Defensive Metrics
by Smoltz's Beard on Jan 15, 2009 12:33 PM EST
Email
PrintI believe it was FrankyWren, apologize if I misremembered, that mentioned it might be helpful to post some threads that could introduce some sabermetric analysis for those who are new to the scene. You may not know that there's a blog on SBNation that has begun to run a series doing just that: Beyond the Box Score. Most recently they've tackled the ever elusive defensive metrics with this post.
If you like what you see, there are additional articles here and here.
*Gondee, if you don't want me linking directly to this other blog please feel free to delete this thread.
This FanPost does not express the views or opinions of Talking Chop.
123 comments
|
1 recs |
Do you like this story?
Comments
can we like…adopt ONE METRIC to use so we don’t have people quoting 4 different stats to state their point?
BIG JOE SUCK ONE
No
When it comes to an argument, people will bring up whatever stats make them right, whether or not anyone else knows what they are or not.
No wonder nobody likes you, Tuttle... everything's a (Pujols) damn debate.
oh no, i do it too. its just getting a little weird trying to pick a random stat to make my case when what i normally use makes me look like an asshat
BIG JOE SUCK ONE
Should just make a list and start crossing some off.
Ultimate Zone Rating
Plus/Minus
Revised Zone Rating + Out of Zone Plays (Hardball Times)
Fielding Runs Above Average (Baseball Prospectus)
FRAA is bad because it uses putouts, assists, and errors instead of zone based stats like the rest. I remember hearing someone call it the RBI of fielding metrics, which is a pretty accurate. The only good thing about FRAA is that it has older data, but it’s still flawed. RZR is nice, but it’s more difficult to convert to runs/wins. It’s nice for just saying player A is better than player B.
Plus/Minus is based on plays. Just multiply the number of plays by 0.8 to convert to runs so it’s on the same scale as UZR. And since both of those can be expressed in terms of runs without the flaws of FRAA, I like those two the best. I’m sure I could find out the methodology behind them both and pick my favorite, but I’m way too lazy for that.
Any other fielding metrics out there? Those are the only 4 I can think of that are well known and fairly easy to access.
UZR/150 gets quoted a lot, though its pretty much the same as UZR
why was 150 chosen anyway?
BIG JOE SUCK ONE
Probabilistic model of range is pretty good as well. I try to use that, UZR, and +/-. FRAA is garbage; the new BP play-by-play fielding stat should be good enough to include.
When does the new BP metric come out? Or is it already out?
I knew I forgot one. PMR is good too.
For those of you who don’t know, a flaw with some of the zone rating ones is that they may be skewed by the park. PMR, if I recall properly, uses a player’s performance at a specific park and measures it against everyone else who fielded the position at the same park. One conclusion that it comes to is that guys like Manny aren’t as bad as some of the other metrics say because of the Fenway effect.
Funny how BP has been completely left in the dust
by BraveBronco0121 on Jan 15, 2009 5:29 PM EST up reply actions
Yeah it's funny
The book that got me into the Saber world was Baseball Between the Numbers (and I actually had to google it since I couldn’t remember the name).
Same with me. Do you wanna…I dunno…like, go out and have drinks sometime?
I guess I should be one to talk.
There's nights that I can't even walk.
There's days I couldn't give a fuck.
And in between is where I'm stuck.
by Smoltz's Beard on Jan 15, 2009 5:49 PM EST up reply actions
Moneyball was mine.
I <3 BP, where’s the Haterade coming from?
by mattdiaz4life on Jan 15, 2009 6:10 PM EST up reply actions
Not for long. They promise that their new annual will address two huge criticisms of their metrics:
1) That the replacement level in their WARP metric is too low. They have said that they are raising the replacement level in line with new research, and their WARP figures will likely once again become the cream of the crop in terms of overall value.
