The Mariners OF Defense 2003-7

by Matthew on Mar 2, 2008 8:15 PM PST

Update [2008-3-4 13:15:51 by Matthew]:: Other teams and entire MLB have been added below in the comments.

Thanks to Derek for the inspiration.

Derek's chart was thought provoking, but it didn't do much to answer an immediate question it raised in me; so how has the Mariner OF defense changed over time? To answer that, we'd need batted ball data for all Mariner pitchers in which to compare the OF putouts against.

To that end, thank you Retrosheet. I wrote up a quick perl script to parse through the event files and pulled the following information for years 2003-7^.

The conclusions are pretty evident.

The 2007 staff allowed many more line drives, about 50 more than normal.
Bavasi's strategy to build a groundball heavy staff failed, FB+LD did not change from 2006 to 2007.
Starting in 2006 and continuing last year, the Mariners' OF was markedly worse at turning flyballs into outs, about 50 plays per year.

Hattip to Nadingo for reminding me about 2004. Now, correlation does not imply causation, but...
2003: Ibanez makes 0 starts in the OF (for SEA).
2004: Ibanez makes 108 starts in the OF.
2005: Ibanez makes 57 starts in the OF.
2006: Ibanez makes 158 starts in the OF.
2007: Ibanez makes 131 starts in the OF.

Raul 4 DH

^I used 2003-7 instead of going back further because Retrosheet made a change in how they categorized batted ball types after the 2002 season.

For John. Primary starters:
2003: Winn, Cameron, Ichiro
2004: Ibanez, Winn, Ichiro
2005: Winn (Ibby after trade), Reed, Ichiro
2006: Ibanez, (Reed+Willie+Doyle), Ichiro
2007: Ibanez, Ichiro, Guillen

0 recs | Comment 44 comments

Story-email Email Printer Print

Comments

How many runs is this costing us?

How significant would the gain of throwing Reed, Corey Patterson, or Lofton in LF be?

"You know, as that was coming out of my mouth, I knew that it was wrong."

by JI on Mar 2, 2008 8:38 PM PST 0 recs

50 plays per year = ~40-45 runs

by Jeff on Mar 2, 2008 8:39 PM PST up 0 recs

good gravy

"You know, as that was coming out of my mouth, I knew that it was wrong."

by JI on Mar 2, 2008 8:42 PM PST up 0 recs

nice work

This is pretty much outfield specific defensive efficiency, no? It comes closer to showing what people were trying to take away from that graph. Still, I wonder how these numbers could have been affected by having 60+ starts go to Ho, Weaver, Baek, Feier and company.

by Mere Tantalisers on Mar 2, 2008 8:41 PM PST 0 recs

Send an open letter

To Ichiro. The managers, coaches, and FO don't listen, but it seems as though the players do.

I will not make jokes in my sig. I will not make jokes in my sig. I will not...

by TIF on Mar 2, 2008 8:47 PM PST 0 recs

This is why I believe even Jeremy Reed

in LF is an improvement. Sure, his bat is worse than Vidro's, but if it gets Raul to DH for the love of god do it!

Wlad, Reed, Jimerson, Lofton, Patterson, anyone but Raul for LF please!!

Yesterday's Pants
A blog-thingy about the Mariners and stuff.

by BrettJMiller on Mar 2, 2008 10:00 PM PST 0 recs

Excellent post.

I was wondering who you would suggest we sign in order to improve our outfield defense? I realize our best option was just traded, so is there a certain outfielder you would suggest seriously going after next offseason?

by Taylor H on Mar 2, 2008 10:06 PM PST 0 recs

Another thing to take away from this:

LDPO% is really random.

by Graham on Mar 3, 2008 12:56 AM PST 0 recs

Scary to think what would have happened

if it had hit the 2006 mark.

...and now I'm here

by Librocrat on Mar 3, 2008 1:01 AM PST up 0 recs

What do 2004, 2006, and 2007 have in common?

Besides having the three lowest FBPO%?

Raul Ibanez, starting Left Fielder!

by Nadingo on Mar 3, 2008 6:34 AM PST 0 recs

Hey matthew - a quick request

can you tell your script to grab the equivalent numbers for some other AL teams?

I'd just like to put these figures in context, and I'd also like to see how well it lines up with THT's outfield +/- figures (or the defensive metrics).

by marc w on Mar 3, 2008 9:27 AM PST 0 recs

Seconded.

I thought of this, too -- I'd like to see how these #s relate to other teams in the league.

Ill Ligitamus Non Carberendum

by PositivePaul on Mar 3, 2008 9:30 AM PST up 0 recs

Yeah, I'll do that later

by Matthew on Mar 3, 2008 10:30 AM PST up 0 recs

Geoff Baker needs to read this.

He doesn't see how Raul KILLS the team defensively in LF. I'm no stathead, but this convinces me...

