Yeah, Jimmy upped it to 105

BuddyGroom wrote:And DS, don't be so sure Hamels won't notice Feliz' glove work. I suspect plenty of those 89-90 fastballs he tosses in there get ripped down the third-base line by righties.

RH hitters, against all pitchers over the last 5 years, hit groundballs 44% of the time. About 65% of those were from the shortstop to the 3B line, and batters hit a combined .202, with a .231 SLG, on those balls.

RH hitters, against Cole Hamels last year, hit groundballs 43% of the time. About 58% of those were from the shortstop to the 3B line, and batters hit a combined .211, with a .257 SLG, on those.

Given that he allowed 109 such groundballs, the Phillies 3B defense cost him 1 hit, a triple, in 2007.

He may well win the Cy Young next year, but it won't be because of Feliz.

Shore gives me wood

shore is not one man

Shore wrote:

BuddyGroom wrote:And DS, don't be so sure Hamels won't notice Feliz' glove work. I suspect plenty of those 89-90 fastballs he tosses in there get ripped down the third-base line by righties.

RH hitters, against all pitchers over the last 5 years, hit groundballs 44% of the time. About 65% of those were from the shortstop to the 3B line, and batters hit a combined .202, with a .231 SLG, on those balls.

RH hitters, against Cole Hamels last year, hit groundballs 43% of the time. About 58% of those were from the shortstop to the 3B line, and batters hit a combined .211, with a .257 SLG, on those.

Given that he allowed 109 such groundballs, the Phillies 3B defense cost him 1 hit, a triple, in 2007.

He may well win the Cy Young next year, but it won't be because of Feliz.

Feliz has the voters in his pocket though.

dear shore and matts --

it seems pretty clear that you guys reject the current state of the art best estimates of defensive value. as phillychuck has pointed out, the defensive runs saved above replacement estimates for feliz are consistent and large. in the neighborhood of 20 -30 runs or 2-3 wins. and helms and dobbs are consistently horrible, something like the opposite of feliz over a whole season without nunez as caddy, negative 1-2 wins. so being 3-5 wins better on defense is consistent with state of the art defensive assessment. as is being only a win worse than ensberg overall in 2006.

look, i know that defensive assessment is an uncertain art, but if you are going to insist on an estimate of value much smaller than those of the smartest guys out there, it seems like you should present more of an argument about why those other estimators are wrong.

Shore wrote:

BuddyGroom wrote:And DS, don't be so sure Hamels won't notice Feliz' glove work. I suspect plenty of those 89-90 fastballs he tosses in there get ripped down the third-base line by righties.

RH hitters, against all pitchers over the last 5 years, hit groundballs 44% of the time. About 65% of those were from the shortstop to the 3B line, and batters hit a combined .202, with a .231 SLG, on those balls.

RH hitters, against Cole Hamels last year, hit groundballs 43% of the time. About 58% of those were from the shortstop to the 3B line, and batters hit a combined .211, with a .257 SLG, on those.

Given that he allowed 109 such groundballs, the Phillies 3B defense cost him 1 hit, a triple, in 2007.

He may well win the Cy Young next year, but it won't be because of Feliz.

And the Phillies made the playoffs by a single game, which easily could have been swayed by a hard grounder down the third base line turning into an extra-base hit.

phorever wrote:dear shore and matts --

it seems pretty clear that you guys reject the current state of the art best estimates of defensive value. as phillychuck has pointed out, the defensive runs saved above replacement estimates for feliz are consistent and large. in the neighborhood of 20 -30 runs or 2-3 wins. and helms and dobbs are consistently horrible, something like the opposite of feliz over a whole season without nunez as caddy, negative 1-2 wins. so being 3-5 wins better on defense is consistent with state of the art defensive assessment. as is being only a win worse than ensberg overall in 2006.

look, i know that defensive assessment is an uncertain art, but if you are going to insist on an estimate of value much smaller than those of the smartest guys out there, it seems like you should present more of an argument about why those other estimators are wrong.

helms and dobbs last year each had terrible defensive years even for them. they were 6 runs below replacement. feliz had his best defensive year ever, 27 runs above replacement. even if you figure nunez played 1/4 of the team and the helms/dobbs pairing would be 8 runs below replacement, that's still 35 runs last year. but that's not 5 wins. it's 3.5. and that is probably not going to be repeated. i'm guessing regression and progression to the mean puts them closer to 2 wins different, maybe 3.

offensively, we need to get into our heads that third base would not have gone as badly this year as it did. helms was going to hit much better, in all likelihood. every projection system has said so.

and furthermore, i'm sick of OPS being a statistic that you use to evaluate Pedro Feliz. It's a fine statistic, but it should be well known that it understates the value of OBP and overstates the value of SLG. That's fine for evaluating hitters who are proportionately good or bad at both, but for a guy like Feliz, it's just ridiculous. use his VORP-- which is negative the past three seasons.

all in all, the difference offensively between Dobbs and Helms (probably 1.5 wins above replacement level) and Feliz (probably 0.5 wins below replacement level), is a drop in hitting of about 2 wins, maybe 1.5 if you take into account the addition of Dobbs and Helms to the bench.

