Brilliant Reader Jon sent in this story by my friend Pat Reusse called “You can take your UZR and …”
The story is actually quite a fun read. I mean, yes, there are a couple of shots at statistics and the people who love them. So what? My friend Dave Krieger in Denver wrote about me being abducted to Planet Bill James (the cable channels up there are INCREDIBLE). I sometimes wonder why anybody in this crazy business of sports would take him or herself too seriously.
Anyway the core of Pat Reusse’s article is an interview with Jim Fregosi, and it’s clear that Fregosi has a pretty good sense of the stats. And he makes some excellent points about the right and wrong way to use them, like this quote:
“If I’m looking for a leadoff hitter, I don’t care about a combination of numbers — don’t care about his slugging percentage. I want to know how often he gets on base and if he can run.”
If the manager for my big league baseball team ever said something that sensible, I would jump up and down and buy a new team cap to wear around town. I’m not kidding. I’m so used to managers saying that they want leadoff hitters who can “handle the bat” or guys who “will fight up there” or some other vague trait that doesn’t mean anything. A few weeks ago, Royals manager Ned Yost was explaining to Bob Dutton why he plans to hit Mike Aviles in the leadoff spot.
“He’s a guy who has a chance to hit .300. He’s a guy who can steal bases. He’s a guy who will give you a decent at-bat, and he finds a way to get on.”
Mike Aviles’ career on-base percentage is .327. That’s below league average. He has walked 42 times in more than 1,000 big league plate appearances. He didn’t walk much more than that in the minors. He has been hit with pitches three times. I think Mike Aviles is a very useful player, and he might even be the Royals best option as a leadoff hitter. But he most definitely DOES NOT find ways to get on. And when you say stuff like that you make it pretty clear what your priorities are — and on-base percentage is not the priority.
So, I like Fregosi’s summation of a leadoff hitter: “I want to know how often he gets on base and if he can run.” That order. I like it. And he also said: “There are too many things that can’t be seen through statistics.” I agree entirely with that too. He talked about how OPS is not especially relevant when he’s looking for a utility infielder. Agree with that too. Almost everything Fregosi said in the article, I agree with.
Fregosi then said that he despises UZR — Ultimate Zone Rating, one of the advanced defensive metrics out there. “I can watch any player for three days and tell you if he has range,” Fregosi said. “And I’ll tell you more accurately than a chart in a computer.”
Jim Fregosi absolutely may be right here too. There are real issues with Ultimate Zone Rating. But I will say one thing … we only ever hear from people like Fregosi on this topic. That is to say we are constantly hearing from baseball people who know how to measure defense better than some statistic like UZR. We are constantly hearing from people who, through well-honed powers of observation and years of visual training, can determine a player’s range and skill and defensive production better than UZR. They don’t need any statistic to tell you who can or cannot play defense. As the headline says, you can take your UZR and …
And you know who we never hear from on this topic? That’s right: UZR herself.
Q: Thank you so much for joining us. I understand this is your first interview.
UZR: Well, I’ve been jumping from mother’s basement to mother’s basement, and I really have not had any free time.
Q: I’m glad you …
UZR: Oh the underwear I’ve seen.
Q: Right. So let’s get right down to it. You are a defensive statistic and one of the more prominent ones out there. First, can you explain yourself.
UZR: What I try to do is determine how many runs a fielder saves his team over the average player. I look at the players arm, his range, his errors and the number of double plays he helps turn.
Q: So you are trying to ruin the game?
UZR: Is this going to be one of those interviews?
Q: No. I just wanted to see how you would react. By the way, you look good, have you been dieting?
UZR: Well, I’m park adjusted.
Q: I thought so. So, you’ve heard what people say about how imprecise you are … what do you say those people.
UZR: I’d agree with that wholeheartedly. I’m a defensive statistic. There are all sorts of quirks and blips involved with measuring defense. I always tell people never to use just a year of me. You have to use at least three years to get real value out of my numbers.
Q: Three years?
UZR: Oh yes. AT LEAST three years. Defense is a complicated matter you know. If I may, I heard what Jim Fregosi said about me, and let me say that I’m a big fan of his. We met once, years ago, but he wouldn’t remember … anyway, I don’t doubt that he could watch a player for three days and determine if the player has range. I bet he could watch a player for two days or one day and say if he has range. I could never do that.
