The 2004 defense post

DMZ · November 24, 2004 at 8:04 pm · Filed Under Mariners 

Updated! Now with more goodness! I’ve mentioned I’ve been working on this a couple of times before, and I’m going to go ahead and post instead of waiting for good defensive stats to come out.

The 2004 Mariners finished 2nd in the AL and 6th in the major leagues in defensive efficiency, which is how many balls in play were turned into outs. Now, park effects have something to do with this, but it remains true that the team played good defense. And yet, you’ll remember that I said some terrible, terrible things about Randy Winn to start the season, and early in the year I bitched constantly about what a terrible defense the team was putting on the field.

So what happened?

There were two things. Dave’s mentioned this repeatedly, but I’ll echo him once more: the 2003 team played spectacularly good defense. The outfield, in particular, was best-in-baseball good, by a long way. The dropoff from 2004 to even being a modest top-third defense was huge, and as fans we reacted to that. It was entirely understandable.

The other big thing that happened over last year, was that they got better. Apx Batting average on balls hit into play*, by month:
April .301
May .266
June .271
July .287
August .279
September .259
October .313

* as I couldn’t get complete split data, I used hits/balls in play, with balls in play as outs-k+h-hr

Now, while it doesn’t break down so clearly, you can see the team starts out terribly, gets much better, is okay, then really plays well in September.

A lot of this is Randy Winn. Winn started the year not just worse-than-Cameron, but flat bad. By the end of the year, he was good out there. As good as you can be with that arm, anyway. That overall, I think defensive metrics will show Winn had an above-average defensive season for the position is amazing, because he really did suck early on.

Ibanez was not as bad I expected in left. From Winn, he’s a huge step down, certainly, but I think he too was playing left far better at the end of the year than he was when he started out.

The changes in player personell helped: Jolbert Cabrera, ride him for his bat as I may, helped defensively as he got more playing time after the first month. Spiezio’s defense contributed up, and using him as a 3B/1B player worked well. Between these guys, the anticipated step down from Olerud didn’t happen — if anything, the team got better.

At short, I think Aurilia was better than Lopez, though we’ll see if the stats end up bearing that out, but the team didn’t miss Aurilia at all.

Boone. Ah. This is totally subjective, and there’s no way defensive statistics are ever going to be able to get this fine. But he seemed to be streaky defensively, even more so than you’d expect from a fired-up dude like Boone. Game-to-game, he was hauling baggage around, but when he looked good, he seemed to keep his head in things and not make the boneheaded blunders that were his undoing this year.

I got a little, uh, frustrated not having good numbers, so… it’s bad number time! Range factors and zone ratings are badly flawed ways to compare defense, unless you’re comparing teammates… which I am. Then they’re only sort of bad. Note particularly that in what I’m doing, playing time drops quickly, so the sample on the top guy is much better… anyway, check this out.

First base
Who, Fielding Percentage, Range Factor, Zone Rating
Oleud, .998, 8.38, .843
Spiezio, .986, 8.87, .830
Bucky, .984, 9.31, .778
Jolbert, 1.000, 9.50, .931

Bucky’s ZR looks bad, but other than that the surprise is that Cabrera does look pretty sweet out there, whils Spiezio’s indeed about Olerud’s level. Raul Ibanez, by the way, comes out awful. Like.. Frank Thomas bad.

Second base
Who, Fielding Percentage, Range Factor, Zone Rating
Boone, .978, 4.33, .790
Cabrera, .987, 4.99, .867

Okay, here’s the Boone-off, then
2003 Boone: .990, 4.54, .814
2002 Boone: .989, 4.36, .843

Big decline for Boone this year — consider he’s playing behind substantially the same staff this year as last, too, so that sticks.

Surprising is that Cabrera, caddying, embarassed Boone.

Third base (200+ IP)
Who, Fielding Percentage, Range Factor, Zone Rating
Spiezio, .964, 2.86, .771
Cabrera, .970, 3.11, .864
Leone, .901, 2.71, .773
WFB, .923, 2.44, .683

Hee hee hee. Bloomquist got outplayed by Justin Leone.

What’s also interesting is to see that everywhere there’s enough innings I’m comfortable even looking at the stats, Cabrera shines.

