ESPN's Neel writes article on how defensive stats, not economy shaped free agency
Eric Neel of ESPN writes article noting how clubs are using Baseball Prospectus and UZR when looking at Free Agents and how the defensive measurements shaped free agency this offseason.
Nothing new for anyone already converted, but good reading for the rest of us.
7 months ago
N Oakley
74 comments
2 recs |
Comments
Gut feel - Soriano defense - an asset or liability?
The Cubs’ Alfonso Soriano, surprisingly enough, has the highest rating among leftfielders over the past three years. He got to just three more balls than the average guy, but because of his strong arm, he saved his team 42 runs.
There was a recent Fandshot where this stat is was discussed in great length. However, the more I see it trickle through several other media outlets just makes my blood boil. I don’t watch the games with a calculater, but I certainly have seen a ton of good and bad outfield play. I love to see him throw somebody out at the plate, but that does not compensate for balls he doesn’t get to, drops outright, misplays, or when he bunny-hops himself to the DL.
Soriano is a defensive liability – period.
by BatCubFan on Apr 1, 2009 8:40 AM CDT reply actions 0 recs
So you dont believe in logic?
Okay, just so I understand it... in your wildest fantasy, you are in hell. And you are co-running a bed and breakfast with the devil.
by bren on Apr 1, 2009 8:43 AM CDT up reply actions 0 recs
Do not confuse stats with logic
I believe what I see, not what some algorithm tells me. If you think “logic” says Sori is a plus defender than why does Lou replace him defensively in late innings?
Soriano is a defensive liability – period.
by BatCubFan on Apr 1, 2009 8:52 AM CDT up reply actions 0 recs
Lou DOESN'T replace him, usually
Although I think he should, whether he does or not gives little support to your argument,
Facts are meaningless. You could use facts to prove anything that's even remotely true! --Homer J. Simpson
by Shanghai Badger on Apr 1, 2009 9:39 AM CDT up reply actions 0 recs
Precisely.
I’m torn on the issue. I’ll admit that, emotionally, I’d rather see RJ play left field in late innings. But Soriano’s arm could literally save games by holding runners on or gunning them down at the plate.
"That little kid at second base - he is after a job, isn't he?" ~ Lou Piniella, 3/9/09
by dat cubfan daver on Apr 1, 2009 9:49 AM CDT up reply actions 0 recs
funny how your eye deceives you
this is why disciplined clinical trials or objective statistics or why math is called the universal language. Soriano as a left fielder only has not gotten to 3 balls against the league average (meaning in MLB they hide many a bad fielders there) but his arm has saved 42 runs….on par with Crawford who is a good fielder and has a good arm.
The numbers don’t lie. This is why the courts are fooled with eye witness testimony where DNA says that guy didn’t rape that woman even though she swears he did, or others who say that person ran a red light while the camera says no that didn’t happen.
kudos, I thought all along Soriano was an asset in LF. And I think he will get better.
Piniella: "This is a tougher job than I thought it would be, I'm going to be honest with you."
by Ivy Walls on Apr 1, 2009 9:55 AM CDT up reply actions 0 recs
Well Ivy,
I think you cross the line twice calling Soriano an asset in LF and using rape metaphors.
As for the Soriano stats, I have a few questions for anyone expert on the “get to balls” measurement. How are dropped balls and pulling up because of the OF wall measured in whether he got to the ball. With a drop, he was there. If he pulls up on the track, he was there. Do those count as gotten to the ball?
if this was still new to me, i wouldn't understand
by N Oakley on Apr 1, 2009 10:00 AM CDT up reply actions 0 recs
sorry about the offense but that is reality
Let me say I knew someone accused of rape by a eye witness and then even with evidence of an alibi took until DNA proved he was not the guy.
Soriano in the relative world is an average “get to” left fielder, he is faster than most, although his jump is less than desired. His arm with 42 put outs is fantastic.
What has happened is a group attitude towards Fonzi that he is not a good OF’er, he is an average one but blessed with a cannon, an accurate one at that. What is Fonzi’s problem is he is paid so much for so long.
Piniella: "This is a tougher job than I thought it would be, I'm going to be honest with you."
by Ivy Walls on Apr 1, 2009 2:39 PM CDT up reply actions 0 recs
The numbers don't lie, they just ...