But that’s not what I’m trying to do. I’m trying to tell you how much a defensive player is helping his team. And I’m doing it with numbers. The managers look at Derek Jeter and tell you he’s a great defensive player, right? They give him the Gold Glove every year. Well, he might be a great defensive player by the eyes. He might be the smartest player in the world. He might have a sense that no other player on earth has. But I rank him 42 runs below average since 2002 because that’s how the numbers add up. I don’t judge based on how good or bad he looks. I mean, he looks good, he’s Derek Jeter, come on. I don’t care if he dives or doesn’t dive. I don’t care about any of that stuff. Does he make the play? Does he get to the ball? Does he turn two? These are the data points in my statistic.
Q: So you’re saying …
UZR: I’m saying that the human mind is better for writing poetry. The closest thing I’ve ever come to poetry is this: “Hat … Pat … Sat.” I’m still thinking a name for it. The human mind is better for literature, for music, for art, for comedy. The human mind is better in billions of different ways that I could never conceive. The human mind is especially better at narrative.
But by being better at narrative, the human mind can and will shift things to make them fit. The human mind will find trends in randomness, and stories in fog, and that’s one of the beautiful parts. I can count better than you can. I don’t mean that in a bragging way. I just can. I can count better, and I can ignore unnecessary data better, and I cannot be influenced by beauty or awkwardness. If you have one day to determine if a guy can play defense, or a week, or a month, you are better off to use your eyes because I need more than three days. If we have five years of data, I’m pretty sure I’ll beat your analysis every time.
Q: Do you understand why people take shots at you?
UZR: Sure I do. I sometimes spit out numbers that don’t match up to what the eyes suggest. It’s not personal with me. But it is personal with baseball fans, and it should be. They are watching the game with love. And they don’t want to be told that their eyes are misleading them, that they might not see the game as well as they think, that their hero doesn’t get to nearly as many balls as they believe.
I tell people that I’m a tool. See, a person would never say call herself a tool. Heh heh. That’s statistical humor. It breaks them up at the conventions. Point is, I’m not perfect. I’m not close to perfect. I’ve gotten better since I was young, and I will keep getting better. Baseball defense has been widely miscalculated for many, many decades. People have judged players on whether or not they fielded ground balls that hit their gloves, or made throws that sailed over their targets. People have judged players on how far they ran to catch a ball or how spectacular their dive. That’s all wonderful to watch, but that’s not what defense is about. Defense is about preventing runs. Defense is about turning batted balls into outs. Over seasons, I believe I can tell you which players are good and not so good at doing that. And I believe I can do it better than you can do it with good eyes and a great memory.
Q: OK, before you go, tell us … which of the advanced statistic is the best.
UZR: Oh, I could never choose, they’re all my friends. But I can tell you that VORP is hilarious — I told you stats are not very funny, but VORP is the exception. He told this one the other day … let’s see if I can remember it: How many RBIs does it take to change a light-bulb? Who cares — RBIs are context-based stats that overstate the importance of certain middle-of-the-lineup hitters! Oh, he had us rolling in the aisles with that one. Also WAR does a great Dave Krieger impression.
Don’t invite WAR to a party. He will drink all your juice, and leave early in the morning.
OT, but the great Charlie Pierce gives a shout-out to Joe (http://www.boston.com/sports/columnists/pierce/2011/03/play_ball.html). Talk about an ideal guest for the PosCast…
Your comments about how the human mind constructs detail (often falsely) is related to a book I’m reading: “Stumbling on Happiness” by Daniel Gilbert. If you want a respected psychologist to back you up on that point, give it a read.
Excellent Joe, but I’m sorry to say that I have to take some issue with your first quote from Fregosi…
[quote]
If the manager for my big league baseball team ever said something that sensible, I would jump up and down and buy a new team cap to wear around town. I’m not kidding.[/quote]
But it isn’t particularly ‘sensible’ at all. It is again taking that non-sensical view that “leadoff hitters” have to have a certain “job” because they “lead off” while in the real world, the leadoff hitter is only guaranteed to leadoff one inning, and for the rest of the game he could have, who know, 5 bases-loaded appearances. So OF course you should care about SLG, even for “leadoff guys.” In fact, I would guess that a .370/.350 guy probably should not bat atop your lineup despite that sexy OBP. That’s just a guess, but to say we should flat-out ignore SLG because he’s a “leadoff guy” is not sensible at all in my opinion.