So, to the outfield…
Left field, Ibanez < < Winn, but we knew that. Center field, Winn, Bocachica and Reed all look about the same, in differnt ways -- weirdly, RF and ZR disagree badly, rating them RF: Reed (huge drop), Winn, Bocachica ZR: Bocachica (big drop), Reed, Winn Generally speaking, here's my caveat on these defensive stats: almost no defensive stat, bad or not, will hide a truly bad or stellar player. When you're working with range factor and zone rating, huge difference are worth paying attention, but like we see here in center field, where there are three guys who appear like they about the same for the position, it's a wash. At that point we want to go to the play-by-play data, which I don't have. Once again, if anyone wants to give USSM many, many millions of dollars in venture funding without the expectation of any return, we'll be happy to look into providing that data. So anyway, how's that for not being a rabid stathead, huh? It's also possible the defensive misadventures of Olivo forced the team to pay more attention with runners on... heh. I learned a couple of important things from all of this. First, defense is amazingly important. The difference between a 2003 OF and a 2004 version, just defensively, is like adding a star player. You can use defense to cover the weaknesses of your pitchers (turning Hasegawa, Franklin, say, into effective pitchers) beyond just preventing runs. And second, defense may be more pliable than offense. In April, I would have guessed that Winn would have finished the season maybe ten runs defensively below an average center fielder. He did much better than that. Which is weird, because I'm a sucker for position switches and think teams should be much more willing to swap guys around, if their defensive skills suit a switch. It would seem that there is some value in leaving a guy out there and having a coach yell at them as they shag balls. In any event, I'm going to be very interested to see what the 2004 play-by-play stats like UZR say about this team.


21 Responses to “The 2004 defense post”

  1. Mark on November 24th, 2004 9:07 pm

    Excellent comments. This is what I love to see from the stathead community: arguments that start from data and move to interpretation, being willing to change one’s opinion based on what the data is saying. Unfortunately this is what we’re seeing less and less of from the more successful of the stats community writers. But it’s still nice to see it here.

  2. Mark (another one) on November 24th, 2004 9:08 pm

    Interesting and nice to see some crow being eaten. Does Winn playing in LF a lot more in August/September contribute to his apparently un-suckiness?

  3. Trent on November 24th, 2004 10:07 pm

    It didn’t hurt him on the whole, but he was nowhere near as bad as we all thought he was. Like DMZ said, we were all really spoiled in 2003, so anyone not named Cammy would’ve looked like chopped liver.

  4. Matt Staples on November 24th, 2004 10:53 pm

    Nice comments as well; it confirms the subjective feelings I had about the team’s defense as the year went on. As for the importance of defense, we may want to — once again — look down to Oakland, or over at Boston, to see what the market inefficiency of the moment is: defense is, or at least was, undervalued.

  5. Matt Staples on November 24th, 2004 10:55 pm

    My allusion to Oakland, by the way, has to do with the move to bring in Kotsay, even at that relatively high salary, and to a certain extent, Miller.

  6. Terry Benish on November 25th, 2004 12:21 am

    Bang on about Aurelia/Lopez. Were Boone’s stats better or worse with Aurelia or Lopez. I’m betting Lopez as partner made him worse…2003 to 2004 there were far more doubles and triples. Almost 1 more per game on average…that’s got to be Winn and Ibanez.

  7. eponymous coward on November 25th, 2004 12:26 am

    Don’t forget, Reed spent some time in CF in September, too. He looked credible out there, in the Mark Kotsay-I’m-kinda-out-of-position-but-not-a-disaster sense. I don’t think there’d be a lot of dropoff defensively subbing him for Winn if any. Offensively is maybe another story- depends on if he can take a step forward from AAA to the majors more in line with his AA and A ball stats. I’d be willing to bet on it if gave us an extra 3 million to play with on free agents, but I could see why other people wouldn’t- he’d likely have to do very well on his PECOTA projections, like 80-90%.

  8. Paul Covert on November 25th, 2004 12:28 am

    Mark (re. #1): The quote concerning the “sabermetric ideal” that has stuck with me over the years was from Bill James’ Baseball Abstract 1983, in which he wrote that “the sportswriter attempts to be a good lawyer; the sabermetrician, a fair judge.” In other words: That the point of sabermetrics was not in arguing for positions, but in weighing evidence rationally.

