This And That

by Jeff Sullivan on Oct 31, 2006 5:58 PM PST in Seattle Mariners News, Regular

According to the new Elias rankings (please don't try to make sense of them, on account of you can't), Gil Meche is a Type B free agent. What this means is that, if and when the Mariners offer him arbitration and he declines in favor of signing a multi-year contract somewhere else, the organization receives a compensatory sandwich pick between the first and second rounds of next year's draft, as stated by the new CBA.

Since the team finished in the lower half in terms of W/L record, their first-round pick is protected in the event that they sign a Type A free agent (Schmidt, Zito). Should that happen, they'll forfeit their second-round pick instead. Which would kind of be the last of my concerns if the Mariners actually went out and paid what it's going to take to land Jason Schmidt or Barry Zito.

Also, a little while back Chris Dial posted his AL Gold Gloves, according to the methodology linked within his article. Of note is that (A) he thinks Ichiro is deserving by virtue of his time spent in center, and (B) yet another defensive metric isn't buying the Betancourt hype. If you think about it, you can begin to understand why Yankee fans have grown so weary of hearing about UZR and Jeter's defense - the numbers simply don't match up with what they observe with their own eyes, just as is the case with our own flashy shortstop. It's not the same, but it's the first example that came to mind.

While we're on the subject, a much more accurate way of measuring defensive performance is to break it down by team instead of individual. And that's precisely what the Hardball Times does here, with values being reported as "plays (not runs) made above/below average." The Mariners came out at +2 overall, +12 on grounders and -10 on flyballs (or, if you prefer, +12 in the infield and -10 in the outfield). If I had to guess where the individuals stood off the top of my head, I'd have Sexson around -8, Lopez around -4, and Betancourt/Beltre each at +12. Might not be 100% accurate, but it makes more sense than some of the other metrics we've seen.

Anyway, let's award some Gold Gloves by team:

Best AL Infield: Detroit, +74 (that's funny to me)
Best AL Outfield: Cleveland, +48
Best AL Defense: Detroit, +56

Best NL Infield: Houston, +86 (probably all Adam Everett)
Best NL Outfield: Atlanta, +51
Best NL Defense: San Diego, +71

It's worth pointing out that fielding flies is more valuable on a play-by-play basis than fielding grounders, since the former generally go for extra bases while the latter go for singles, which means you probably have to apply some sort of weighting system in order to determine who was really the best team defense in each league, but I don't feel like messing around with that right now.

This goes a long way towards explaining why Detroit and San Diego were able to make the playoffs with what were, on the surface, relatively underwhelming rosters. 50 runs in the field are every bit as good as 50 runs at the plate or on the mound, after all, but they're much harder to notice, particularly when you're just looking a team over on paper. Also, special kudos go out to the Indians and Devil Rays, who were each about as bad in the infield as Houston was good. I don't know what -80 looks like over the course of a season, and God willing it'll stay that way for as long as I'm alive. Never leave us, Yuni.

Tweet Comment 27 comments | 0 recs |

Do you like this story?

Comments

This looks like a good time to field one of my

questions about defense.

So all the time, we say that this player or that player is +/- so many runs. This kinda makes sense and is easy to think about because at the end of the year you can look at the number of runs allowed and just say that if you would have had an average defensive player we would have had x number of runs instead of what we ended with.

The problem I may have with this is that we use it to also say that player A may be +4 runs for defensive but -4 runs for offense. Does this mean that he is an average player or is a run saved more that a run scored like some people say?

by Edgar for Pres on Oct 31, 2006 10:21 PM PST reply actions

The difference between

the value of a run saved vs. the value of a run produced is very small. So while adding them up directly isn't 100% accurate, for all intents and purposes it's very close, and making the mathematical adjustment probably isn't worth doing (especially since defensive numbers are so unreliable anyway).

by Jeff Sullivan on Nov 1, 2006 5:14 AM PST up reply actions

Well,

I'd still prefer punting defense for offense at some positions for the given offensive upgrade... Defense stops runs, but you can never run a positive differential solely on defense. You can save as much as you want, but the score will never be above 0-0 (though this is coming from the extreme 1.000 OBP sorta view).