Ill Ligitamus Non Carberendum

by PositivePaul on Mar 3, 2008 9:29 AM PST 0 recs

give him some credit

he acknowledges Raul's shortcomings. he just stops short of ripping into him every day like we do cause he, you know, talks to him all the time. And Raul's a big dude.

by Mere Tantalisers on Mar 3, 2008 10:06 AM PST up 0 recs

Geoff has stated that he paid close attention

To the balls Raul did not catch. His argument was that it never hurt the team. I like Geoff a lot, but I have no idea what games he was watching to come up with that conclusion.

by Sec 108 on Mar 3, 2008 10:49 AM PST up 0 recs

This is a very convincing chart!

It would be even more useful (as an article) if you listed who the outfielders were for the years you studied (for those too lazy to go look it up.) And provided comparisons to at least one other team... maybe LAA.

by johnbai on Mar 3, 2008 11:01 AM PST up 0 recs

starters added

by Matthew on Mar 3, 2008 11:19 AM PST up 0 recs

Worse Pitching does play a big role in this

When you have pitchers serving up goferballs like Weaver and Horam, that will result in an increase in line drive rate and a decrease in the "catchability" of line drives and fly balls in the outfield.

by vkut79 on Mar 3, 2008 10:52 AM PST 0 recs

Line drive rate I buy, that's obvious

the "catchability" of flyballs I do not. I want to see numbers on this.

HoRam's and Weaver's HR/FB% (and HR/G) were actually lower than career norms. That doesn't lead me to think they were serving up gopherballs at an increased rate.

by Matthew on Mar 3, 2008 11:02 AM PST up 0 recs

Well if the pitcher throws a ball that is

easier to hit (as Weaver and Horam did last year), and the hitter hits a flyball, there is a greater chance that the ball will go closer to the wall, where it is harder to catch, and this will make it harder for any fielder to catch the ball. Its a smaller effect than with line drives, because there is more time to get to the ball, but it will still reduce the PO %.

When I am talking about Weaver and Horam throwing balls that are easier to hit, I am talking about the Mariners rotation overall and how the addition of these two pitchers last year increased the amount of easy-to-hit-hard pitches thrown over the course of the year (from 06 to 07). So if you agree that FB and LD PO%'s decrease when more easy-to-hit-hard pitches are being thrown by a bad pitching staff, then that could help explain the decrease in these numbers from 06 to 07. Still, the decrease is not even that major. I am skeptical of how much credit people are giving to the Mariners poor defense in explaining this trend when the pitching also got significantly worse from '06 to '07 (although I don't know the numbers exactly for this).

by vkut79 on Mar 3, 2008 11:36 AM PST up 0 recs

My whole point is that I don't agree

that poor pitchers give up harder to catch flyballs or linedrives. I need to see a study.

For instance, if Weaver was easier to hit and thus the flyballs he gave up were hit 5% further than those of say Washburn's (even this I'm not granting is true, just hypothetical), the OF could compensate for that by playing deeper.

What you would need is not a longer distance on the average flyball yielded, but a wider dispersion. Again, it's not something I can buy until I see a conclusive study on the matter.

by Matthew on Mar 3, 2008 11:43 AM PST up 0 recs

Hmm okay

Of course having a detailed study would be best.

I don't think playing deeper would help much. I think if the balls are being hit harder, then there will be a wider dispersion, because the distribution of distance travelled will have more variation. This is just based on simple physics. If balls are hit harder, on average, they have access to more space where they could land. The exact location would depend on bat alignment. Its not like you would take all the balls hit and increase their speed by 5% - that wouldn't increase the variation - there would still be balls hit at lower speeds, but also some hit at higher speeds.

So I think if a hitter can read the pitcher, and can see his pitch well, and is able to make good contact with the pitch, he can hit it harder and as a result there will be greater variation in the distribution of places where that ball could go, which would decrease the PO%.

by vkut79 on Mar 3, 2008 11:57 AM PST up 0 recs

The outfield is finite

Really hard hit balls are called 'home runs'

by Graham on Mar 3, 2008 11:59 AM PST up 0 recs

Yes, its finite

But as you know its somewhat rare for the ball to actually leave the field. Within the boundaries of the field, there is still room to increase the variation.

Also I'm not necessarily talking about "true" fly balls, the kind that always leave the yard if hit hard enough. I'm also talking about "high drive" hits that are sort of in between the two, and these are the ones that have the greatest amount of variation in where they land. For example, the kind of ball that typically settles into the gap about 4-5 seconds after being hit.

by vkut79 on Mar 3, 2008 12:04 PM PST up 0 recs

What you are talking about is guesses.

