Chone Smith's defensive projections for 2009

Obviously, these are just projections, and the less time a player has, the less data there is, so the usual caveats apply about defense need at least three years to get a sense of true talents, small sample size for part time players, fielding slumps, etc.). There is also some "controversy" because he regresses to the Fans Scouting Report rather than league average, but hey, I'm just reporting. Read about the methodology here.

Below are what he has for many Royals, as well as some other players "of interest." I've done it in runs @position format. At least I think the numbers represent runs. "corner" = corner outfield (all one position here, for some reason). Not sure how the got the numbers for CF for some guys, maybe they just played an inning way back when and it's still in Smith's database, or morel likely there's a formula for extrapolating what a player would do at the other outfield positions from their performance elsewhere.

Infielders (primarily):

Alex Gordon -9 @3B
Tony Pena, Jr. +3 @SS
Mike Aviles +4 @2B, +3 @3B, +5 @SS
Alberto Callaspo -7 @2b, -4 @3B, -8 @SS
Esteban German -15 @2B, -13 @3B, -20 @SS, -7 @corner, -23 @CF

Billy Butler -4 @1B
Ross Gload -1 @1B
Ryan Shealy +3 @1B
Mike Jacobs -8 @1B
Russell Branyan -1 @1B, -5 @3B

Orlando Hudson +3 @2B
Rafael Furcal +2 @SS
Mark Ellis +15 @2B

Outfielders (primarily):

David DeJesus +9 @Corner, +3 @CF
Mark Teahen +2 @1B, -11 @3B, +4 @corner, -5 @center
Jose Guillen, -11 @corner, -23 @center (?)
Joey Gathright +11 @corner, +4 @center
Mitch Maier +4 @corner, -4 @center

Franklin Gutierrez +22 @corner, +19 @center
Ryan Langerhans +14 @corner, +9 @center (sign him!)
Matt Stairs -13 @corner, -27 @center (thanks, we needed to know whether or not to acquire Stairs to platoon in center with Gutierrez)

Nick Swisher +5 @corner, -3 @center, 0 @1B
Ben Francisco -2 @corner, -11 @center
Adam Dunn -13 @corner, -27 @center (Stairs=Dunn)
Pat Burrell -14 @corner, -28 center (Stairs>>>Burrell)
Bobby Abreu -11 @corner, -24 @center

I'm sure that's enough. Anyone else that interests you can be looked up.

Again, remember that these are these are projections for next year, not evaluations of the last year (although past performance obviously is the main factor in the projections). Sample sizes vary greatly depending on how much time a player has played at a position, how long ago it was, how long they've been in the majors, etc.

Enjoy! Discuss! Argue!

about 1 year ago devil_fingers 60 comments 1 recs |

Story-email Email Printer Print

Comments

If defensive metrics are suspect, defensive projections are crap

Sure the great defenders should be great and the awful ones should be awful. But these projections, for the huge mass of guys in the middle are pretty worthless. Both Aviles and Callaspo should be better at SS than 2B? Why? Because small sample sizes lead to poor metrics and worse projections? Yeah.

This is just my opinion. I could easily be wrong.

by NYRoyal on Oct 28, 2008 12:58 AM EDT reply actions 2 recs

yes to this.

Rowdy Hardy Fan Club member.

by doublestix on Oct 28, 2008 2:32 AM EDT up reply actions 0 recs

and +1 to that

This space intentionally left blank.

by marbotty on Oct 28, 2008 7:17 AM EDT up reply actions 0 recs

for players with very little data, there are implied huge error bars. get over it.

Beyond the Boxscore // Calling BJ Upton lazy is lazy.

by Sky Kalkman on Oct 28, 2008 11:54 AM EDT up reply actions 1 recs

+1 and rec'd

by DarthYoshi on Oct 28, 2008 12:01 PM EDT up reply actions 0 recs

can you explain why the projections for the guys in the middle are worthless?

is it anything other than small sample sizes? if so, please elaborate. if not, then i think we’re on the same page here: little data bad, big data good.

Beyond the Boxscore // Calling BJ Upton lazy is lazy.

by Sky Kalkman on Oct 28, 2008 12:24 PM EDT up reply actions 0 recs

I'll give you th short answer

First, worthless was an overstatemnt. “Of low value” would have been more appropriate. Second, I still think that even the best advanced defensive metrics aren’t great. Defensive metrics are in their fetal stage of development and have a long way to go. So, when you add the limitations of any projection system to the limitations of defensive metrics (along with a lot of small sample size issues for many players) you’re left with something pretty unreliable for projecting defensive performance.