So in reality, the difference is probably positive but less than a win. I'm sick of 5 wins being thrown around. If you sent Ryan Howard to third, you might not get a five win difference between him and Feliz.

I'm not dismissing the value of defensive statistics-- I'm just highlighting that this is less than a one-win difference. Signing Kyle Lohse would probably be a two-win difference. Now it seems that doing so would put the team around $110MM instead of $105MM which might make them unwilling to do so.

Furthermore, I want to point something else out-- Feliz being a bad hitter does more damage to a team like the Phillies than other teams. It means that he will cost them more runs by getting outs and double plays with runners on base more often and not getting on base at the bottom of the lineup as much before our .500 slugging leadoff hitter gets up. my suspicion is that the offensive difference may be larger than we're guessing using numbers like VORP which effectively discuss the value of adding a player to an average offense.

phorever wrote:dear shore and matts --

it seems pretty clear that you guys reject the current state of the art best estimates of defensive value. as phillychuck has pointed out, the defensive runs saved above replacement estimates for feliz are consistent and large. in the neighborhood of 20 -30 runs or 2-3 wins. and helms and dobbs are consistently horrible, something like the opposite of feliz over a whole season without nunez as caddy, negative 1-2 wins. so being 3-5 wins better on defense is consistent with state of the art defensive assessment. as is being only a win worse than ensberg overall in 2006.

look, i know that defensive assessment is an uncertain art, but if you are going to insist on an estimate of value much smaller than those of the smartest guys out there, it seems like you should present more of an argument about why those other estimators are wrong.

I apologize if I've run over your dog, phorever.

Estimates are great; I've provided - for Hamels, at least - hard data to suggest that defense on that side of the infield was not a problem, relative to the league. That's based on play-by-play data, and I used Dan Fox's BIP Charts to find it (http://danagonistes.blogspot.com/).

Same guy has created Simple Fielding Runs (SFR), which were discussed at BP last week (http://www.baseballprospectus.com/article.php_articleid=7072. Via SFR, which combines "traditional" defensive stats with Ball-in-Play data, Pedro Feliz was, in fact, the best defensive 3B in the league last year... at 15 runs. Abe Nunez, in limited time, was worth 4, and Wes Helms 0 (well, 0.1). Dobbs was a disaster at -6.7.

2007 was Feliz's best year of the last 3, as his SFR were only 9.6 and 5.2 in 2006 and 2005. So, over the last 3 years, Feliz has been worth, defensively, 30 runs (by this system). That's 3 wins, in 3 years.

Wes Helms hovers around 0 - .1, -.9, 1.3. Dobbs has last year's -6.7, and then 2 year right at 0 due to limited PT.

All of which would seem to indicate that we'd be around a -10 if we'd played Dobbs, Helms, and a caddy like Bruntlett (who's also about 0). And Feliz would be about +10, giving us 20 runs difference, or 2 wins, defensively. Given that Dobbs/Helms is likely to be a win better offensively, we'll call Feliz a 1-win better option than them.

I, however, have argue that Morgan Ensberg was a better option. Using SFR again, Ensberg's last 3 years are -9.5, 11.7, and 10.8. Better than Feliz 2 of the last 3 years, and significantly worse last year. Still, a +13 over 3 years, about 2/3 of a win per year behind Feliz. I believe him to be at least 2 wins better offensively, putting him more than a win better overall than Feliz, and more than 2 better than our current situation. Obviously, one would have to identify the cause of the precipitous drop in both defense and offense for Ensberg, but assuming injury, it seems like he's a far more solid choice - at least a wash, potentially superior on both offense and defense.

interestingly, this pretty much sets the 2009 roster on the hitting side pretty clearly.

Ruiz
Howard
Utley
Feliz
Rollins
<LEFTFIELDER>
Victorino
Jenkins

Coste or Jamarillo
Dobbs
Bruntlett
Werth
<5TH OUTFIELDER>

If we resign Burrell and pick up Taguchi's option, that pretty much sets the hitting as is.