    Of course, this ideal can never really be applied with total purity. In the same essay, James also added that “I do of course have positions for which I argue from time to time” (quote inexact, as I don’t have the book with me at the moment). Anytime you review evidence, make up your mind on an issue, and try to explain why your conclusion is correct, you have now become a partisan. Conversely, if for that reason you refuse to ever make up your mind on anything, you may succeed in avoiding error, but you’ll be of limited use in a world where people do have to make decisions that matter every day.

    The best solution, I think (if I may be pardoned for philosophizing here), is to be willing to be convinced, but not to be convinced lightly. I have commented to friends elsewhere that statistics is about “knowing when it’s safe to reach a conclusion from data.” In some cases the data really won’t be clear (as for example in the evaluation of managers); in other cases (as for example in the relative values of walks vs. stolen bases) the trend will be clear beyond a reasonable doubt, and may be ignored only to one’s own harm.

    I do think that the advantage of taking a “scientific” approach is that it does, indeed, allow us to learn from our mistakes much better than a gut-feeling-based approach does. When I met Derek for the first time at a BP pizza night a couple years back, we talked about my prospect studies, and I was mentioning that, based on admittedly sketchy data, it seemed that BA seemed to be coming out ahead of the stat-based guys in the long-term performance of their chosen top prospects. Derek cringed and said, “Oh, noooooo… have I been wasting my life?” But I replied: No, it’s natural for science to start out behind the curve compared to intuitive observation. But if we can learn from our mistakes *coughJeremyReed#2overallprospectcough*, we can steadily improve our methods, and there will remain good hope that we can eventually make a real contribution to advancing the state of public knowledge.

    As for the “stathead community”: I suspect that the Internet, though it’s helped us in publicity and communication, has harmed us in our standards of proof. When Pete Palmer and Bill James were the only guys in the market, publishing books once a year (James) or less (Palmer), they could afford to stick with their solid theories as the mainstays of their writing, and to treat their speculations as speculations. But with the Usenet in the ’90’s and now with blogging, their tends to be more of an immediate payout (in public attention) for making bold statements, and so the standards of proof tend to be lower than we might like them to be.

    When I wrote my guest article on prospects for BP a few years back, I was actually hoping it would be put through a peer-review process and that somebody would try to find the weak points and challenge me on them. It eventually got published, but with only editiorial-type changes from its original version. But I still think that a peer-reviewed sabermetric journal would be a good thing. People’s opinions of the day (which is what blogs are) can be fun for a nice little daily shot of mental energy; but a systematically developing body of knowledge would in the long run be, in my view, more helpful.

    As for the guys here at USSM: Part of why I’m here is because I’ve seen and appreciated how Derek and Dave have improved since the old days when I was a lurker at (Jason seemed to me not to have been quite so forcefully opinionated in the first place.) I still occasionally think they’ve rushed to judgment on something (perhaps they think the same of me? I don’t know), but for the most part, they’ve gotten to the level where, when they reach a conclusion, it’s usually one on which I can agree with them.

  9. JPWood on November 25th, 2004 2:18 am

    Besides what David, Derek and Jason write here, it’s readers like Paul Covert who make this interesting.
    I guess that means that pizza is a brain food.

  10. Shawn on November 25th, 2004 2:27 am

    #8 — A baseball research journal? I like the idea. If the transport properties of nanoscopic materials can make it in some wacky engineering journal, then baseball defensive metrics should have its own place too! Make it happen!

  11. DMZ on November 25th, 2004 3:08 am

    There is a baseball research organization, SABR, which publishes baseball research. Magazines and everything. Check it out. A lot of interesting work, and they have different committees that focus on cool stuff, so if you’re, say, fixated on dead-ball era history, there’s a group for you.

    I’m a member.

  12. vj on November 25th, 2004 3:46 am

    Carbrera being surprisingly good defensively gives me the idea to use him as next year’s shortstop and keep Lopez in AAA for seasoning (as David has been suggesting). What do you think?

  13. Jerry on November 25th, 2004 9:34 am

    I am wondering if the rest of the league sees Winn as a decent, or slightly above average, defensive centerfielder. If so, he would be a pretty valuable player. After Beltran and Finley, there really aren’t any CFers out there. Kenny Lofton? I am just hoping that the M’s can get a decent player back in a trade for Winn. Perhaps a reliever. With Reed, I think that Winn is very very tradable.