This is just my opinion. I could easily be wrong.

by NYRoyal on Oct 28, 2008 2:28 PM EDT up reply actions 0 recs

so we're good on small sample sizes.

and i probably like defensive metrics more than you, but understand your points.

i just think the main thing I disagree with is that a projection is worse than a year’s worth data. i’d rather have a projection than just 2008 data, because it actually contains MORE information (specifically 2007, 2006, etc data). take hitters for example. we know that Manny was awesome in 2008 and we know his stats exactly. but now that’s irrelevant. we care about how good he actually his and how good he’ll be next year, not what he did in 2008. so you take all the data you can get, analyze it well and get a picture going forward. no, 2009 results will not match 2009 projections as well as some people hope, but they DO describe true talent pretty well.

i mean, you KNOW a coin will come up heads 50% of the time. but if you flip it ten times, you’ll rarely peg how many heads come up. does that make the knowledge that it’s a 50/50 coin of low value?

Beyond the Boxscore // Calling BJ Upton lazy is lazy.

by Sky Kalkman on Oct 28, 2008 3:14 PM EDT up reply actions 0 recs

i just think the main thing I disagree with is that a projection is worse than a year’s worth data. i’d rather have a projection than just 2008 data, because it actually contains MORE information (specifically 2007, 2006, etc data)

That’s a valid point. I think multiple years of data on fielding (or pitching or hitting) are more meaningful than one year, even if that one year is the most recent year. And I realize that projections are made from multiple years of data. But projecting the future is inherently difficult. It is much easier to describe the past than predict the future. Creating projections from inherently soft metrics is, I think, adding one layer of fuzzy uncertainty to another. That doesn’t mean it isn’t worth attempting.

I do think that creating a dichotomy between one year of fielding stats on the one hand and projections on the other hand is a false choice. I think the better comparison is between multiple years of multiple fielding metrics vs. projections based on multiple years of a small subset of fielding metrics. I’d go with a good combination of data from multiple years over any projection like this. Of all of the fielding metrics (and I find them all meaningful but suspect), I value UZR and Dewan’s +/- the highest. So I really didn’t like Chone’s methodology of using ZR, RZR and Totalzone (and then running the projections even before Totalzone’s 2008 numbers came out). I assume he used those because they were only ones available to him for 2004-2008. If that was the best he had available, then it is understandable and I appreciate the effort. But I don’t think it makes for a meaningful and/or reliable projection.

This is just my opinion. I could easily be wrong.

by NYRoyal on Oct 28, 2008 4:19 PM EDT up reply actions 0 recs

there's an inherent "loss of accuracy" when going from past results to future results

you have it with hitting stats, pitching stats, and fielding stats. sure, trying to predict the future will widen your error bars, but you should also widen/lower your expectations with projection. a projection is exactly the same thing as trying to peg the true talent of a player over the past few years. and true talent is NOT just the past stats we’ve recorded.

your point about using the best fielding stats available is a good one. UZR and Dewan are both better, individually, than any of ZR, RZR, and TZ. but UZR and Dewan are both BIS-based metrics and are pretty similar. You’d want to ditch one in favor of a STATS-based advanced metric, and PMR is really the only one out there. Including TZ is done because it is yet a third data set (retrosheet). and you could think of the fans scouting report as a fourth data set instead of just a regression target.

i’m not trying to refute what you said, just expand on it.

Beyond the Boxscore // Calling BJ Upton lazy is lazy.

by Sky Kalkman on Oct 28, 2008 4:30 PM EDT up reply actions 0 recs

I think our disagreement isn't fundamental; only a matter of degree

On a slightly different subject, how much value do you think there is in the Fans Scouting Report? My gut tells me that it is inherently unreliable because it reflects a bunch of fan gut opinions. Is it more reliable or meaningful than that? If so, is it useful for any purpose?

This is just my opinion. I could easily be wrong.

by NYRoyal on Oct 28, 2008 5:08 PM EDT up reply actions 0 recs

Sky can give more details

but Tango has found that the FSR actually matches up surprisingly well with “objective” metrics.