… aren’t sophisticated enough to take in all the real world variables. Yes, for what they measure they generate an interesting, useful for comparison number. But, the stats are limited in their scope, and do not have ability to account for “what your eyes” are telling you. Yes sometimes people put on blinders to statistics and make bad decisions, but just as many blindly trust the numbers and use them to make bad decisions.
Bottom line — 7th game WS, 9th inning, we have the lead — I’m pulling Sori for defensive replacement — statistics be damned!
by BatCubFan on Apr 1, 2009 11:07 AM CDT up reply actions 0 recs
I'm hopeful someone with better knowledge of the
formulas can help with my question above as to what counts as getting to a ball. Statistics don’t lie, but the the data entered can’t be suspect.
if this was still new to me, i wouldn't understand
by N Oakley on Apr 1, 2009 11:10 AM CDT up reply actions 0 recs
its like making sausage
The reality is ugly.
by ol Pete on Apr 1, 2009 12:57 PM CDT up reply actions 0 recs
Wait, wait
Fonz is the best LF by statistical measurement because of lips and assholes?
if this was still new to me, i wouldn't understand
by N Oakley on Apr 1, 2009 1:47 PM CDT up reply actions 0 recs
or the casing he is in
just as long as Randall Simon doesnt bop him on the head with a bat
Friendship is like peeing on yourself: everyone can see it, but only you get the warm feeling that it brings.
by Cubbie-Tim on Apr 1, 2009 1:55 PM CDT up reply actions 0 recs
Strange choice
Because if there is a runner on second and a base hit to left field, Soriano will gun his ass down, Reed Mantle won’t.
www.talkingchicagobaseball.blogspot.com
by nji232 on Apr 1, 2009 12:12 PM CDT up reply actions 0 recs
Clinical trials and the data they produce are often contentious for years at a time. They are done scientifically and are evaluated in the open by professional statisticians, all of which is in contrast to the production of the stats you refer to. Even after all that, the conclusions turn out to be erroneous.
I have no idea what on objective statistic is. There is no way a proprietary sales products like UZR or PMR could be referred to as objective.
by ol Pete on Apr 1, 2009 1:03 PM CDT up reply actions 0 recs
being that I was in the field of stat analysis and data mgmt in pharma
I will tell you that they are corrupt in how they manage the data sets as to what or whom are in the protocols and who are not, after the trials begin.
The contentious rates they produce are because the sample sizes are too small and within prescribed data sets so to produce the best results.
Best example was a quinolone, that was meant heavy pneumonia patients, the drug was tested only in data sets under 45 years old but when it hit the market it was given as last resort to those over 70 and it shut down vitals in a huge way. Many deaths. The doc’s prescribed out of its unrealistic protocol.
Recently I see that a doc who studied Celebrex is now up for false data.
The issue is not the science of clinical trials it is the execution of it.
Piniella: "This is a tougher job than I thought it would be, I'm going to be honest with you."
by Ivy Walls on Apr 1, 2009 2:47 PM CDT up reply actions 0 recs
LOL
"I've got an idea...an idea so smart my head would explode if I even began to know what I was talking about." ~Peter Griffin
by Goodie1969 on Apr 1, 2009 6:31 PM CDT up reply actions 0 recs
you're flat out wrong
about the contentiousness being produced by small sample sizes in clinical trials.
The issue is the quality of the attempts to measure defense and quantify the results. Do you know how Dewan and BIS produce their products?
by ol Pete on Apr 1, 2009 8:07 PM CDT up reply actions 0 recs
Theres an objective measurement that says he's one of the best LF in the league
the past three years and you ignore them b/c you “feel” or “think” or that “he looks like” a poor LF, when the stats prove he is an above avg LF, and his strong arm makes up for his lack of range, which is still above avg. anyway
Okay, just so I understand it... in your wildest fantasy, you are in hell. And you are co-running a bed and breakfast with the devil.
by bren on Apr 1, 2009 10:09 AM CDT up reply actions 0 recs
Correct, but I believe people keep
taking “above average LF” and translating that to good fielder.
if this was still new to me, i wouldn't understand
by N Oakley on Apr 1, 2009 10:32 AM CDT up reply actions 0 recs
Maybe try
best left fielder in the league? Does that mean good fielder?
