How much more rope for Andruw Jones?

13 thoughts on “How much more rope for Andruw Jones?

  1. jay_robertson

    How much rope? One vote for "not much." If we're upset by the production of Jeter, Swisher, Posada, Texiera, and for the last few weeks, A-Rod, then why keep someone on the bench and at the plate with those numbers?

    Unless its a case of playing the option/call-up game with a player's time, I would guess sooner rather than later. By mid June at the very latest. The team has from 3-5 black holes in its production at any time – it would have to make Girardi's job a lot easier if he could sit whoever is cold secure in the knowledge that the player he's putting in will do a better job. With Chavez on DL, his options are Jones, Nunez, Cervelli, and Dickerson. Not exactly Murderer's Row.

    • Ted Nelson

      "why keep someone on the bench and at the plate with those numbers? "

      Because it's a ridiculously small sample size. What he did in that 10 game sample (5 PAs/game… 50 PAs = 10 games) does not predict what he'd do in his next 10 games. Jones has struggled to a .260 wOBA this season, but he's done well the past two seasons: .364 and .338.

      Even this season, the problem is really just May. In April he had a wOBA of .334 (not stellar, but for a 4th OF you'll take it), and in May to date it's dropped to .197… This is what baseball people call "a slump." It's been 9 games and 27 PAs in May (plus whatever struggles he may have ended April with)… 5 games worth of PAs. If a starter has a cold 5 games, you don't cut him. Same goes for a proven bench player like Jones.

      • BrienJackson

        Well, yes you do, because to some extent he's there to be DFA'd. He was signed to a cheap one year deal as an expendable part. He's mostly got one job; hit lefties. If he doesn't do that, you jettison him and try to find someone who can. That's the whole idea of signing cheap bench parts to one year deals.

  2. andrew

    where is montero?

  3. Ted Nelson

    It's not impossible that Jones could turn it around? It's actually very likely that he does turn it around. It's 50 PAs. Those 50 do not predict what will happen the next 50. While those 50 stink, the Yankees have to project what they think he'll do going forward and not worried about water under the bridge. Jones has hit pretty well all but one season of his career. There is a really good chance he turns it around. You just seem to be patting yourself on the back for predicting Jones would struggle.

    The Yankees don't have to send Jones packing to bring up Maxwell. They've had a spare roster spot since Chavez got hurt and could have opened up a 40 man spot for Maxwell if they felt like it. He could be on the team, instead they went with Dickerson.

    • BrienJackson

      "It's actually very likely that he does turn it around."

      Jones hasn't been much more than a replacement level player since 2007, his current BABIP is actuall higher than it's been in the past few years, and he's striking out in over a third of his at bats. So no, in actuality, there's absolutely no reason whatsoever to assume Jones must have a rebound in him.

      • Ted Nelson

        The reason is that he doesn't have to rebound… you're looking at 5 games worth of PAs as if they mean something. They don't. It's 5 games. That's a little, tiny mini slump. Players will have 2-25 type slumps sometimes. If you're A-Rod or Cano you get 600+ PAs in a season and the 2-25 goes unnoticed. If it's to start a season–Gardner, Posada, etc.–or you only have 50 PAs… it's really painfully obvious because it's a huge weight on your season average to that point. Luckily the Yankees don't make rash decisions based on 25 PAs, or A-Rod also would have been DFA'd during his recent struggles and Gardner would be in Scranton after the way he started the season.

        In terms of BABIP… look at his splits. April it was .357, May is has been .143. .143 is not in-line with anything. These are little sample sizes, though. Performance varies from month to month, and you don't cut every guy who has a bad month. If you have some real incite into his stats or swing… ok. Just looking at 50 PAs and deciding a guy whose hit for years can't his is something you can't do, though.

        Jones have been above replacement every season besides 2008. I don't have a problem with DFAing him for Maxwell or whoever, but I want that to be based on more than a cursory look at the results from 27 PAs in May and a pre-season prediction he was washed up.

        • BrienJackson

          I'm not even sure what you're trying to argue at this point. Yes, it's a pretty small sample size, but such is the nature of things when you're dealing with an expendable role player. Jones isn't going to get 600 plate appearances, he's going to get 200-300, so 50 plate appearances is roughly 20% of his season. It's not fair, but such is life. Players like Jones are a short term option anyway, so they've got to perform in the short-term. Especially when they've got a 6-1 K/BB ratio and show no signs of being particularly unlucky.