OMG Banny. FWIW I am only crdtng u w/3 runs allwd bc of DDJ OMFG

by devil_fingers on Oct 28, 2008 5:10 PM EDT up reply actions 0 recs

Then why don’t we stick with the objective measures (and continuing to improve them) instead of relying (to any degree) on the vagaries of subjective fan opinion?

This is just my opinion. I could easily be wrong.

by NYRoyal on Oct 28, 2008 5:17 PM EDT up reply actions 0 recs

I should really let Sky answer this

I think it started out as an “experiment,” to see how well fans do. He then found they did pretty well. Moreover, the fans who participate in it tend to be pretty hardcore. By adding it to defensive evaluations, it is supposed to correct some of the problems that people like you and doublestix have with defensive metrics.

But I’m getting dangerously close to talking out of my butt (if I’m not already), so that’s my guess. I think you’d find that you liked the projections less if they didn’t have the FSR element as part of it.

OMG Banny. FWIW I am only crdtng u w/3 runs allwd bc of DDJ OMFG

by devil_fingers on Oct 28, 2008 5:21 PM EDT up reply actions 0 recs

Perhaps so

This is just my opinion. I could easily be wrong.

by NYRoyal on Oct 28, 2008 5:34 PM EDT up reply actions 0 recs

Wait, when did UZR switch to BIS?

According to MGL, he uses STATS for UZR.

by Gopherballs on Oct 28, 2008 7:17 PM EDT up reply actions 0 recs

well i'm completely wrong then

thanks for clearing that up. a combination of UZR and Dewan would be great

Beyond the Boxscore // Calling BJ Upton lazy is lazy.

by Sky Kalkman on Oct 28, 2008 8:40 PM EDT up reply actions 0 recs

My kingdom for a unified fielding metric

This is just my opinion. I could easily be wrong.

by NYRoyal on Oct 28, 2008 8:53 PM EDT up reply actions 0 recs

hee hee… I get your point. But at this point, given the known lacunae (hey, Firefox says that’s right on the first try) in fielding systems in general, isn’t it best to have a couple to look at to get an “overall” feel for how guys are doing out there?

OMG Banny. FWIW I am only crdtng u w/3 runs allwd bc of DDJ OMFG

by devil_fingers on Oct 28, 2008 10:00 PM EDT up reply actions 0 recs

Yeah, I think at this point you have to look at every good fielding metric and try to come up with as complete a picture as possible. I find it hard to hang my hat on any one of them.

This is just my opinion. I could easily be wrong.

by NYRoyal on Oct 28, 2008 10:54 PM EDT up reply actions 0 recs

I know that numbers don't play the game...

but they are real enough to serve as a coatrack?

OMG Banny. FWIW I am only crdtng u w/3 runs allwd bc of DDJ OMFG

by devil_fingers on Oct 28, 2008 11:01 PM EDT up reply actions 0 recs

They are real if we believe in them

And if we all wish hard and clap together, we can bring Earl Weaver back from the dead.

This is just my opinion. I could easily be wrong.

by NYRoyal on Oct 29, 2008 12:00 AM EDT up reply actions 0 recs

I was going to say... you had me pretty mixed up

I thought the idea of “averaging” ZR and RZR was to get a budget “average” of Dewan’s and UZR

OMG Banny. FWIW I am only crdtng u w/3 runs allwd bc of DDJ OMFG

by devil_fingers on Oct 28, 2008 10:00 PM EDT up reply actions 0 recs

it's better to regress to a scouting report than league-average

regression is just saying “hmm, we’re not totally confident in the data, so let’s hedge our bets towards our best-guess”. a scouting report is a better best-guess than league-average.

the default difference between a corner outfielder and a center fielder is 10 runs, which is how Sean gives every OF a rating at both positions (there’s some black box combination of actual performance and the 10-run rule going on). he obviously didn’t use the typical position adjustments for the infield positions, just running individual projections at each position.

and while I get the point about Aviles at 2B/SS, come on now. we KNOW Aviles didn’t play much 2B, so it should be obvious the error bars are quite high for that projection. i mean, if a guy has 50 at-bats, hits .350 and his projection going forward is to hit .280, are we going to call the projection crap? A -1 projection at 2B is basically just calling him average, because there’s almost no data. Sean’s not making some big bold claim that Aviles is without a doubt a better 2B than SS.

for guys without only one year (or partial years) these fielding projection are obviously not that great — but the same goes for a hitting projection. but for guys with a few years now, they are really really useful. for one, you’ve got multiple years of fielding data, which presents a much better picture. and two, you’ve got some nice regression towards what visual observers actually think of a fielder.