2) That they have no play-by-play defensive metric. Clay Davenport has reportedly been working on The Fielding Stat to Rule Them All, and since I still believe Davenport is the smartest sabermetrician in the business, I don’t doubt that BP’s new defense metric will be a world-beater. Clay doesn’t get quite the recognition of some others because he doesn’t open-source his work, but BP has been listening to the sabermetric crowd more than you might think.
Look for all of this in their new annual, due out in February. They also say that there could be a revised version of Dan Fox’s Baserunning Runs included, which would give their WARP figures a boost over others by the inclusion of baserunning. They know that they are perceived as having been “left in the dust” and I don’t think they are particularly content with that.
That’d be lovely. I’m stuck between +/- and UZR, though…
I guess I should be one to talk.
There's nights that I can't even walk.
There's days I couldn't give a fuck.
And in between is where I'm stuck.
by Smoltz's Beard on Jan 15, 2009 12:53 PM EST up reply actions
Yeah, but I’d like to use the one that is widely considered more accurate. Is there somewhere that I can read what they do that is similar and what they do that is different?
I guess I should be one to talk.
There's nights that I can't even walk.
There's days I couldn't give a fuck.
And in between is where I'm stuck.
by Smoltz's Beard on Jan 15, 2009 5:06 PM EST up reply actions
Dammit
I want that, and I’m also lazy…
by Bravenewworld on Jan 16, 2009 11:26 AM EST up reply actions
Why not use both to describe the whole picture and let the interpretation decide?
by soup du jour on Jan 15, 2009 3:24 PM EST up reply actions
Good stuff
I haven’t been to BtB in a while so thanks for pointing the article out. I have to say I wish he would have addressed the pros/cons a little more, but at the very least he did a good job of informing peeps on how to get started.
"Debated ya right not one person agreed with me" by ATLsportsfrk on Dec 27, 2008 6:31 PM EST
Thanks SB
This kind of stuff is helpful. And for those of you who refuse to acknowledge stuff like this/prefer shooting from the hip (e.g., “KJ is a great second basemen. I love what I’ve seen of the guy!”) … LISTEN UP (or I guess read up)!
SB is trying to help …
"He's getting better, but he's not there yet ..."
- Bobby Cox (talking about Boyer)
leary......
I have always been a bit leery of defensive stats in general. Granted it might be because I have never really tried to sit down and understand them. When I see one stat declare a player and ‘excellent defensive player’ and another declare the same player ‘below league average’ i just lose my ability to consider any of them worthwhile. But I will look into the magic of UZR and +/- to make sure you guys really are just insane.
It’s important to remember to take a broad look at the stats available, include as many data points as possible and a large enough sample size. They can tend to bounce around, but I feel like at the very least you can find a player’s true talent region (ie, +3 to +6, -7 to -10, etc.). That’s really what matters; exactness isn’t terribly important.
It helps to throw away the idea that any stat is either GOOD or BAD.
Everything’s in between. Even FRAA is a 3/10, not a 0/10. UZR, +/-, and PMR, given current data available are probably “only” 6 or 7 out of 10. But that’s better than other options, including most individuals’ personal observations.
Beyond the Boxscore // Calling BJ Upton lazy is lazy.
What the hooey how often do you lurk on this board Mr. Kalkman?
Not grading everything in binary probably applies to a lot of what’s said on this site.
Also, I heard that UZR called BJ Upton lazy.
Seriously…I saw the he posted on this thread and figured I’m being stalked or something…
I guess I should be one to talk.
There's nights that I can't even walk.
There's days I couldn't give a fuck.
And in between is where I'm stuck.
by Smoltz's Beard on Jan 15, 2009 2:54 PM EST up reply actions
I'm not here much. But more often than I leave a mark.
Beyond the Boxscore // Calling BJ Upton lazy is lazy.