Detect-O-Vision

by IcebreakerX on Nov 1, 2006 6:45 AM PST up reply actions

I'm with you there.

Give me Derek Jeter over Adam Everett any day.

And on top of that, the biggest problem with trying to build a team defense-first is that we still don't really know how to measure defense, but we have a million ways of measuring offense. So building with bats gives you a high degree of certainty, whereas building with defense doesn't at all.

by Jeff Sullivan on Nov 1, 2006 7:28 AM PST up reply actions

Another thing is...

Wouldn't defensive value also depend on other factors, such as pitching? (I'm not familiar enough with the more modern defensive metrics to know whether this is addressed.)

Let's continue with the extremes and say, we have a staff of pitchers that run a K/9 of 27.00 and a H/9 of 0.00.

In other words, a team that never lets the ball in play.

That would basically make the defense effectively worthless. You could field 9 Giambis or Edgars or Mannys (ok, I guess you'd still need a catcher) and it wouldn't matter. The team could have perfect or utterly flawed team defense (aside from catcher), but it'd have little to no outcome on the game.

The same would go for a 100% FB or 100% GB staff. You'd effectively reduce the amount of runs the defense could save depending on the staff's tendencies. A defense behind high K/9, low H/9 staff would inherently save less real runs.

So wouldn't the value of the virtual runs saved be extremely prone to variation of real runs? This intuitively makes sense as defense is the third line of action in a single play (Pitch -> swing -> in/not in play).

Ok, I'm going to shut up now since I'm sleepy, which means I'm probably not thinking straight.

Detect-O-Vision

by IcebreakerX on Nov 1, 2006 7:44 AM PST up reply actions

You are right

and I'm not sure whether or not the better metrics (predominantly UZR) adjust for fielding opportunities. I'd imagine they do, but I don't know how, so I can't really speak to that.

Teams that have strikeout pitching staffs (Cubs) can afford to have a worse defense than teams with contact pitching staffs (Royals), but the difference is as little as one or two plays a game. Over the course of a full season a terrific defense might turn those into 10-20 more outs than an awful defense, which - while significant - isn't that large. So it's better to build a defense independent of your pitching staff's tendencies, because a strikeout staff won't spare you the harm thata bunch of bad gloves can do.

The same goes for groundball/flyball staffs, although again, the difference across teams isn't huge - the most extreme GB staff got an average of about two more grounders a game than the most extreme FB staff this year, so building a defense for the specific purpose of turning more grounders or flies into outs isn't as wise as just putting together a solid group in general. You can adjust it on a day-to-day basis depending on whether you have Brandon Webb or Jarrod Washburn on the mound, because then it'll make a much more significant difference, but over a full season it's not that important.

by Jeff Sullivan on Nov 1, 2006 9:05 AM PST up reply actions

Groundballs / flyballs

I'm not sure you can say that over a full season it's not that important.

For example, THT has Detroit at +74 on groundballs while Cleveland is -81 on groundballs. That's a huge difference.

visiting A's fan.

by rfloh @ Lookout Landing on Nov 1, 2006 9:21 AM PST up reply actions

Re: Yes, but

It's not a predictable difference. Both Cleveland and Detroit look like outliars.

I'm one of the biggest proponants of the undervaluedness of defense, and I don't totally agree with Jeff - I think building a pitching staff and defense that complement each other can be a great way to maximize return on investment - but there's no way the Tigers are really a +74 defense on ground balls.

They may have performed that way during the past 162 games, but it's not a true talent level. Much like the way the Mariners defense played way over their heads in '01 or the Angels outfield defense was so ridiculously awesome in '02, I believe Detroit's defense was really good and somewhat fortunate at the same time.