Beyond the Boxscore // Calling BJ Upton lazy is lazy.

by Sky Kalkman on Oct 28, 2008 11:54 AM EDT reply actions 0 recs

it doesn't matter how much you qualify something here, Sky, as I did above

people are going to get mad if it says what they don’t want to hear… I might have well have posted that DMGM was dumb enough to trade for Mike Jacobs.

OMG Banny. FWIW I am only crdtng u w/3 runs allwd bc of DDJ OMFG

by devil_fingers on Oct 28, 2008 12:01 PM EDT up reply actions 1 recs

Sorry

that was just an over-reaction to the over-reactions above. I actually thought this morning that I should post a note saying that, in addition to the numerous qualifications I made in the post, I should add that the projectinos for players playing multiple positions only look at the positions separately. The program obviously doesn’t do any sort of regression to the most difficult/most frequently played. So I agree that this doesn’t tell us much about Aviles at second, or Callaspo at short, etc.

I to think it tells us how the combined metrics (and hardcore fan scouting report, which has been shown to be pretty damn good) see Aviles at SS and Callaspo at 2nd going forward.

Also tells us what every metric I’ve seen confirms — Ross Gload is replacement level on his best day. But everyone already knows that.

OMG Banny. FWIW I am only crdtng u w/3 runs allwd bc of DDJ OMFG

by devil_fingers on Oct 28, 2008 12:19 PM EDT up reply actions 0 recs

i wish my comment wasn't so long

the point is that faulting a projection based on its projections of players with very little data is short-sighted.

Sean very well could have just filtered out everyone with less than half a season’s data at any one position and the specific complaint above would be moot.

the thing is that a projection like this is actually BETTER than a single year of data, not worse. more data + plus fans’ scouting report + regression = GOOD.

maybe Sean should have included error bars based on each player’s sample size, but that doesn’t mean there aren’t there.

Beyond the Boxscore // Calling BJ Upton lazy is lazy.

by Sky Kalkman on Oct 28, 2008 12:23 PM EDT up reply actions 0 recs

I think people who are nerdy enough to read this

probably already know how much playing time these guys got.

not to nag, but I just sent you another email, Sky…

OMG Banny. FWIW I am only crdtng u w/3 runs allwd bc of DDJ OMFG

by devil_fingers on Oct 28, 2008 12:25 PM EDT up reply actions 0 recs

or he said

OMG Banny. FWIW I am only crdtng u w/3 runs allwd bc of DDJ OMFG

by devil_fingers on Oct 28, 2008 12:25 PM EDT up reply actions 0 recs

Wait.

There’s more than one guy in the world named Chone?

Kansas City Royals: your 2006 and 2007 NL Central champions!

by mazoboom on Oct 28, 2008 12:29 PM EDT reply actions 1 recs

+1,000,006

OMG Banny. FWIW I am only crdtng u w/3 runs allwd bc of DDJ OMFG

by devil_fingers on Oct 28, 2008 12:31 PM EDT up reply actions 0 recs

if you're kidding, ignore the following comment

His real name is Sean Smith, but sometimes goes by Chone online since he’s an Angels fan. And his projection system (which is insignificantly better than PECOTA) is called CHONE for a similar reason. Smart dude. Big fan.

Beyond the Boxscore // Calling BJ Upton lazy is lazy.

by Sky Kalkman on Oct 28, 2008 1:04 PM EDT up reply actions 0 recs

That's awesome

I should have figured that, since he’s an Angel’s fan…

Really, CHONE is as good/better than PECOTA? I thought that Zips was the best “free” one out there.

(CHONE is free access, right?)

Now, if I could just figure out how to get all this into Excel, I’ll never have to pay for fantasy advice again…

OMG Banny. FWIW I am only crdtng u w/3 runs allwd bc of DDJ OMFG

by devil_fingers on Oct 28, 2008 1:14 PM EDT up reply actions 0 recs

yes, free

but none of CHONE, ZIPS, THT attempt to project playing time, which is actually probably the more improtant piece for fantasy.

but yes, here’s a good article comparing projection systems for 2008 — past years are similar. basically, if you use Marcel as the benchmark (a simple three-year weighting of stats with some simple regression), none of the high-falutin’ systems are a huge improvement. and none of them are significantly better than the others. with a gun to my head i’d go with PECOTA for pitchers and CHONE for hitters.