Greatly enjoyed the rest of your post.
Great read Joe – as usual.
I really like to read Reusse. He’s wonderful. But his UZR article left a bit to be desired.
You hit the nail on the head, or should I say that UZR did. I didn’t know UZR was female, but that is clear in the interview, even if you hadn’t mentioned it. . .
“I’m saying that the human mind is better for writing poetry. . .
. . . The human mind is especially better at narrative.”
and. . . “But it is personal with baseball fans, and it should be. They are watching the game with love. . .”
What I think that Reusse and Fregosi fail to realize, is that for many of us – (some, like me really don’t understand UZR any better than they do. I digress.) – the new stats HELP in our enjoyment of the narrative of the game. They ADD to the poetry of baseball that is what feeds most all of our enjoyment of the greatest of games.
That interview is a real coup! UZR!
Yost is following Aviles with Melky Cabrera in #2 spot. His lifetime OBP is .328. I think the second spot is Yost’s “Handle the Bat” spot. He is keeping it warm for Jason Kendall, who had a .318 OBP last year with Zero triples or home runs. Now that’s handling a bat!
Don’t know which is more awesome: that your mind goes to places like this, or that you can share the adventure with us so beautifully.
In the same vein as @RJL: I agree that you shold care about your lead-off hitter’s OPS and Slugging. I think the only guy who should be judged pretty much on OBP first is a #8 hitter in the NL. Gotta clear that pitcher’s spot. . .
RJL: that’s not quite true. The #1 spot is usually batting behind the worst batters on the team. So even in later innings, he is more likely to be batting without runners on base than any other batting position.
So given that a typical (good) team will have at least one player with a good OBP but lousy SLG, you want that player hitting #1.
Matt has the right idea about why OBP is important for leadoff hitters. Additionally, while the leadoff hitter might not leadoff any more than once a game you still want a guy who gets on base at a high rate batting in front of your middle of the order bats. If the goal is to maximize the number of runs scored, then you definitely want a guy who gets on base 40% of the time batting in front of your 2, 3, and 4 hitters rather than a player who only reaches base 30%.
Can UZR team up with the RZA ?
Here in the Twin Cities area it’s always a treat to hear Reusse go back and forth on the radio with Phil Mackey, who at 25 years old has probably aged another 10 years by trying to explain the concept of Wins Above Replacement. I’m on the stats side of the divide, (as I suspect are most who frequent Joe’s blog) but entertainment is entertainment. Of course, no one should call Reusse a curmudgeon, because Sid Hartman is still on the loose!
@RJL and Stephen: To supplement what Matt said…the most important aspect to scoring runs is getting on base. As your #1 spot will come to the plate more than any other slot, it is even more important that he reaches base. Tango and MGL’s The Book goes in depth on how to properly construct lineups.
P “N” K – now live in Wyoming. Sid’s no curmudgeon. Too many syllables.
Reusse is though. In the best sense of the word.
I don’t get to listen to him, but I found his dismissal of WAR and such sort of disappointing, because he clearly loves baseball, and has a good feel for the game.
oops – signed in on my wife’s account
Underminer
P “N” K: Yeah, some “treat” it is to have Reusse blather on about, well, anything. Sid is merely keeping the seat warm until Pat finally falls over the edge. Should happen in a year or two.
Laura: Ha! If Curmudgeon was a Gopher, Sid would LOVE him!
Joe – get some sleep! You’re going loco!!
I have always suspected that the real value of UZR is that it finally allowed people to find a flaw with Derek Jeter. I say this because 99 percent of all articles about UZR (including UZR’s own explanation of itself) cite the fact that stats prove Jeter is really a horrible fielder. I wonder if UZR had shown he’s an awesome or even adequate fielder if people would be as interested in the stat. I will say that if UZR is accurate about Jeter, it shows that people have long been wrong about the value of needing a good-fielding shortstop to have a winning team.