"Manny Trillo is coming in to pinch run. You know, for a lot of teams, you would pinch run for Manny Trillo." - Harry Caray
by Archie on Apr 1, 2009 10:44 AM CDT up reply actions 0 recs
Depends.
How does the best LF compare to the worst RF, average RF, etc.?
if this was still new to me, i wouldn't understand
by N Oakley on Apr 1, 2009 11:03 AM CDT up reply actions 0 recs
Thats irrelevant
Hes a LF what he does respective to RF doesnt matter. I think everyone is surprised at how good Sorianos defensive numbers are, partly b/c of his lousy defensive reputation at second and his reluctance to move to LF to begin with, but the numbers show he’s amongst the best LF in the league,he might not be the most graceful, but he saves runs and gets to more balls than the avg left fielder.
But should we be surprised, he’s enormously athletic
Okay, just so I understand it... in your wildest fantasy, you are in hell. And you are co-running a bed and breakfast with the devil.
by bren on Apr 1, 2009 12:14 PM CDT up reply actions 0 recs
In comparison of Left Fielders, how he compares with
a RF is irrelevant. When trying to determine if he’s a good fielder, how he compares with players at the other corner position, the one deemed for better fielders, seems relevant to me.
if this was still new to me, i wouldn't understand
by N Oakley on Apr 1, 2009 1:49 PM CDT up reply actions 0 recs
I dont think so
b/c he doesnt play RF, so its not a proper comparison
Okay, just so I understand it... in your wildest fantasy, you are in hell. And you are co-running a bed and breakfast with the devil.
by bren on Apr 1, 2009 3:22 PM CDT up reply actions 0 recs
I understand.
I just wish someone with more knowledge could answer my question above on what constitues getting to the ball. Soriano is very athletic and has the ability to cover a lot of ground for a left fielder.
I want to know if a ball dropped out of the glove is “getting to the ball” or not.
if this was still new to me, i wouldn't understand
by N Oakley on Apr 1, 2009 3:34 PM CDT up reply actions 0 recs
“Getting to the ball” is how the results are referred to in conversations like this one. The measurement is “outs” and “expected outs.” If a fielder covers an enormous amount of ground and gets his glove on the ball, that’s not an expected out. When a ball is dropped by the fielder in the zones assigned to that fielder he doesn’t make an out – so it depends. One of the many critical flaws in the metric is that they assign static zones. If Sori moves way over to center in one of those extreme shifts and the ball lands in LF, he is considered to have failed although he didn’t fail. A team that shifts a lot will rate more poorly defensively than one that doesn’t although it isn’t necessarily so. (sausage making)
by ol Pete on Apr 1, 2009 8:15 PM CDT up reply actions 0 recs
Only a team that shifts poorly.
IOW, if you have a lot of cases where a ball is droping in and a fielder is being inappropriately charged with responsibility for a ball due to a shift, then the team is simply doing poorly with the shifts.
In a total value metric, that’s useful information. A team that shifts well won’t be overly penalized in the ratings, because they’ll catch more balls than the other team, and the balls they catch due to the shift have a higher value than balls caught in zone anyway.
by cwyers on Apr 2, 2009 12:20 AM CDT up reply actions 0 recs
Thank you Colin.
All the questions that come to mind are directly related to whether routine balls dropped by Soriano trigger a positive or negative result in the statistical model. If he gets a glove on the ball and it tallies an expected out, and an uptick in his measurement, then we may have a problem.
I don’t believe Soriano is a horrid left fielder, given his peers, but I cannot accept that by being decent for a left fielder, he deserves to be qualified as good in the field.
if this was still new to me, i wouldn't understand
by N Oakley on Apr 2, 2009 9:44 AM CDT up reply actions 0 recs
Not everything Lou does is correct
Some of his moves still seem rooted in old thinking
Okay, just so I understand it... in your wildest fantasy, you are in hell. And you are co-running a bed and breakfast with the devil.
by bren on Apr 1, 2009 10:10 AM CDT up reply actions 0 recs
Agreed, but
Left Field is where defensive liabilities reside in the NL. I can’t imagine how the Fonz finishes ahead of Crawford, but I could see Soriano being in the top five of leftfielders when his throws to home are included.