Beyond the Boxscore // Calling BJ Upton lazy is lazy.

by Sky Kalkman on Oct 28, 2008 1:31 PM EDT up reply actions 0 recs

that link

http://www.insidethebook.com/ee/index.php/site/comments/evaluating_the_2008_forecasting_systems/

and i like this quote:

“"Chone, among these 4, is the clear leader. I hope no one tries to tout their system as the best, other than maybe Chone.—Tangotiger, Oct 7, 2008"”

Beyond the Boxscore // Calling BJ Upton lazy is lazy.

by Sky Kalkman on Oct 28, 2008 1:33 PM EDT up reply actions 0 recs

thanks

I guess I’d read an article that PECOTA was the best, then ZiPS. Maybe I’ve been reading too much insidethebook, but I’m at the point where I’m convinced that PECOTA is the only reason to pay for a BP subscription, and now I’m not sure what to do. But if if their fantasy stuff uses PECOTA, and its at least as good as Shandler’s forecasts, then that is the better investement.

OMG Banny. FWIW I am only crdtng u w/3 runs allwd bc of DDJ OMFG

by devil_fingers on Oct 28, 2008 1:40 PM EDT up reply actions 0 recs

ZIPS and PECOTA were the best according to last year's THT article

and if I remember, CHONE was not far behind. Apparently, CHONE did better this year.

by Gopherballs on Oct 28, 2008 7:20 PM EDT up reply actions 0 recs

PECOTA may cost more, but still has the best name

although now that I know the origins of “CHONE,” it’s a lot closer

OMG Banny. FWIW I am only crdtng u w/3 runs allwd bc of DDJ OMFG

by devil_fingers on Oct 28, 2008 9:58 PM EDT up reply actions 0 recs

Here's one sort of good thing about Gordon

Now, I personally think he will be better than -9 at 3B next year.

But even assuming that projection, it doesn’t mean he’s needs to move (aside from the fact that the ROyals don’t have anyone else to play third, especially with Teahen at -1) to an easier position.

The difference in “value” between a equal defenders as 3B (+.25 wins) and LF/RF (-.t5 wins) is one win over a season. Let’s call that 10 runs (it’s more like 10.5). With everything else being equal, then, if Gordon would be a league average defender at LF/RF, then he’s still worth about 1 run less than he is as a -9 at 3B.

I know that position shifts aren’t something done just with the positoinal adjustments and a calculator — it involves considering the other available personnel, and this is also an area where the judgments of scouts are obviously crucial. But thinking about realtive positional value in this way does keep things in perspective.

OMG Banny. FWIW I am only crdtng u w/3 runs allwd bc of DDJ OMFG

by devil_fingers on Oct 28, 2008 12:36 PM EDT reply actions 0 recs

It is not hard to imagine that Gordan would be a above average oufielder though.

Gordan runs pretty well, has a good arm, and has a good reaction time. If he is good at reading fly balls, he would be better value wise as an outfielder. I think Gordan is in the outfield in two years anyway when Moose is ready. Once that happens and we get rid of Guillen “the statue” and if we get Gootz then thats about 3-4 wins just for outfield defense, thats very nice.

Go Royals!

by BabyBlues on Oct 28, 2008 1:39 PM EDT up reply actions 0 recs

maybe (although the "Gootz" rumor was squashed pretty hard by DMGM)

I agree that it shows that Gordon is at the very least good enough to hold the position down until Moustakas is (hopefully) ready.

I still think Gordon is better than 9. It’s only two years of data. He was above average last year, and had a bad year this year. But some people think he looiked better after he was back form injury, and there’s some statistical evidence for that. It’s also interesting that part way through 2007, UZR had him at about -7/150, but he ended up at +4 - so maybe he’s a “second-half defensive performer” just like he is on offense, and this year would have ended up better. Assuming that first/second-half splits have predictive value (which is questionable).

