THE BOOK--Playing The Percentages In Baseball

Besides that, it is a SAMPLE statistic and even the perfect statistic, dropped onto us like manna from heaven (I am practicing for Passover), is going to be “wrong” X percentage of the time, regardless of the sample size.  People who don’t understand that concept have no business commenting on how good a stat is…

Tango, you are kidding, right?  You know you butchered his name, right?

And that is why we have nicknames like Tex, Grudz, Douggie Eyechart, etc…

At #1:

People are taught from a young age that 4=4.  With statistics we say that 4 might equal 4.  It also might equal 3.  The problem is that most people are never taught that.  They make it through Algebra and perhaps even Calc where 4=4 but never catch enough statistics to critically evaluate things like UZR.  So in their lack of understanding, they grasp onto the first tangible means to understand or dismiss the statistic, in this case Tex’s seemingly incongruous UZR.

The old McDonald way works, but only if you get the placement right:

TeiXeiRa

Name: ouch.  I can’t believe I butchered his name in a post about the perfect way to remember his name.

It’s fixed, but you can see the butchered name in the URL.

I use “‘i’ before ‘e’ except in ‘Teixeira’” to remember.

One Red Sox fan said that Tex plays much closer to the bag than Youk, and that he goes to the bag, leaving the attempt for the play to Cano, much more than Youk. If that is true, Tex can look great, and be up there with the best in his zone, but have bad UZR because his zone is smaller than average 1B zone (due to his style of play or due to manager’s preferences). Can anyone who has watched Yankees and other teams with great defensive 1B comment on that?

Davor,

That’s interesting.  As a Twins fan, I always remember “Eye Chart” diving to stop balls that the second basemen easily would have fielded.  This was way before I had heard of sabermetrics.

Is there a way to correct for this?  Or is it all ready corrected for?

A player gets credit for [1 - (% of times that a ball in that bucket is recorded as an out)] outs each time an out is recorded, right?  If so, then this problem is all ready accounted for for the most part.  If the second basemen would have easily made the play, then there’s probably a close to 100% chance that a ball in that bucket. is turned into an out, so the fielder doesn’t get much credit.

Davor, I don’t think that is going to make much difference, to be honest.  For one thing, in UZR, no one gets penalized when someone else makes a play, so it minimizes the effect of ball hogging, which is what you are talking about. Plus, I don’t think that Cano has any unusually good UZR numbers (I’m not sure without looking).

Anyway, if anyone is worried about that sort of stuff, than simply take Cano’s and Tex’s UZR combined and split it up any way you want.

Now, if Tex is looking good because he dives for a lot of balls to his right that other 1B do not have to dive for, that’s too bad for him - it doesn’t make him a better fielder than what UZR thinks.  And of course if he plays closer to the line, what he loses on balls in the hole, he will make up for to some extent on balls down the line and get extra credit for that.  If Cano gets to more balls because of where TEX plays, as compared to the average 1B, then it is true that Tex would be shortchanged and that Cano would be unfairly helped (in UZR), but, as I said, I doubt that is much more than a couple of runs a year.  Of course none of that changes how the entire Yankee defense (Cano AND TEX) looks to UZR and how they are in reality.

Just read the article.  It is not bad, but it is the usual B.S.

To wit:

“That’s Teixeira’s real beef with UZR. He’s so good, he makes difficult plays look routine — and no software program could interpret that.”

Of course if the author bothered to call someone like Tango or myself, he would find out that whether a fielder makes a tough play look easy or not has nothing to do with UZR whatsoever and it won’t make a good fielder look bad.  Quite the opposite of course.

But, sports journalists don’t do what real journalists have to do, which is to insure the integrity of their stories by doing the proper work and research, which is why sports journalism is the Professional Wrestling of the journalistic world, and an embarrassment to real journalists.

And of course there is the usual BS (in this article) about how sabermetrics does not acknowledge how players like Jeter are so great in the clutch.  I love how they use him as the poster boy for “clutch.” Does he even have good clutch numbers?  I realize there are many ways to define clutch and we can probably find one or two that make him look very good in the clutch.