Food for thought
Ok, I am sure I am going to get lamblasted by all here, but here goes anyway:
What about fielding %? Ok, ok…I know that it doesnt reflect the range of a player and it gets skewed when a player can reach a ball, not make a clean play that another player wouldnt even be able to reach at all. BUT, it does at least indicate how cleanly a player fields the ball, when he DOES get to it. All of these UZR-type ratings, at least to my ignorant of the matter eyes, seem to base all defensive ability on range alone. Range is indeed an important factor to consider, but what about getting the ball in the glove, transitioning from the glove to the throwing hand, and the throw (and with the throw, you gotta take into consideration arm strength as well as accuracy).
Plus, some 1B can bail out a lot of SS/2B/3B from an error by having a great glove/stretch/reaction. Are there any metrics out there that include the WHOLE picture of what it is to play defense? That is to say, is there a metric that accounts for: Range, “glove ability”, transition ability, throwing accuracy, arm strength, getting bailed out of an error by a great 1B?
"Success is not final. Failure is not fatal. It is the courage to continue that counts." - Sir Winston Churchill -
by justincredubil02 on Jan 15, 2009 2:25 PM EST reply actions
fielding percentage is dependent on errors, not all of which are consistent, charged fairly, or have the same criteria behind them. you get penalized the same if a ball rolls between your legs as you do if you make an amazing diving play on a ball with a fast runner going, and wing it off the first baseman’s glove.
BIG JOE SUCK ONE
good point
i forgot to include that in my post. but, it does bring up another question: aren’t ALL of these sort of metrics subjective and flawed? what makes FP so much worse than others?
I am not advocating FP here, as i know there are flaws with it. i am just wondering what makes it so much worse than other metrics?
"Success is not final. Failure is not fatal. It is the courage to continue that counts." - Sir Winston Churchill -
by justincredubil02 on Jan 15, 2009 2:36 PM EST up reply actions
The problem with fielding % is that players aren’t punished for balls they can’t get to.
Say hypothetically player A and B have 100 balls hit to their left, 100 hit to their right, and 100 hit right at them.
Player A gets 70 balls to the left, 70 balls to the right, and 100 right at him and makes 0 errors. Therefore, he made 240 plays out of 300 possible.
Player B gets 90 balls to the left, 90 balls to the right, and 100 right at them, but makes 20 errors, failing on 20 balls he got to. Therefore, he made 260 balls out of 300 possible.
Fielding % would say Player A is 100% while player B would have have a fielding % of 92 or so (260/280).
However, a range based metric would say Player B is 20 plays better than player A in this scenario.
i get why there are problems
with FP…i am just wondering why it is WORSE than the other flawed defensive metrics…
"Success is not final. Failure is not fatal. It is the courage to continue that counts." - Sir Winston Churchill -
by justincredubil02 on Jan 15, 2009 3:02 PM EST up reply actions
let's use your example
if player B’s errors on his 20 balls end up in a runner getting to an extra base, or allowing a run to score, as opposed to player A, who just didnt get there, Player A’s defensive play would have been better for his team than player B’s. Also, your example only takes range into the equation as the difference. Let’s say player A has a better arm/more accuracy than Player B, and also that player B had more balls saved by the defensive ability of his 1B than player A…your metric would indicate that player B is a better defensive player bc he has more range, while player A, with less range, is a better (cleaner) fielder…
I am not arguing in favor of one metric over another, and certainly not in favor of FP as the best metric. I am just curious as to why it is generally considered a WORSE metric than the others. I am also wondering if there are any metrics that take every aspect of defense into consideration (see post above…)
"Success is not final. Failure is not fatal. It is the courage to continue that counts." - Sir Winston Churchill -
by justincredubil02 on Jan 15, 2009 3:06 PM EST up reply actions
Errors and missed plays due to range are the same thing. They are both simply plays the defense failed to make, but for different reasons. A 1B knocks down a line drive but can’t make the play to the pitcher ends in an error, but even though he turned a double into a single. A 1B who can’t go to the ball at all doesn’t get the error but lets a double through. Missed plays are just as bad as errors.