OMG Banny. FWIW I am only crdtng u w/3 runs allwd bc of DDJ OMFG

by devil_fingers on Oct 28, 2008 1:43 PM EDT up reply actions 0 recs

dang cross out thing

should be

I still think Gordon is better than 9. It’s only two years of data. He was above average last year, and had a bad year this year. But some people think he looiked better after he was back form injury, and there’s some statistical evidence for that. It’s also interesting that part way through 2007, UZR had him at about -7/150, but he ended up at +4

OMG Banny. FWIW I am only crdtng u w/3 runs allwd bc of DDJ OMFG

by devil_fingers on Oct 28, 2008 1:44 PM EDT up reply actions 0 recs

better than -9

OMG Banny. FWIW I am only crdtng u w/3 runs allwd bc of DDJ OMFG

by devil_fingers on Oct 28, 2008 1:44 PM EDT up reply actions 0 recs

If Gordon is -10 runs as a 3B and adjusts to a corner spot like a typical player, he'd be league-average there

so if he’s better than average, it’s a good move (assuming he can’t improve at third, which is a big assumption.)

Ryan Braun’s a good example — he went from being at least -20 runs at third base to +5 in LF, a brilliant (if not obvious) move by the Brewers.

Beyond the Boxscore // Calling BJ Upton lazy is lazy.

by Sky Kalkman on Oct 28, 2008 1:45 PM EDT up reply actions 0 recs

I can't speak for others

but I have a lot of “sports emotion” dork-ely invested in Gordon. As Rany Jazayerli pout it earlier this year, he was touted of having the bat of Braun and the glvoe of Zimmerman, but sometimes it seems like he has the glvoe of Braun and the bat of Zimmerman. Sigh.

It’s just a clever line, of course. I don’t think he has the raw power of Braun or Zimmerman’s glove, but I think he already has better plate discipline than either, will have more isolated power than Zimmerman, and has a chance to at least stick at third.

His basestealing is underrated, too. 14 for 18 last year, 9 for 11 this year.

OMG Banny. FWIW I am only crdtng u w/3 runs allwd bc of DDJ OMFG

by devil_fingers on Oct 28, 2008 1:49 PM EDT up reply actions 0 recs

I think Braun's a fluke

the Brewers should trade him for Ramon Ramirez while his stock is high.

OMG Banny. FWIW I am only crdtng u w/3 runs allwd bc of DDJ OMFG

by devil_fingers on Oct 28, 2008 1:49 PM EDT up reply actions 0 recs

Defense=overrated

Accidentally not thedude925 anymore. I do hate this new name.

by wildthang on Oct 28, 2008 4:07 PM EDT reply actions 0 recs

that = wrong

the above-average fielding teams in 2008 by combined ZR and RZR:
PHI 51
OAK 49
STL 47
HOU 41
TOR 38
MIL 29
ATL 27
SD 26
TB 25
LAN 22
CHN 20
NYN 14
LAA 8
CLE 7
SF 3

You’ve got 6 of the 8 playoff teams, plus some competitive teams — TOR, STL, HOU — and Oakland was surprisingly good for a while until they traded everyone away.

The bottom half:
KC -57
NYA -56
MIN -42
TEX -41
BAL -38
BOS -31
FLA -27
PIT -25
CHA -25
ARI -23
SEA -15
COL -12
DET -11
CIN -8
WAS 0

You’ve got a bunch of bad teams, with a few decent ones. Hmm, I wonder why the Yankees missed the playoffs?

Beyond the Boxscore // Calling BJ Upton lazy is lazy.

by Sky Kalkman on Oct 28, 2008 4:35 PM EDT up reply actions 0 recs

Hmm, I wonder why the Yankees missed the playoffs?

The Curse of Clay Bellinger.

OMG Banny. FWIW I am only crdtng u w/3 runs allwd bc of DDJ OMFG

by devil_fingers on Oct 28, 2008 5:09 PM EDT up reply actions 0 recs

UPDATED

Responding to the suggestions of Sky and MGL, Chone made some changes to weight the relative positions of infielders as dealed in the linked thread. The changes to Royals infielders are reflected above — it hurt Esty, made Callaspo’s 2B projection 1 run better, and helped Aviles 2B and 3B projections significantly.

OMG Banny. FWIW I am only crdtng u w/3 runs allwd bc of DDJ OMFG

by devil_fingers on Oct 28, 2008 10:30 PM EDT reply actions 0 recs