    I am also wondering how much a few good signings would improve the defense. Everyone in the hardcore fan community seems to want Beltre and Drew as the big moves. If Drew is in CF, with Reed in LF, Beltre at 3B, Lopez at SS, and Ibanez at 1B, I wonder how much better the defense would be. Ibanez is the biggest liability, but the M’s always have Spiezio as a solid defensive replacement. I wonder how Drew would compare to Winn in CF. Obviously, his arm is far better. However, if they did sign those players, I would think that moving Ichiro to CF would make a lot of sense. Drew is an above-average RFer, and Ichiro could be great in CF. I don’t know this empirically, but it seems to me that Ichiro has a lot more range. Regardless, with Reed, Ichiro, and Drew, you would have three outfielders who can cover a lot of ground. Reed would be much better than Ibanez in LF. And Beltre, assuming no major defensive change from 2004, would probably be the best defensive thirdbaseman in the AL (or at least comperable to Chavez).

    This is another side of the argument on why Delgado or Sexson is not the right move for the M’s. While Sexson is a good defensive firstbaseman, having strong players at CF and 3B is a lot more important. I totally agree with DMZ that having an incredible defense is like adding an elite player. So if the M’s are going to be signing players for elite money, why not go after younger guys who can help the team on both sides of the equation. Thus, I would rather see the M’s focus on CF and 3B, and look at 1B either through trades or just stick with Ibanez with Spiezio as a replacement. If Bucky is better than Ibanez defensively, then they can switch. Whichever scenario they go for, 1B is easier to address internally than 3B and CF.

  14. IgnatiusReilly on November 25th, 2004 9:44 am

    But his arm…his ARM!

  15. Evan on November 25th, 2004 10:46 am

    Winn’s arm is dreadful, but I don’t think these numbers wree measuring his arm.

    And his arm only matters with runners on. As long as the pitchers keep giving up home runs to every third batter, his arm doesn’t matter.

    Cameron was still way, WAY better than Winn could ever hope to be.

    And Boone needs to package up that gold glove and ship it to Orlando Hudson. The O-Dog was robbed.

  16. David J Corcoran on November 25th, 2004 11:29 am

    Now I remember why I read this blog. Good post!

  17. Matt Williams on November 25th, 2004 2:59 pm

    But…I’ve been trained to believe that my chosen talking head (no matter which side he/she/it is on) is always right, was always right, and will always continue to be right. Going back and re-evaluating an earlier position just isn’t the way the world works. If, through some trick of the other side, it turns out an idea was wrong not completely accurate we have to bury the evidence and turn the attention to something else.

    Just to note, the above wasn’t meant to be (entirely) political. It seems like most people on any side of any issue or connected with any idea/cause operate in a blind manner, dismissing anything they don’t like.

    Gotta agree with everyone above, this was a great post. I remember getting annoyed every time I heard the announcing crew call the Seattle defense great because they had so few errors. Guess they were more right than I thought, although for the wrong reason.

    I think a lot of the critiquing of the defense was because it was put together wrong with Ichiro in right. Like ordering a new car in royal blue and having them accidentally paint it navy blue. Sure, the car may otherwise be wonderful, and maybe you really like navy blue too, but just the fact that it’s wrong will make you focus on everything negative rather than the positives.

  18. Scraps on November 26th, 2004 11:19 am

    I’ve seen this mentioned before, but I haven’t seen it addressed in anin-depth way. Much of the complaint about the Mariners’ defense has been that they gave up a lot more more extra-base hits than before (and I remember seeing early-season numbers showing that the Mets’ extra base hits allowed were way down with Cameron out there).

    Is there a version of Defensive Efficiency that tries to measure not just percentage of balls in play that are turned into outs, but total bases vs balls in play? Is this possible?

  19. Adam M on November 26th, 2004 11:50 am

    Great analysis, glad to see Randy get some credit – from his number with the M’s and the times he played CF and Tampa, it didn’t look like the dropoff would be calamitous.