          The Yankees won't dump him just for the sake of dumping him, but if they can get a better option or if he becomes pure dead weight he's gone. Again, that's the whole point of filling your bench with players like Jones in the first place.

          • Ted Nelson

            I'll break it down. I shouldn't have assumed you understood statistics.

            A sample can be used to represent an underlying population. The assumption is that with enough observations, the sample is representative of the entire population. Thus, you don't have to ask every voter how they're going to vote to take a poll. You ask enough people and you get a representative sample. There is a margin of error, but you can be pretty confident that's how the population feels about the election. Likewise, a sample of baseball stats–a season, 1/2 a season, etc.–is thought to be representative of the player's underlying ability and long-term performance once it reaches a large enough size. Performance will vary from year to year, but after 600 PAs it usually comes pretty close to a player's long-run ability. Within a given confidence interval (your margin for error) anyway. With the occasional outlier.

            Too small a sample, however, is not representative of the underlying population. There is too much variation. If I walk down one block and poll just 27 voters… I will not have an accurate representation of the entire population. These people all live on the exact same street, for one thing, so it is not representative of the population. In Jones' case, all these PAs come in the same month… which is called a "slump." More importantly, each result has a huge weight on my outcome. Perhaps one family on that block will all vote for the mother of the house, and there are four registered voters in the house. My poll concludes that 15% of the population will vote for Jane Smith on 123 Main Street… when really only 4 people in the whole population will vote for her.

            This is why the results in too small a sample mean absolutely nothing… too variable, too unpredictable. They have no predictive value. I can not look at Jones/A-Rod/anyone's last 5 games to conclude with any certainty how they'll hit in the next 5 games. It might represent the population, it might not. You might poll 27 people and get a breakdown that matches almost perfectly to the whole population. Totally possible. However, the chances are a lot higher that you'll get the right results with 27,000 people than with 27. Likewise, with 200 PAs you have a better idea of A-Rod's ability than with 27 May PAs from Jones in which he's been awful. This is also the case with Brett Gardner after his start, or Posada. Everyone was sure Posada would never hit well again based on a small sample, and he's proceeded to go on a tear.

            You cite a 6/1 k/bb ratio… but in this small a sample it's almost meaningless. No predictive value. His next 50 PAs could maintain that ratio, sure. But it's equally likely that it gets better or worse. There is not a critical mass of PA at this point.

            That Jones gets sporadic PAs does not mean he's less likely to reach his equilibrium performance (regress to the mean). It just means that it will take longer because he gets fewer PAs over the same period of time.

            20% of this season doesn't mean the next 80% will follow suite. It has nothing to do with fairness. It has to do with predicting future performance with as much accuracy as possible. I would honestly take a course or at least buy a book introducing you to basic statistical principles if you want to predict performance in baseball. I would actually take several courses and at least get to basic econometrics, but an intro will help a lot if you don't understand why small samples are more variable and less likely to represent the underlying population.

    • BrienJackson

      Also, the Yankees do not have an open spot on the 25 man roster, and Maxwell is already on the 40-man.

      • Ted Nelson

        Dickerson is the open roster spot… my point is that if they wanted to look at Maxwell they don't have to cut Jones, they could have just called him up rather than Dickerson. That Maxwell is on the 40-man–which I didn't realize–only makes my point stronger.

        • BrienJackson

          But there's no reason to have Maxwell and Jones on the roster at the same time. If Maxwell comes up, or the Yankees trade for another right-handed corner outfielder to hit lefties, then there's Jones serves no purpose.

          • Ted Nelson

            The purpose would be to see what Maxwell can do… give him an audition without losing Jones. The roster spot is open only temporarily. If Maxwell is a better player than Dickerson, just give him the PAs against RHP and ignore the L-R convention. You don't have to platoon good hitters. Then when Chavez returns you can decide between Jones and Maxwell.

            There's no reason to do it, because the odds are stacked in Jones' favor as far as who will hit MLB pitching better from here till the end of 2011. Again, Jones has had a 5 game slump and you're making a mountain out of a mole-hill. Maxwell has a k% of 42.3 against AAA pitching… how exactly do you think he'll do against MLB pitching?

Comments are closed.