You're going to have to produce some actual numbers on this in order to show anything.

by Matthew on Mar 3, 2008 12:09 PM PST up 0 recs

I'm not sure how I could get numbers on it

Educated guesses is the best I can come up with.

by vkut79 on Mar 3, 2008 1:45 PM PST up 0 recs

I believe what you're talking about is BABIP

As in, Batting Average on Balls in Play. And if I understand my defense-independent pitching statistics properly, the vast majority of pitchers have very little if any control over how often a ball in play gets turned into an out. League-average BABIP for pitchers tends to be .300, and while the actual number for a pitcher may fluctuate around that in a given season, maintaining a low BABIP is generally not a repeatable skill.

by Nadingo on Mar 4, 2008 6:15 AM PST up 0 recs

Actually the 06-07 doesn't really even matter

That's not the real point of this analysis - the 5 year trend is what matters. Never mind that.

by vkut79 on Mar 3, 2008 11:44 AM PST up 0 recs

Suggestions Needed Regarding Other Teams

Any particular OFs (keep it between 03 and 07) you'd like to the see the above numbers for? Could be our division foes or OFs that could illustrate the values of particular people (for instance San Diego 2005 to 2006 which went from Klesko-Roberts-Giles to Roberts-Cameron-Giles).

by Matthew on Mar 3, 2008 11:14 AM PST 0 recs

I'd be interested in Boston and Cincinnati

for 2007, at least, to see the effect of Manny and Griffey+Dunn. As well as two teams with phenomenal OF defense and par OF defense. Off the top of my head, I think Twins and Angels were ~average and... I can't think of any phenomenal OF defense right now, but I just woke up...

...and now I'm here

by Librocrat on Mar 3, 2008 11:26 AM PST up 0 recs

Cincinnati especially

since their two corner outfielders are atrocious. It would be interesting to see how we compared.

...and now I'm here

by Librocrat on Mar 3, 2008 11:27 AM PST up 0 recs

Boston I won't do

because the Monster gums up everything, but Cincy and ANA I'll do.

by Matthew on Mar 3, 2008 11:35 AM PST up 0 recs

How about

Anaheim, Oakland, New York.

Anaheim because I'm always so surprised that defensive metrics LOVED Erstad and Anderson;
Oakland because they've had so much turnover, and because they had to use Jack Cust and Nick Swisher in the OF last year and STILL had a better DER than Seattle;
New York because I think it might be interesting to see the DER plummet as Bernie got old, then to see what difference Melky made last year.

Cinci is ok, but I'm a bit wary of extreme pitcher's or hitter's parks...

by marc w on Mar 3, 2008 11:40 AM PST up 0 recs

Oops, one more

Detroit.

by marc w on Mar 3, 2008 11:41 AM PST up 0 recs

Thanks for updating this Matthew

with the new suggestions. Should be an awesome document when finished... something that could draw the attention of other blogs if not the M's FO.

by johnbai on Mar 3, 2008 1:44 PM PST 0 recs

Well, Tango put a link up

on the Book Blog and some asshole indirectly linked to it at THT.

by Matthew on Mar 3, 2008 1:56 PM PST up 0 recs

Per Request

Here are 6 additional teams:
OAK, ANA, NYY, TB, DET and NYM

WTF is up with the '04 Tigers? That seemed like one of the worst collections of OF in recent history. Craig Monroe, Bobby Higginson, Nook Logan, Rondell White and Alex Sanchez. Were they secretly good? I think Detroit's park is messing with some of these numbers.

Man, the 2003 Mariners OF was insane. How insane? Well, here's the entire MLB:

And the plus/minus plays for the Mariners compared to MLB average:

Year: (LD/FB plays), (total plays), (est. runs)

+42/+67, +110, +88

+17/+16, +33, +27

+24/+47, +71, +57

+5/-1, +5, +4

+20/-28, -8, -6

Keep in mind that traditionally the NL >>>> AL when it comes to defense so any AL OF that's MLB average is going to be one of the best OFs in the AL.

BIG CAVEAT: NUMBERS NOT PARK ADJUSTED. SAFECO FIELD HELPS BOOST OF NUMBERS!!!!!!!

by Matthew on Mar 3, 2008 11:53 PM PST 0 recs

I imagine there will be a lot of surprise

over the '07 Ms total figure. Well, I'll point out that it does jive with the THT numbers which state that the problem with our defense was actually in our infield, not the outfield.

by Matthew on Mar 3, 2008 11:58 PM PST up 0 recs

Ichiro!

by Jed MC on Mar 4, 2008 7:30 AM PST up 0 recs

Why?

The M's seem solidly below average in FBPO%, and weirdly above average in LDPO%.
LDPO% seems to vary from year to year, probably randomly (that's your next assignment - how does LDPO% correlate with.. oh, FBPO%, RZR, UZR, PMR?).

THT numbers show the M's were below average in RZR for the outfield, but made more out of zone plays (Ichiro). That makes sense, and while it's not out of line with what you've presented here, I do wonder how to weight the low RZR versus the high OOZ. These data would tend to suggest the RZR matters a bit more (as you'd expect, really).

It's also interesting to see how much the league averages move in LDPO and FBPO%.

by marc w on Mar 4, 2008 8:46 AM PST up 0 recs

Doyle

I just wanted to mention that in 2006 Snelling is listed as having played CF, when in fact he played RF after Ichiro! took over CF. In fact Doyle only played one inning in CF that year.

Wrad Bilkerson!

by larrybowa on Mar 4, 2008 8:37 AM PST 0 recs