More PMR defense goodness
DMZ · November 21, 2006 at 1:03 pm · Filed Under General baseball, Mariners
David Pinto’s been putting out the 2006 defensive ratings in bits and pieces, so let’s catch up on the first three.
LF: Raul Ibanez is horrible. He’s above Carlos Lee, Chris Duncan, Bobby Kielty, and Manny Ramirez.
CF: Ichiro is ridiculously good. Reed was the worst in the league in his limited time.
1B: Sexson’s pretty bad.
Update: Dave’s comment makes me want to clarify this — he’s right, and I should have said “Raul Ibanez’s rating is horrible” or something similarly clear.
[duhr]
[retraction of dumb comment]
Anybody else suprised by how awful Kielty is? I didn’t think he was good, but ManRam and Chris Duncan bad? That stuns me a little.
Wow – Barry Bonds is a better fielder than Ibanez? I was also surprised to see Scott Podsednik so low in the rankings. I would figure a speedster (He stole 40 bases in 2006) would cover a lot of ground, and get to a fair number of fly balls and line drives.
Is it that he lacks a good first step?
You fools! Now Dave’s mystery Reed trade will never go through!
[thanks given for deletion of said dumb comment]
Does anybody know where I can find a detailed description of those defensive models?
Obviously the system still needs a little tweaking, you can’t convince me that Delgado is a better fielder than Helton.
Nor can I be convinced that Bonds is better than Carl Crawford. Huh?
I’m suprised that Kotsay finished so low. I’ve always thought of him as a great fielder.
PMR is kind of like ERA for fielders – you wouldn’t (hopefully) look at ERA and say “Jake Woods is better than Felix!”, so try not to look at PMR and say “Carlos Delgado is better than Todd Helton!” It’s a metric based on what actually occurred, and doesn’t completely adjust for things out of the realm of a player’s abilities.
PMR is useful, especially when combined when viewed in combination with other stats like UZR, Dial’s ZR, and Dewan’s +/-, but it has some issues that hurt its individual reliability. Most notably, there isn’t an adjustment made for popups, so the PMR first base ratings could be easily skewed if one first baseman called off the rest of his infielders a lot while another deferred to others.
ZR and UZR don’t have this issue, so I’m guessing that accounts for a large majority of the differences among infielders. It almost certainly accounts for the differences on guys like Pujols. There’s just no way Pujols was really 31 outs better than expected. No way.
What is all those stats Dave and what do they mean?
Man, the n values seem really low to be able to state definitively that anyone but the extremes are good/bad/indifferent. Even Sexson missed 6 fewer plays, or 2%, than expected. That level of variation seems quite close to random, though perhaps the ‘expected’ number deals with this in some way.
What I’d love to see is a side by side with the newer UZR numbers; it looks like Chris Dial and David Pinto are both sure that Ben Broussard is an outlier, only the former thinks he’s Giambian and the latter thinks he’s quite good. I don’t get it. I’d love to see what UZR pegged Coco Crisp at too – it loved him as a LF; what does it say now? It seemed like UZR was something of an outlier last year on Crisp; should we now take his solid ratings (and that’s putting it mildly) more seriously?
So a lot of the ‘feeling’ that Reed’s D was somehow ‘plus’ defense is incorrect? Taken with his bat…are we looking at a total failure or would his historical data indicate it was an outlier of a year for him? It seems to be adding up that poor Jeremy just can’t function outside SoCal at even replacement level.
There arent just a few players whose rankings seem out of whack here. Torii hunter and Mike Cameron are right about average? Sorry, not buying it. Coco Crisp is one of the best in the league? In the august series at seattle he took worse routes than I ever saw from Randy Winn in center. Ben Broussard is no stud at first base, and konerko is nowhere close to average. The system seems to, in the most general sense, rate really good players as good, and really bad players as bad, but thats about it. There is too much noise in the middle, and too many wierd aberrations (hunter, cameron.)
I do not know much about PMR, or if it even purports to be a way of ranking a players total defensive value. It may be useful in measuring what a player accomplished in a given year, but it certainly does not look at all useful in comparing how effective a player actually is.
What is all those stats Dave and what do they mean?
I did a whole post on defensive stats last year.