Defense matters, for sure. But the difference between what you can expect from the best and worst defense isn't 150 outs a year.

by davidcameron on Nov 1, 2006 9:28 AM PST up reply actions

Well, that's not really what I meant.

What I meant was that building a strong infield/outfield defense to complement a groundball/flyball pitching staff isn't that important because your team with a great infield will still have to field a bunch of fly balls, and your team with a great outfield will still have to field a bunch of grounders. That's why I pointed out the teams on either end of the spectrum - the difference between the most and least GB-friendly pitching staffs in the league last year was a hair over 2 grounders a game. When you're already dealing with 25-30 balls in play each day, 2 represents only a small fraction.

I do agree with Dave above that having a defense that complements the pitching staff can help some - there's always room for improvement, and doing a better job of fielding those two grounders a game will save one or two dozen runs over a season. Where I think that idea runs into problems is that it's tough to put that plan into practice, because pitching staffs have a fairly high turnover rate, and if you (say) build a team with a better infield than outfield you're going to end up limiting yourself to groundball pitchers when you need to fill a hole in the rotation. If you've got a number of starters sticking around for the long haul, then sure, build a defense that helps them, but that usually isn't the case.

by Jeff Sullivan on Nov 1, 2006 9:41 AM PST up reply actions

Basically I agree with you both

I agree with you that the idea is great in theory but not all that easy to implement in practice, and with Dave that outliers like Cleveland happen, and should have put that in my post.

Interestingly, I believe THT's 2006 annual had Cleveland as one of the best defenses in the majors.

visiting A's fan.

by rfloh @ Lookout Landing on Nov 1, 2006 11:39 AM PST up reply actions

Indeed they did

but a lot of that was the Crisp/Sizemore/Blake outfield. The infield was just a little above average (and appears to have completely tanked in 2006).

by Jeff Sullivan on Nov 1, 2006 2:58 PM PST up reply actions

Well

Their 2006 Opening Day 2B has a ring. Can't say much for the rest of them, though.

by Gomez on Nov 2, 2006 7:38 AM PST up reply actions

Yeah

From the modelling I've done it looks like a run saved is worth roughly 1.02 times a run scored.

There's no way defensive metrics are that accurate.

by Graham MacAree on Nov 1, 2006 7:08 AM PST up reply actions

Neifi Perez

I note with great interest that Alias rates Neifi Perez higher than Betancourt & Lopez. Must be because of his grittiness & clutchiness.

We are the holler men, we are the glove men, mouthpiece filled with plug: ptoo!

by Celadus on Nov 1, 2006 12:10 AM PST reply actions

Typo

Um . . . make that Elias

We are the holler men, we are the glove men, mouthpiece filled with plug: ptoo!

by Celadus on Nov 1, 2006 12:11 AM PST reply actions

Also...

"23t Julio Mateo 68.268 A"

Bahahaha

by Graham MacAree on Nov 1, 2006 7:34 AM PST reply actions

Does infield include catcher?

by Graham MacAree on Nov 1, 2006 9:35 AM PST reply actions

Sort of.

But at least as far as THT is concerned, how many grounders do catchers really field each season?

Catcher defense is important, but it doesn't really show up here.

by Jeff Sullivan on Nov 1, 2006 9:42 AM PST up reply actions

I remember looking at THT's

Defensive stats halfway through the season, and noting that the top catchers were corresponding to the top infield defenses, is why I ask.

Was that just random coincidence?

by Graham MacAree on Nov 1, 2006 10:12 AM PST up reply actions

I'm guessing it is.

I can't imagine why they'd have a real significant influence, unless you buy the whole "catcher = team leader" idea (which I don't, at least not in this respect).

by Jeff Sullivan on Nov 1, 2006 10:15 AM PST up reply actions

Well, the only other idea

That springs to mind is that they integrate the running game into their 'plays'.

Since I have no idea as to their methodology (and care -just enough- to be really curious, but not enough to devote time to research), I find that plausible, but could be dead wrong.

by Graham MacAree on Nov 1, 2006 10:24 AM PST up reply actions