The “value” of holds

Thursday morning, an Adam Hayes-penned an article appeared here at The Hardball Times regarding relievers and the shortcomings of the mainstream stats used to evaluate them.

Thursday evening, the Pirates lost their game against the Brewers to fall below .500 on the year as Pittsburgh continues to do a nifty imitation of last year’s collapse.

These two items are related because of the box score that game produced.

After climbing out of an early 4-0 hole to take a 7-4 lead, the Buccos coughed up their late lead and fell by a score of 9-7. One of the pitchers most responsible for this loss was Chad Qualls, who surrendered three runs on three hits while retiring a single batter.

Qualls was credited with a hold.

Chris Resop came in next and gave up a run on two hits and a walk while recording two outs.

Resop took the loss.

Obviously, neither hurler pitched well, but Qualls clearly was worse. It is absurd for him to receive positive credit for his “contribution” while Resop was on the hook for the loss.

Holds, saves, wins, losses, blown saves—these traditional counting stats we attribute to pitcher performances simply don’t do a sufficient job of assigning credit and blame. Yes, those with a sabermetric bent are well aware of this, so situations like this simply serve to provide more ammunition in the assault on these stats and the significance many fans—and mainstream media—attribute to them.

Print Friendly
 Share on Facebook0Tweet about this on Twitter0Share on Google+0Share on Reddit0Email this to someone
« Previous: And That Happened
Next: Ichiro and the road ahead »

Comments

  1. Paul G. said...

    WARNING: Rant.

    The Hold is not an important statistic, really.  Most of my encounters involve one of the following scenarios:

    1. John Barten mocking it in the THT Awards as only John Barten can mock.
    2. When I am staying in a hotel and get to amaze my co-workers with my explanation of “H,3” in the complimentary copy of USA Today.
    3. On the occasional TV broadcast where it gets mentioned by the geekiest member of the announcing team.  This is typically met with (a) the color commentator having no clue what a Hold is and having to have it explained to him, despite the fact that it was explained to him last week, or (b) the other guy treating the trivia with disdain as yet another one of those things that the eggheads love,  which he then uses to segue into an ode to whatever intangible he is infatuated with today.

    Honestly, the Hold is one of those stats without an audience.  Casual fans don’t care, statheads like us treat it with suspicion, and baseball people are probably oblivious.  This is the purview of emerging statboys who embrace it as a step up from calculating ERA and then will toss it aside as they learn about more advanced metrics, girls, or both.  And who can blame us?  It is an unofficial statistic that does not even have a consistent definition.  As Adam was good to point out, It does not even get the dignity of being blown.  (Does “Blown Hold” sound safe to say in mixed company?  Just askin’.)

    I think the main problem with the Hold is it doesn’t make sense.  The Save rules, while flawed, do make sense.  If the closer is granted a 3 run lead, promptly gives up a couple of homers and loads the bases, yet despite the slings and arrows of outrageous fortune and the hanging curve manages to get the elusive third out without further damage, he has still accomplished something.  He held the lead.  Chance of victory: 100%.  The fact that he was terrible is unfortunate, especially to his agent who will have to settle for the 30 foot yacht when this reduces his free agent desirability, but he did do his job.  This same scenario that results in an emptied dugout gathering in the infield for high-fives, bro hugs, and the occasional somersault as Frankie croons “New York, New York” into the canyons of Gotham (or whatever) doesn’t work with Joe Seventh Inning.  Poor Joe, assuming he managed to get the third out, has greatly harmed his team’s chances.  More likely he has made his way to the showers to wash off his failure and the random spilt beer of woe as his successor gets the joy of surrendering runs that belong to someone else for a change.  The Save rule wasn’t meant for open ended situations like this and it shows.

    So, anyway, Wins, Losses, and Saves are flawed stats, but they generally measure something useful.  What do Holds measure?  Surviving?  Avoidance of not being too terrible?  Being able to blame someone else for your failures?  Who needs a stat for that?  Well, besides Gregg Jefferies…

  2. Paul G. said...

    Now that I have stopped ranting, I still think the Hold could be a useful statistic.  Actually, right now it is a fairly useful stat because as much as we love to rip into the ridiculous situations like the one above, most Holds are legitimate.  But really the rules should be modified so they apply to success as a middle reliever as opposed to success as a hypothetical closer. 

    The simplest form would be “enter in a Save situation, get at least one out, don’t give up any runs, don’t get the Win or Save.”  You’ll still get a few weird situations where the honoree loaded the bases with one out (on a sac bunt) and had to be bailed out, but at the very least it measures something that is actual success as opposed to what might be success if the next reliever performs a miracle.

    Of course there is also the tricky situation of a middle man entering in a tied game or even down a run or two.  Holding an opponent where they are in a close game is one of the defining roles of a middle reliever.  The Hold ignores this completely in favor of something like looks like a group Save.

    So, yeah, it needs work.

  3. bpdelia said...

    @ above ranter:

    exactly wins and losses are goawed but by and large they do have meaning and these ststs need to be simple enough for the casual fan.  in reality the whole system needs a tweak.  Sorry boys but WPA isnt going to become a mainstream metric used to assign win shares.

    firstly they need to actually give credit for the win to the best pitching performande. bthe 5 ip starter requirement is fine.  but a 1-0 11th inning game in which a lefty specialist gets a win for one out while the sp pitched 9 shut out innings need to go.  Most wins are legit.  the other frsction needs to be assigned based on best pitcher.  you can use GAME WPA or gamescore to eecide who gets the win in non traditional outcojes.

    the uold:  the hold should only be given if a reliever keeps the lead and either A)records the final out of an inning.  or b)leaves the game without having put the tying run ON BASE.  if you come in with a two run lead, liad the bases and get pulled.  no hold whther a run scores or not.

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>