Man, the n values seem really low to be able to state definitively that anyone but the extremes are good/bad/indifferent. Even Sexson missed 6 fewer plays, or 2%, than expected. That level of variation seems quite close to random, though perhaps the ‘expected’ number deals with this in some way.
Yea, the first base rankings need to be taken with more grains of salt than any others. It doesn’t even attempt to incorporate things like scooping ability, which clearly has an effect and is also a skill, just one we don’t know how to measure very wellyet.
What I’d love to see is a side by side with the newer UZR numbers; it looks like Chris Dial and David Pinto are both sure that Ben Broussard is an outlier, only the former thinks he’s Giambian and the latter thinks he’s quite good. I don’t get it. I’d love to see what UZR pegged Coco Crisp at too – it loved him as a LF; what does it say now? It seemed like UZR was something of an outlier last year on Crisp; should we now take his solid ratings (and that’s putting it mildly) more seriously?
MGL’s posted a lot of individual UZR numbers for 2006 on The Book blog and over at Baseball Primer. If you google for them, you can find a handful.
So a lot of the ‘feeling’ that Reed’s D was somehow ‘plus’ defense is incorrect?
If you’ve been hanging around here for long, you know that I’ve been saying that Reed’s defense was significantly overrated for a couple of years now. There’s no way he’s the worst defensive CF in baseball, but but he’s not particularly good out there. He’s a corner outfielder who is stretched to the max playing center.
Torii hunter and Mike Cameron are right about average?
Torii Hunter played hurt most of the year, and he’s always been a bit overrated defensively. I’m actually surprised his rating was that high.
And I love Mike Cameron, but he is 33-years-old. Defensive ability generally peaks about age 24. There’s no reason to think Cameron is still as good as he was in Seattle. Average might be low, but it’s very unlikely he’s still the best CF in baseball.
Coco Crisp is one of the best in the league?
Left/Center at Fenway presents all kinds of problems for defensive stats. Take anyone playing those positions with many grains of salt.
Ben Broussard is no stud at first base, and konerko is nowhere close to average.
As noted in another comment, take the first base numbers here with some grains of salt too.
Most notably, there isn’t an adjustment made for popups, so the PMR first base ratings could be easily skewed if one first baseman called off the rest of his infielders a lot while another deferred to others.
This is why PMR said that Orlando Hudson was worth +60 runs on defense in 2004. He was a ball hog on pop-ups (and actually stole some outs from the outfielders), so he looked like an MVP candidate.
Low scores for Vernon Wells, too. I’ve never thought he looked particularly good in CF (he looks slow when he runs), but he does have a strong accurate arm (see the 2006 All-Star game), so maybe that’s why he won the gold glove.
It’s not a UZR v. PMR comparison, but it *is* interesting: It’s PMR vs. ZR.
It looks like Ibanez is one of those players that really divides the metrics, with PMR seeing him as terrible and ZR having him as average to a bit above.
One more link; great discussion at tangotiger’s blog in the comments here.
Betancourt comes in at fourth for shortstops after Adam Everett, Bill Hall and Craig Counsell(!)
How does Sean Casey have a job, much less a $4M contract? I know you said to take 1B numbers with a grain of salt, but he’s a replacement level hitter whose power seems to be permanently gone (52 2B and only 17 HR in his last 1000 PA) and he’s a bad fielder. When I saw the contract yesterday and the negative offensive value, I assumed he was a defensive wiz.
I know Shelton tanked after a hot April, but doesn’t he have to be better not to mention $3.5M cheaper?
Casey still has a rep as a defensive wiz, and he’s beloved in every clubhouse he’s ever been in. He’s getting paid for who he is as a person, not who he is as a player.
but, Dave, he’s The Nicest Guy In Baseball (TM)
Adam Dunn is below average but while the ’06 UZR suggests he’s a -20 run guy, PMR rates hims as much less damaging… The PMR suggests Dunn and Ibanez are pretty much equivalent defenders……
[i]Casey still has a rep as a defensive wiz[/i]
This is the first I’ve heard of it.
Betancourt came out near the top. First PBP metric to do so.
Betancourt is sweet. I feel sometimes like Gammons when he said that he didn’t care about RF if RF said Jeter wasn’t the best player in baseball, but having watched Yuni this long and noted how far he moves to make sure-handed plays on balls, I have to admit that if someone came up with a defensive metric that said he sucked, I’d be extremely resistant to their pleas.