The pitching side of the equation is a little more open:

Hamels
Myers
Kendrick
Eaton
<STARTER>

<CLOSER>
Madson
Romero
<4 MORE RELIEVERS>


There are probably going to be raises for Howard, Utley, Victorino, Werth, Hamels, and Madson, and clearly 8 more roster spots to fill. Taking into account the Thome money, this seems to be about $85-90MM of 2009's payroll already spent.

BuddyGroom wrote:

Shore wrote:

BuddyGroom wrote:And DS, don't be so sure Hamels won't notice Feliz' glove work. I suspect plenty of those 89-90 fastballs he tosses in there get ripped down the third-base line by righties.

RH hitters, against all pitchers over the last 5 years, hit groundballs 44% of the time. About 65% of those were from the shortstop to the 3B line, and batters hit a combined .202, with a .231 SLG, on those balls.

RH hitters, against Cole Hamels last year, hit groundballs 43% of the time. About 58% of those were from the shortstop to the 3B line, and batters hit a combined .211, with a .257 SLG, on those.

Given that he allowed 109 such groundballs, the Phillies 3B defense cost him 1 hit, a triple, in 2007.

He may well win the Cy Young next year, but it won't be because of Feliz.

And the Phillies made the playoffs by a single game, which easily could have been swayed by a hard grounder down the third base line turning into an extra-base hit.

This is where I say the Phillies made the playoffs by a single game, and by adding Pedro Feliz to the lineup, we've taken 2 wins off of the team. Guess where that puts us.

One thing no one mentioned is Greg Dobbs' offense. Dobbs had the perplexing split of a .641 OPS as a third baseman and a .977 OPS when he wasn't. He had significant plate appearances in both capacities. He was just that much better when he didn't play third.

In 2006 he hit in a minimal numbers of at bats and didn't come to bat as a third baseman. In 2005, with 142 ABs, he didn't hit but played only 2 games at third. In 2004 he played pretty much exclusively at third and didn't hit.

Most people chalk this up to a fluke. I can't think of another player who could hit at one position and not at another. We do know Dobbs struggles as a third baseman. What if his play at third affects his concentration at the plate? It probably isn't the case but his poor OPS at third was a concern.

It seems to me that if Feliz is a 1-1.5 win added player, 4.25 mil is a good value for a team in the Phil's position. Obviously 1 year would have been far better but the deal is cheap enough that it's very hard for me to complain.

Laexile wrote:One thing no one mentioned is Greg Dobbs' offense. Dobbs had the perplexing split of a .641 OPS as a third baseman and a .977 OPS when he wasn't. He had significant plate appearances in both capacities. He was just that much better when he didn't play third.

In 2006 he hit in a minimal numbers of at bats and didn't come to bat as a third baseman. In 2005, with 142 ABs, he didn't hit but played only 2 games at third. In 2004 he played pretty much exclusively at third and didn't hit.

Most people chalk this up to a fluke. I can't think of another player who could hit at one position and not at another. We do know Dobbs struggles as a third baseman. What if his play at third affects his concentration at the plate? It probably isn't the case but his poor OPS at third was a concern.

those differences are remotely statistically significant. he had 48 at-bats as a pinch-hitter. the batting average is only 1.25 standard deviations above what you'd expect and the on-base percentage is only 0.87 standard deviations above what you'd expect. that's just not a big difference.

i agree with you that may have factored into their decision making, but it's proposterous if it did.

regardless, this is not a signing that has a huge effect on anything.

Shore wrote:
I apologize if I've run over your dog, phorever.

Estimates are great; I've provided - for Hamels, at least - hard data to suggest that defense on that side of the infield was not a problem, relative to the league. That's based on play-by-play data, and I used Dan Fox's BIP Charts to find it (http://danagonistes.blogspot.com/).

Same guy has created Simple Fielding Runs (SFR), which were discussed at BP last week (http://www.baseballprospectus.com/article.php_articleid=7072. Via SFR, which combines "traditional" defensive stats with Ball-in-Play data, Pedro Feliz was, in fact, the best defensive 3B in the league last year... at 15 runs. Abe Nunez, in limited time, was worth 4, and Wes Helms 0 (well, 0.1). Dobbs was a disaster at -6.7.

2007 was Feliz's best year of the last 3, as his SFR were only 9.6 and 5.2 in 2006 and 2005. So, over the last 3 years, Feliz has been worth, defensively, 30 runs (by this system). That's 3 wins, in 3 years.

Wes Helms hovers around 0 - .1, -.9, 1.3. Dobbs has last year's -6.7, and then 2 year right at 0 due to limited PT.