Being compared to the likes of Abreau, Burrell, Duncan, Dunn, etc. allows Fonz to move up the list.
if this was still new to me, i wouldn't understand
by N Oakley on Apr 1, 2009 9:17 AM CDT up reply actions 0 recs
I don't necessarily agree...
…but we’ve been all through this. As awkward as Soriano may look in left field, he has adequate range and a killer arm. All things considered, he’s average to slightly above average in left field.
"That little kid at second base - he is after a job, isn't he?" ~ Lou Piniella, 3/9/09
by dat cubfan daver on Apr 1, 2009 9:48 AM CDT up reply actions 0 recs
I didn't post this to rehash
the great debate “Soriano, not a great fielder, but good for a left fielder.”
I’m sure there are books on similar comparisons, “Handsome Boxers in their Fifties” and “Good Looking for a Nun.”
if this was still new to me, i wouldn't understand
by N Oakley on Apr 1, 2009 9:52 AM CDT up reply actions 0 recs
Oh, I know.
I think this is a good and very timely article. If memory serves, several people on that Soriano Fanshot scoffed at the idea that teams were actually using defensive metrics to help them decide whether to acquire a player. Neel makes a pretty strong case that they are.
"That little kid at second base - he is after a job, isn't he?" ~ Lou Piniella, 3/9/09
by dat cubfan daver on Apr 1, 2009 10:06 AM CDT up reply actions 0 recs
+1
I love to see him throw somebody out at the plate, but that does not compensate for balls he doesn’t get to, drops outright, misplays, or when he bunny-hops himself to the DL.
and there is a reason he throws so many out, base runners challange him more often, since they know there is a decent chance he misplays the ball
Friendship is like peeing on yourself: everyone can see it, but only you get the warm feeling that it brings.
by Cubbie-Tim on Apr 1, 2009 11:20 AM CDT up reply actions 0 recs
-1
Throwing someone out at the plate or at third is worth much more, double, triple, even quadruple, giving up a single you shouldnt have. A man on first scores anywhere from 10-40% of the time depending on how many outs there are. A man on third scores anywhere from 30-85% of the time. And getting a guy at the plate literally saves 1 full run. Last year Fonzie had 5 errors and 10 assists. That is a huge net positive because of how many more runs an outfield assist saves even accounting for non-error errors that dont show up in the stats. Fonzie would need closer to 30-40 errors versus 10 assists for it just to be a wash, and for him to be an average outfielder (which probably still makes him an aboves average left fielder)
Reed Ballgame - best CF in the MLB
by californiachicagoan on Apr 1, 2009 1:37 PM CDT up reply actions 2 recs
Yes. Rec'd.
"That little kid at second base - he is after a job, isn't he?" ~ Lou Piniella, 3/9/09
by dat cubfan daver on Apr 1, 2009 1:53 PM CDT up reply actions 0 recs
I am not saying that it is not important
I am saying that many of the runners he guns down would not be running against other LFers
Friendship is like peeing on yourself: everyone can see it, but only you get the warm feeling that it brings.
by Cubbie-Tim on Apr 1, 2009 1:56 PM CDT up reply actions 0 recs
Well, if that's the case...
…maybe we should give Soriano points for strategery. Perhaps he deliberately bobbles a few balls every week or so just to make runners think he sucks at fielding just so he can gun their sorry asses down at the plate. See what he does there? You don’t get a Hummer like that being stupid, y’know.
"That little kid at second base - he is after a job, isn't he?" ~ Lou Piniella, 3/9/09
by dat cubfan daver on Apr 1, 2009 2:04 PM CDT up reply actions 0 recs
We're talkin' about a car, right?
if this was still new to me, i wouldn't understand
by N Oakley on Apr 1, 2009 2:20 PM CDT up reply actions 0 recs
Either way, really.
"That little kid at second base - he is after a job, isn't he?" ~ Lou Piniella, 3/9/09
by dat cubfan daver on Apr 1, 2009 2:23 PM CDT up reply actions 0 recs
HA!
Friendship is like peeing on yourself: everyone can see it, but only you get the warm feeling that it brings.
by Cubbie-Tim on Apr 1, 2009 3:46 PM CDT up reply actions 0 recs
I do understand your point
and I hope mine is clear as well. I am not saying it is good or bad that more runners seem to test him, and that he seems to be able to stop the majority. I would love to see actual stats to see how many runners are held or sent against carious OFers in general, and see where he ranks in that.
