August 16, 2010

ESPN: Umps Miss 1-in-5 Close Calls: After reviewing every umpire's call in the 184 Major League games from June 29 to July 11, except for balls and strikes, ESPN's Outside the Lines has determined that 1.3 calls per game were close enough to require replay review and the umps were wrong 20.4 percent of the time. "Fans need to feel good about the fact that we're going to try to get the call right," said Hall of Famer Tony Gwynn, one of many baseball luminaries contacted on the subject.

posted by rcade to baseball at 05:24 PM - 18 comments

Not surprising at all. If you added balls and strikes it would look even worse.

I have no idea what the answer is. I'm against anything more than minimal technology being used, and yet it's frustrating to see just how bad the umpiring is currently.

posted by justgary at 05:17 PM on August 16, 2010

rcade, the article says 1 in 5 plays, but you were in the neighbourhood.

posted by tommybiden at 05:47 PM on August 16, 2010

Nate Silver's response.

Phrased differently, and less sensationally, about one call is missed for every four games.

posted by cl at 06:07 PM on August 16, 2010

Batting over .650 is pretty good in baseball. I think this is really only an issue if the play has some significant meaning. For example, on a close play at first, the ump says safe when the runner is really out but the runner does not score anyway, then not signficant.

It would be interesting to see a study on the significance of the close calls. Also, whether missed calls were spread evenly among umpires or if there were some umpires who blew more calls? Could be useful in weeding them out.

posted by graymatters at 06:19 PM on August 16, 2010

Phrased differently, and less sensationally, about one call is missed for every four games.

But that's taking into consideration every call isn't it? What I'm reading is that there's about 1 close call a game, and there's a 20 percent chance they will miss that call.

Watering down the stats to include calls that could be made by someone from their couch seems meaningless when the study was to find out how many 'close' calls are missed.

Batting over .650 is pretty good in baseball.

But a fielding percentage of .650 sucks.

the ump says safe when the runner is really out but the runner does not score anyway, then not signficant.

I don't think that's true at all. Runner is called safe. Next batter goes to a full count before getting on base with a single. The next hitter works a walk after 10 pitches. With the bases loaded the pitcher gets the 3rd out on a 2-2 count.

The pitcher could easily throw 20 more pitches because of one bad call (and could easily throw more). If your ace is on the hill that puts your reliever in the game earlier than he would have been, and if your relief pitchers are your weakness, that's a problem.

And of course, losing an out could mean that A-Rod comes up to hit with the tying runner on base in the ninth instead of swisher.

It could be very significant in more ways than if the runner scores. And it could be much more significant than just that one runner scoring.

posted by justgary at 06:37 PM on August 16, 2010

When choosing to believe Nate Silver or ESPN, I'm going to go with Nate Silver and be right 99% of the time, with a 2% margin of error.

posted by grum@work at 07:34 PM on August 16, 2010

But that's taking into consideration every call isn't it? What I'm reading is that there's about 1 close call a game, and there's a 20 percent chance they will miss that call.

Watering down the stats to include calls that could be made by someone from their couch seems meaningless when the study was to find out how many 'close' calls are missed.

I disagree. They arrived at those statistics by counting up a total and dividing it out over the number of games. Removing the "not close" calls from the sample to increase the percentage is whatever the opposite of "watering down the stats" is--say, condensing the stats.

Ultimately, their conclusion is that there is on average one blown call in every 4 games of baseball that could perhaps be avoided through replay. I, for one, don't find that terribly objectionable.

While your rundown of the butterfly effect of a blown call at first base is certainly valid, there are similar permutations of a single play that would make replay complicated to enforce in the first place (someone detailed something like this in another thread a few weeks ago), and furthermore, I can also live with the scenario you described.

posted by bender at 07:34 PM on August 16, 2010

I disagree. They arrived at those statistics by counting up a total and dividing it out over the number of games.

We'll disagree. The high majority of calls are easy enough for an average little league umpire to call. The close calls are the reason mlb relies on better than little league umpires. Blowing 1 out 5 is going to be acceptable to some, and not to others, but I don't see how the fact that they call 100 percent of the easy plays correct means much. I'd say the umpires can be thankful there's only 1 close play a game.