    Some questions for the gen. public:
    1) With Winn’s arm, it seems like he’d be very solid in left, correct? Could Ibanez play RF regularly? I have never seen him play this position and have no idea. And is Ichiro rock-solid dead-set against playing center regularly? I know Melvin kept shooting down the suggestion, but there were some reports that Ichiro was stubbornly against it, and BoMel was just covering for him.
    2) Did Cam not get a GG this year because he didn’t deserve it this year, or because GG voters are ignorami? I e-mailed Rob Neyer, and he said Cam was clearly the best defensive CF in the NL, better than Jones or Edmonds (who I think is vastly overrated with the glove).
    3) How important is a 1st basemen to an infield’s overall defensive rating? Can you find a clear case of 2 different 1Bs making a huge difference in the performance of their teammates (prob. an ideal test case would be Giambi vs. Nick Johnson with the Yankees a few years ago). i.e. is the ability to “dig tough balls out of the dirt” that important? It seems like, outside the Tino/Sorrento era, the M’s have been pursuing a philosophy of popless, high-OBP, leather-flashing 1B for the past, oh, 20 years or so (Davis, O’Brien, Segui, Olerud). It seemed like they got more out of the pop guys than the gold glovers.
    4) How well does the old adage that you have to be defensively solid “up the middle” (CF, 2B, SS, C) hold up when analyzed statistically? I mean since those guys are going to field most of the balls put in play anyways, it seems like it proves itself. But does anyone know of a study that looked at this?

  20. Evan on November 26th, 2004 1:41 pm

    Cameron is clearly the best defensive CF in baseball, and has been for years. And it’s not particularly close.

    Cammy was injured a bit this year. But my guess is that the NL voters just don’t yet realise how good he is. I don’t think the Mets do either, given their musings about moving him to RF.

  21. Bela Txadux on November 30th, 2004 1:31 am

    Thanks for the intensives here, Derek; food for though a-plenty.

    Defense matters a very great deal, yes. I have thought so for many years, and watching Pat Gillick’s Ms, the outstanding defensive team of this decade and perhaps of their time, I think so now more than ever. The analogy of adding a star player is very apt. Bavasi’s announced transition of last offseason of “trading a little defense for more offense” was, to me, the equivalent of saying “We’ve traded Ichiro for Aurilia to get a little better.” Yup, that’s just what it looked like on the field.

    I hadn’t really though about Cabrera as a _defensive_ factor, although he came with the reputation of a sound glove man. I’m pleased to see that he was considerably better in the field than I’d realized, and happy to see him get credit for real value-added there; I’ve been fixated on his empty bat in looking at him.

    Ibanez simply cannot play 1B. Any talk about his doing this for ’05 is the equivalent of the rumor-mongering a sign of Travis Lee ‘for his offense;’ real and just cause for open derision. I’m not surprised to see that Bucky had reasonable range, and the numbers suggest that if his knee really is cleaned up he might actually have some positives at 1B. However, movement, i.e. range, wasn’t his real liability in my direct observation of him: he has an iron glove. Literally, the ball bounced off his glove at least twice while I was watching him. And those errors, friends, go to other fielders much of the time. I continue to believe that he is a very, very dubious proposition wearing leather. Ole Olerud, as we see, had miserable range; he has NO lateral movement. But he was a high-G gravity well for anything that came into his arm range. And that is where the real value for 1bmen is as I see it, holding onto the ball, not getting to it.

    I still find it very difficult to accept the proposition that Randy Winn was even ‘decent’ in CF. The team’s defense was sharply _better_ in September? Really?? The month where Randy was in LF more than half the time while others got some play in CF, especially Reed??? Yeah, that’s how I saw it, too. Of course some of this may simply be the upgrade in LF for Winn relative to Ibanez, where Raul was _not_ a disaster, just limited. I watched Randy Winn run bad routes and get late breaks in CF right down to the final days of the season. He certainly didn’t LOOK like he was playing the postion at any passable level, although I’ll grant that he was so terrible in April that anything he did subsequently was bound to look like and measure like an improvement. Count me as among the unconvinced here. Perhaps the fact that he looks terrible getting to the ball should not obscure for me the fact that in the end he got/gets to the ball, but I watched him _not_ get to any number of balls, so . . . hhmmmm.

    Spezio’s numbers at 3B came out better than I’d thought, too, but that’s something I’m willing to accept and happy to see. I always believed he was trying hard there, and the numbers say he got some passable results. I’ve always found it hard to judge 3B defensive play simply by watching it. So many balls get driven by the third baseman that someone who cuts off even a few, or makes a few bang-bangs off a hard charge will stand out more statistically it seems to me than at any other position even if day to day you can hardly see the difference on the field.