Friendship is like peeing on yourself: everyone can see it, but only you get the warm feeling that it brings.
by Cubbie-Tim on Apr 1, 2009 3:47 PM CDT up reply actions 0 recs
carious = various
Friendship is like peeing on yourself: everyone can see it, but only you get the warm feeling that it brings.
by Cubbie-Tim on Apr 1, 2009 7:35 PM CDT up reply actions 0 recs
I’ve seen that. IIRC he rates very high.
by ol Pete on Apr 1, 2009 8:20 PM CDT up reply actions 0 recs
So, lemme ask you.
How many runners does Soriano throw out?
How many runners hold up, rather than trying to stretch a base hit, because they are afraid of being thrown out by Soriano?
How does this compare to the average left fielder?
How many balls does he fail to get to?
How many balls does he drop?
How many balls does he misplay?
How does this compare to the average left fielder?
by cwyers on Apr 2, 2009 12:30 AM CDT up reply actions 0 recs
comes down to trust
There are two unambiguous metrics in your list. Thrown out and drops, you could add errors if you want. All other defensive metrics have a degree of subjectivity to them. Now, I think there is value in capturing the data, they are useful for comparison purposes, and for generating discussion.
However, the metrics should be just one channel for decision making. Whether you manage in sports or business, you develop an intangible level of trust in a person’s ability to perform. Regardless, of how many people Sori throws out, or what Chone says, I don’t trust him in the field. I don’t think Lou does either. In most cases, his offensive potential cancels out that risk. However, in a big game with a lead, my strategy would be to put in a late inning defensive replacement.
by BatCubFan on Apr 2, 2009 9:40 AM CDT up reply actions 2 recs
Ok, I need a little education here
here is a quote from the article:
“Over the past three seasons, Abreu has made plays on 28 fewer balls than a league-average rightfielder would be expected to make, costing his teams (the Phillies and Yanks) 19 runs.”
So did Abreu cost his teams 19 runs over the three year period, or is the 28 fewer balls reached and 19 runs per year?
"Enough foreplay- let's get crackin'"- Fred Garvin
by davidalanu on Apr 1, 2009 8:55 AM CDT reply actions 0 recs
I am no expert, but
I believe the 28 fewer balls when compared with average cost 19 runs per year.
if this was still new to me, i wouldn't understand
by N Oakley on Apr 1, 2009 8:56 AM CDT up reply actions 0 recs
That seems like a lot
Without reading the article, the excerpt reads to me like it’s a total over the three years.
Facts are meaningless. You could use facts to prove anything that's even remotely true! --Homer J. Simpson
by Shanghai Badger on Apr 1, 2009 9:40 AM CDT up reply actions 0 recs
likely neither
but the author is using his poor writing to indicate over three years. Twenty-eight balls over three years is probably within the margin of error. In reality Abreu’s defense could have saved his team runs over the three years.
by ol Pete on Apr 1, 2009 1:17 PM CDT up reply actions 0 recs
You are correct - over the three year period.
if this was still new to me, i wouldn't understand
by N Oakley on Apr 1, 2009 9:48 AM CDT reply actions 0 recs
F*&%!!!
Reply to Badger.
if this was still new to me, i wouldn't understand
by N Oakley on Apr 1, 2009 9:49 AM CDT up reply actions 0 recs
If that's the case
then I have to wonder what all of the fuss is about. If a player like Dunn costs his team ~6-8 runs/year defensively, doesn’t his 100 points of ops above league average more than make up for that? Like way more than make up for it?
"Enough foreplay- let's get crackin'"- Fred Garvin
by davidalanu on Apr 1, 2009 10:00 AM CDT up reply actions 0 recs
I'm the last person to ask, but wouldn't
the offensive measurements be compared added in. The article compares Abreau with DeJesus and how the offensive and defensive together have DeJusus as more valuable.
if this was still new to me, i wouldn't understand
by N Oakley on Apr 1, 2009 10:02 AM CDT up reply actions 0 recs
I guess I found this article interesting with the
increased cites from Cwyers and Josh77 on defensive matrices and Daver’s recent jabs with UZR.