RJL and Stephen should be aware that it is generally understood that a point of OBP in the leadoff position is worth three times as much as a point of SLG.
Cyril Morong did most of the original number-crunching and then some others ran with it.
http://www.beyondtheboxscore.com/story/2006/2/12/133645/296
http://thepastime.net/2006/02/lineup-optimization.html
The value of slugging never goes away, but yes, your highest OBP should hit # 1.
So Fregosi and Joe (and Aaron) DO have it right.
Joe, you’ve a committed a bad cut-and-paste error. The Dave Kreiger link should be http://www.denverpost.com/krieger/ci_17715462
rasternomicals: Question about those articles: They’re nearly five years old and are filled with preliminary data and a few theories. I wonder: has anything changed at all? I could’ve sworn I used to hear that it was proven that batting position didn’t matter much (and that baffled me).
I also never disagreed with Joe or Fregosi (and being a Mets fan, I am grateful that we seem to finally have guys in place who think like Joe and Jim). I just said that ignoring SLG would be dumb. And then I made some convoluted point about the 8th spot on an NL team (which is probably a manifestation of again, being a Mets fan; thank you Luis Castillo).
I enjoyed the UZR interview, although I prefer defensive runs saved. (There is more “seeing” put into it, as well as number crunching.)
I agree that OBP is the #1 requirement for a leadoff guy. I prefer he be faster than average but speed is secondary.
One of the biggest lineup mistakes made by managers is the habit of putting your best stolen base guy, even if he has a low OBP, in the leadoff spot, and then putting a weak hitter who can bunt in the #2 hole.
Giving up an out is bad enough. Doing it in front of your 3-4-5 hitters is criminal. (I saw a manager with really good 3-4-5 hitters last year who tried to bunt his fast guy over after a single-in the first inning. The lead runner was thrown out, and then the #2 guy was caught stealing. The #3 guy homered, now a solo shot. The team lost the game by a run.)
I hit those type of guys 7th and 8th. (It could be 6th and 7th if I have a weak hitting team.) I want my stolen base guy to be in front of my weakest hitters. I want them to see more fastballs. I want the threat of a bunt to change the defense, and at the bottom of the lineup, there might even be a situation where it is called for (although not in the first inning.) If my stolen base guy steals second, I want the opposing manager to be tempted to walk a weak hitter to keep the double play a possibility. Also those weaker hitters behind my stolen base guy are less likely to string hits together or get an extra base hit. The difference between having a runner on first and one on second is magnified.
Well of course UZR is a woman! Stat geeks are madly in love with advanced metrics. I believe that stat geeks are overwhelmingly male, and if they are comparable to the general population, perhaps 95% of them would be straight. UZR must therefore be female!
Joe, this post is particularly inspiring to me. I work as a conservation planner, and a lot of what I do involves developing spatial models (computer maps) of conservation priorities, things like rare species habitat, large intact natural areas, water resource priorities, etc. Our analysis is used to inform public policy decisions, but we constantly get criticism along the lines of “the data is missing this” or we’re just “turning the dials”. This essay, especially the section that starts with “the mind is better for writing poetry” is the most compelling argument against the critics that I’ve come across. I’m going to start quoting you to the skeptics!
In Cliff Lee on 32 best, you wrote (and the SI editor didn’t catch) “ferryman who took home across “. It should be him, not home.
“there are a couple of shots at statistics and the people who love them. So what?”
Poz, I think you’re letting Reusse off the hook in a major way here. Much of that article is as ignorant as anything penned by Murray Chass or Dan Shaugnessy, or the Bruce Jenkins paragraph you shredded a couple of weeks ago.
Not that you should tear into Reusse (articles like that are dime a dozen at this point, and carry almost no weight), but linking the article and calling it an “enjoyable read” is like telling your friends that “Winter’s Bone” was a lighthearted laugher because of a handful of clever lines.
All that said, you once again showed your knack for supporting stats with tact and humor, not animosity frustration.
RJL – I guess this means Aviles and Cabrera are the ideal 1-2 hitters for the Royals. Thanks for clearing that up.
By the way, you look good, have you been dieting?
UZR: Well, I’m park adjusted.
Hillarious Joe, loved it