Grimey the Geek says:

    Next year i bet the over

    Vote -1 Vote +1