I thought through FA that the lack of big contracts to sluggers was strictly the economy. Any usage of other concepts in valuing players is good for baseball. Well good for baseball if you like plays at the plate, triples and the hit and run. If you only like mashers, you will disagree with me.
if this was still new to me, i wouldn't understand
by N Oakley on Apr 1, 2009 9:56 AM CDT reply actions 0 recs
Good article
"Manny Trillo is coming in to pinch run. You know, for a lot of teams, you would pinch run for Manny Trillo." - Harry Caray
by Archie on Apr 1, 2009 10:43 AM CDT reply actions 0 recs
Lies, Damn Lies, and Statistics
"Prince Fielder Dies Of Inside-The-Park Homerun" - The Onion
by DTJchris on Apr 1, 2009 11:23 AM CDT reply actions 0 recs
I wonder about the amount
of variance in winning that defense is able to predict.
- You have the dependent variable, which is winning %.
- Then you have the predictors which are pitching, offense and defense (i.e., you can have subtypes such as starting pitching and bullpen, or you can have just 1 weighted score for a team’s pitching).
You do a pretty straightforward multiple regression and see how much each is able to predict winning. I wouldn’t expect the predictors to be highly correlated so interpretation should be possible.
My guess is that this must have been done many times, but we don’t hear clear results such as “defense is 30% of winning” or “offense is twice as important as defense”. I wonder why we don’t see results such as this talked about commonly, especially since the models and data available now can give us a good grasp about how important each part of the game is.
by Luis on Apr 1, 2009 11:54 AM CDT reply actions 0 recs
The Bill James Fielding Bible
For the sake of a counter viewpoint the Fielding Bible has Soriano ranked the 11th best LF.
by dorf on Apr 2, 2009 6:47 AM CDT reply actions 0 recs
Do you mean the Bill James Handbook?
The Fielding Bible is written by John Dewan.
"That little kid at second base - he is after a job, isn't he?" ~ Lou Piniella, 3/9/09
by dat cubfan daver on Apr 2, 2009 9:34 AM CDT up reply actions 0 recs
What if I took both and shuffled them like a deck of card?
if this was still new to me, i wouldn't understand
by N Oakley on Apr 2, 2009 9:45 AM CDT up reply actions 0 recs
I think your brain would implode.
Either that or the earth would start spinning backwards. You just can’t go around jumbling up sabermagicks like that.
"That little kid at second base - he is after a job, isn't he?" ~ Lou Piniella, 3/9/09
by dat cubfan daver on Apr 2, 2009 10:04 AM CDT up reply actions 0 recs
Well, what if I told you that the Earth's axis of rotation was slowly changing?
Would you blame the sabermagicians?
I love to play baseball. I'm a baseball player. I've always been a baseball player. I'm still a baseball player. That's who I am. - Ryne Sandberg
by Trey2317 on Apr 2, 2009 10:17 AM CDT up reply actions 0 recs
James features prominently in Fielding Bible II, actually.
And Dewan works on Bill James’ books. So the metrics in any James books are, in fact, Dewan’s +/-. (Or Fielding Win Shares but… please don’t use FWS.)
by cwyers on Apr 2, 2009 10:15 AM CDT up reply actions 0 recs
Ah, gotcha.
I’m really close to buying The Fielding Bible II here. After several of these discussions, my interest is piqued.
"That little kid at second base - he is after a job, isn't he?" ~ Lou Piniella, 3/9/09
by dat cubfan daver on Apr 2, 2009 10:22 AM CDT up reply actions 0 recs
That's where it is...
I glanced at it walking out to work this morning; and I swear it said Fielding Bible
by dorf on Apr 2, 2009 11:04 AM CDT up reply actions 0 recs
and reading the rest of the follow ups before posting would have informed me that somebody cleared that up…
by dorf on Apr 2, 2009 11:07 AM CDT up reply actions 0 recs
Appropos of nothing...
…I would also like to give Eric Neel huge points for the Spinal Tap reference in the title of the article in question.
I mean, you can’t exactly dust for vomit, now can you?
"That little kid at second base - he is after a job, isn't he?" ~ Lou Piniella, 3/9/09
by dat cubfan daver on Apr 2, 2009 10:06 AM CDT reply actions 0 recs


