The 1B’man’s skill in getting to balls thrown to 1B, and providing a nice target and having good reach is not really addressed by the fielding stats like UZR. 

IMHO this skill is undervalued, which is probably why it is not measured. I have seen some not very convincing studies that attempt to show there is little difference between 1B’man in this regard, but ask Edgar Renteria if he preferred Kevin Millar at 1B or Olerud, and ask Jeter if he felt a bit more comfortable with Tex at 1B than Giambi.

If Tex hugs the line, he is preventing doubles at the expense of giving up singles, even if it is at the expense of his UZR.  Maybe that’s his or his teams positioning strategy.

Also, UZR and Plus/Minus seem to be cumulative stats and not sample stats as claimed in a earlier comment, at least as far as I understand the definitions.  You consider every ball hit into his zone, and do not sample the data via exclusion.  You count the number of plays made or not made above average. 

In this regard, it is just like an offensive stat, due to SSS, a batter may hit well under or over his true talent.  Fielding stats have a SSS over 1 season, but just like the offensive stats, it shows what he did, even if this does not reflect on his talent. 

A poor UZR over a season means a fielder did not field that well relative to others, perhaps due to positioning or bad luck, just like a 300 batter who hits 200 over a 2 month period, perhaps a slump or a low BABIP, or both.

Unlike hitting (it’s either ruled a hit or not), there has to be some uncertainty by those who judge where a ball is hit in the zone and how hard it is hit.  I read somewhere that interns who change from year to year at each park make these calls.

IMHO this skill is undervalued, which is probably why it is not measured. I have seen some not very convincing studies that attempt to show there is little difference between 1B’man in this regard, but ask Edgar Renteria if he preferred Kevin Millar at 1B or Olerud, and ask Jeter if he felt a bit more comfortable with Tex at 1B than Giambi.

You must not have seen this:

http://www.fangraphs.com/blogs/index.php/first-basemen-scoops/

Also, UZR and Plus/Minus seem to be cumulative stats and not sample stats as claimed in a earlier comment, at least as far as I understand the definitions.  You consider every ball hit into his zone, and do not sample the data via exclusion.  You count the number of plays made or not made above average.

The two are not mutually exclusive.  When MGL say’s “sample”, he is referring to the fact that a season of a player’s career is simply a sample of his true ability, as you correctly noted below. 

A poor UZR over a season means a fielder did not field that well relative to others, perhaps due to positioning or bad luck, just like a 300 batter who hits 200 over a 2 month period, perhaps a slump or a low BABIP, or both.

That’s basically right.  A player’s UZR score is a function of 3 things:

1) His true talent fielding skill
2) His fluctuation around that
3) Measurement error

The problem that everyone seems to have with UZR and Texeira is that they believe that his 2009 UZR only measures #1 above.  And that’s of course incorrect.  And when people are misinformed about something, yet somehow feel the need to take a strong argument on it, you end up with the pile of BS that is this post.

"And when people are misinformed about something, yet somehow feel the need to take a strong argument on it, you end up with the pile of BS that is this post.”

Ooh, that is a nice line!  I hope you don’t mind if I borrow that from time to time.

(And I think he means the article and not pft’s post above.)

MGL,
I haven’t watched enough games to be able to comment, but I understood that Red Sox fan that Youk plays both deeper and more towards 2B than Tex, and also that Tex is more likely to break for the bag instead the ball or borderline plays. If he really plays too shallow, and he covers the bag on some plays that Cano can’t make, and he would have had a chance, he may well allow 5 - 10 singles a year that other fielders with his apparent skills wouldn’t have.
I would like to hear what other who have seen him and other good 1B play think - does Tex play closer to the bag than most?

I agree, a beautiful line that can be used on many occasions:

“And when people are misinformed about something, yet somehow feel the need to take a strong argument on it, you end up with a pile of BS.”

It is kind of funny that when a great hitter has an off year and hits .275 when he was expected to hit .315, no one says “That batting average stat is such crap, he’s a way better hitter than .275, I can’t believe you think (insert favorite player here) is a .275 hitter, you idiot!”