And for what it's worth, I believe balls and strikes are a much bigger problem, and I'm against replay for anything other than home runs. But I think MLB would do well by having a heavier hand in getting rid of bad umpires.

posted by justgary at 08:49 PM on August 16, 2010

This site is phenomenal. I was going to link to 538 and found somebody did it within an hour of this being posted. Amazing.

posted by DudeDykstra at 09:20 PM on August 16, 2010

rcade, the article says 1 in 5 plays, but you were in the neighbourhood.

Crud. I mislead people in 1 of 4 links.

posted by rcade at 10:24 PM on August 16, 2010

That is phenomenally bad science. What a worthless study. Blowing one out of five "close" calls? What the fuck is a close call?

posted by WeedyMcSmokey at 11:40 PM on August 16, 2010

Ultimately, their conclusion is that there is on average one blown call in every 4 games of baseball that could perhaps be avoided through replay. I, for one, don't find that terribly objectionable.

OK, but that also means that on a typical day (not a Monday) during the season, there'll be 2 to 4 blown calls in Major League Baseball. Over the course of 6 months, that's quite a few blown calls.

posted by LionIndex at 01:12 AM on August 17, 2010

OK, but that also means that on a typical day (not a Monday) during the season, there'll be 2 to 4 blown calls in Major League Baseball. Over the course of 6 months, that's quite a few blown calls.

The question is, how accurate is that? I mean, in tomorrow night's [national/regional/local] round-up of the MLB action, will the announcers point out 2/3/4 blow calls? Have they done so the last few games this week?

I don't recall them pointing out that many mistakes over the weekend.

posted by grum@work at 01:56 AM on August 17, 2010

And for what it's worth, I believe balls and strikes are a much bigger problem, and I'm against replay for anything other than home runs. But I think MLB would do well by having a heavier hand in getting rid of bad umpires.

Fair enough.

posted by bender at 08:20 AM on August 17, 2010

For example, on a close play at first, the ump says safe when the runner is really out but the runner does not score anyway, then not signficant.

Ask Jim Joyce how significant that runner can be.

posted by Ricardo at 09:01 AM on August 17, 2010

What the fuck is a close call?

A call that requires replay to determine whether it was correct. It's not like there's a more rigorous definition ESPN could have used. Observing umpires isn't an exact science any more than umpiring itself is.

posted by rcade at 09:56 AM on August 17, 2010

So Umpires get 4 out 5 calls right that you need a replay to see? That's pretty good. Of course, they are closer.

If the argument is that MLB Umpiring is bad, I'm not sure this makes that case. If you look at it from an overall, it could mean that umpires are 99.08% right.

But that doesn't mean it's a scientifically sound study. It's arbitrary. They identified 230 calls over 13 days that they deemed to be "close". Is 13 days a good enough sample size? Is their definition of close measurable? We're talking about a bunch of ESPN guys sitting around a monitor, mind you.

I'm saying if there is a study out there that genuinely shows the effectiveness/ineffectiveness of MLB umpires - this ain't it.

posted by WeedyMcSmokey at 06:11 PM on August 17, 2010

A call that requires replay to determine whether it was correct.

I don't believe that helps very much. If I'm sitting on my couch and I need a replay to see if the ump was correct or not, that's one thing. But I'm not a highly trained umpire. And I'm not in perfect position 10 feet from the play to make the call. And I don't get the benefit of sound (the player hitting the bag, the ball hitting the glove) which umpires are taught to use.

Thinking about this after my first response, I tend to agree with Weedy. The study is too vague to be of much value. I have no idea how close these close plays are (I'm also not claiming to know how to improve the stats).

I'd also like to know if the 1 out of 5 close calls missed are mainly due to a small number of umps or if it's pretty evenly spread out. If it's the latter, then perhaps missing 1 out of 5 is simply what a human umpire will miss, no matter how well trained. If it's the former, then the number could be even better if the worst umps are replaced with better ones.

posted by justgary at 12:47 PM on August 19, 2010

You're not logged in. Please log in or register.