All of which would seem to indicate that we'd be around a -10 if we'd played Dobbs, Helms, and a caddy like Bruntlett (who's also about 0). And Feliz would be about +10, giving us 20 runs difference, or 2 wins, defensively. Given that Dobbs/Helms is likely to be a win better offensively, we'll call Feliz a 1-win better option than them.

I, however, have argue that Morgan Ensberg was a better option. Using SFR again, Ensberg's last 3 years are -9.5, 11.7, and 10.8. Better than Feliz 2 of the last 3 years, and significantly worse last year. Still, a +13 over 3 years, about 2/3 of a win per year behind Feliz. I believe him to be at least 2 wins better offensively, putting him more than a win better overall than Feliz, and more than 2 better than our current situation. Obviously, one would have to identify the cause of the precipitous drop in both defense and offense for Ensberg, but assuming injury, it seems like he's a far more solid choice - at least a wash, potentially superior on both offense and defense.

thank you... that's what i was looking for. you should have figured out by now that i can't resist holding people who really know and love their stats to a higher standard. maybe hypocritical of me, since i save my own hard-core stats work for my job and and pretty sloppy and lazy with my baseball stuff. it is great to be able to take advantage of someone like you who will find an answer for me if i bug you enough. and when something you assert seems to fly in the face of other data points... that seems like a good time to bug you.

now, do you have a feeling for why sfr might be better than the measures that give the higher defensive run values for feliz? i'm must say that i too am inclined to believe the lower number (now that there is more than one of them). my instincts say that several of the current systems have some instabilities at the high and low end of their scales... in other words, the distributions of defensive run values look heavy-tailed and i suspect the true distribution to be cconsiderably more normal. is sfr explicitly addressing that problem (sorry... i'm not only lazy, but cheap... i don't have full access to bp).


about ensberg... what you you make of his unstable walk rates? losing power and plate discipline is kinda unusual. looks to me more like a vision problem than any other injury. if i was considering a contract for him, i would try hard to talk him into an eye test, and see if it's something lasik can fix. i'm really surprised that no one seems to want him despite the rebound in the second half of the season. that's what makes me think there is more to the situation than meets the eye... maybe something known to be wrong with >his< eyes that lasik can't fix. or is it really the case that every single team that needs a 3b is too stupid or broke to take note of ensberg's continued availability?

as far as the dobbs sucking offensively when at third base thing, that probably had more to do with coincidence... he rarely played third in the beginning of the season because helms was penciled in as the starter, and that's when he happened to be at his hottest for the whole year. as the year went on, his hitting performance diminished and he got more starts at third out of necessity, but i doubt the two were related.

MattS wrote:
So in reality, the difference is probably positive but less than a win. I'm sick of 5 wins being thrown around. If you sent Ryan Howard to third, you might not get a five win difference between him and Feliz.

I'm not dismissing the value of defensive statistics-- I'm just highlighting that this is less than a one-win difference. Signing Kyle Lohse would probably be a two-win difference. Now it seems that doing so would put the team around $110MM instead of $105MM which might make them unwilling to do so.

Furthermore, I want to point something else out-- Feliz being a bad hitter does more damage to a team like the Phillies than other teams. It means that he will cost them more runs by getting outs and double plays with runners on base more often and not getting on base at the bottom of the lineup as much before our .500 slugging leadoff hitter gets up. my suspicion is that the offensive difference may be larger than we're guessing using numbers like VORP which effectively discuss the value of adding a player to an average offense.

i get your point. i'm still willing to put up with claims of a 4-5 win improvement. feliz will also take fielding ops away from the fielding-neutral bruntlett, and helms and dobbs 2007 3b performances weren't out of line with their career numbers from what i can see, and nunez appears to have had a lot more than 1/4 of the 3b chances. so 4-5 wins on defense is plausible. helms 2008 was truly flukish and i think it's inflating his projections. and dobbs is almost certain to decline significantly. i can easily see those two plus the nunez-esque bruntlett putting up combined offensive numbers as bad or worse than feliz.

yeah that's a high-side estimate, but more like 1 standard deviation high than 3-standard deviation high.

your page 13 recap: Feliz still suck-spicious

I think I'm going to lose a friend over this Pedro Feliz signing.

I am thinking that all of the variables in this, mostly of the 'what-if nature, that we will not know until July if we should greet him or beat him. Good thing that it is Philly we are talking about, so if the suck factor is high, the explosion will occur quickly.

The talk of 'great fielding' and 'offensive prowess' might make some loon try to go the MJS route to show, what we all know, the he's no MJS.