Zone based metrics take both fielding and range into account. A player with a better arm will show up in Zone based stats because it’ll allow him to make plays that a player with an inferior arm can’t make. Likewise, if Yunel bobbles a ball he got to, the stat punishes him the same as Jeter who didn’t even get to the ball. Zone based metrics don’t care how the player failed to make the play. They only care if the play was or wasn’t made.
FP only looks at balls the player was capable of getting to and the judgment is subjective and inconsistent. FP is extremely flawed and doesn’t look at the whole picture.
As for what you’re asking for in the first post, most fielding metrics don’t break things down like that. Fielding stats, in general, just give an overall grade for a fielder’s ability to make plays. I think, however, Field Bible does. Mouse over Fielding Bible Excerpts. They even have a thing about a 2B’s ability to turn double plays. There’s other places where you can find more on arm strength and stuff too.
i get what you are saying
about turning a double into a single, but the opposite is true as well…how many times can a infielder turn a single into a double due to a bad throw, kicking the ball, etc…
"Success is not final. Failure is not fatal. It is the courage to continue that counts." - Sir Winston Churchill -
by justincredubil02 on Jan 15, 2009 3:43 PM EST up reply actions
Just as often as Raul Ibanez lets a ball drop for a double (but isn’t given an error for).
Yeah, sometimes an error is worse than not making a play, but sometimes not making a play is worse than an error. It evens out. These metrics use the average value of specific plays to give them run values.
In the grand scheme of things, errors don’t even make up a large number of plays. A “bad” SS makes like 20 errors? But they have over 500 putouts + assists and probably a couple hundred balls in the zone they never got to. Why are we evaluating a player’s defensive ability on 20 plays when we have 600+ plays to evaluate them on? And it really seems like a lot of writers do they that cause at the end of the day someone gets labeled as a defensive crippled because “he makes to many errors.”
kind of like
fookie and escobar were labeled that way bc of their errant throws…or how for some reason, known only to God, that Jeter is regarded as a great fielder?
seriously (i am going to rant here) Jeter is a great hitter. He is a good ball player. he is NOT a great defender. he is average. why in the name of all that is sacred and holy is that play against the A’s back in the ALCS regarded as a top play of all time??? wtf is the ss doing cutting off a throw from RF anyway? that should be either the P or the 1B. and if Jeremy Giambi had simply slid, we would never even talk about that play again, unless it was the play that ended the yankees run that year.
and the play with jeter “diving” into the stands to catch that foul ball…makes me want to puke. He did a good job of running that ball down, but “diving” into the stands was just dramatic…he was a good 10-15 feet away from the wall when he caught the ball. he JUMPED into the stands AFTER he caught the ball. that play was no greater than other players running out to CF or LF to make a catch. He just happened to run to his right, and didnt have as much room as other players to finish his run. I hate the media’s man-crush on Jeter.
"Success is not final. Failure is not fatal. It is the courage to continue that counts." - Sir Winston Churchill -
by justincredubil02 on Jan 15, 2009 4:00 PM EST up reply actions
There’s a great chapter in Baseball Between the Numbers about Jeter’s defense.
I guess I should be one to talk.
There's nights that I can't even walk.
There's days I couldn't give a fuck.
And in between is where I'm stuck.
by Smoltz's Beard on Jan 15, 2009 4:03 PM EST up reply actions
does it agree with me?
"Success is not final. Failure is not fatal. It is the courage to continue that counts." - Sir Winston Churchill -
by justincredubil02 on Jan 15, 2009 4:07 PM EST up reply actions
Read it and find out yourself, fool!
I guess I should be one to talk.
There's nights that I can't even walk.
There's days I couldn't give a fuck.
And in between is where I'm stuck.
by Smoltz's Beard on Jan 15, 2009 4:08 PM EST up reply actions
awww, come on...
i am too busy to read for pleasure. between work and getting my master’s i am swamped in books! :)
"Success is not final. Failure is not fatal. It is the courage to continue that counts." - Sir Winston Churchill -
by justincredubil02 on Jan 15, 2009 4:11 PM EST up reply actions
What? So you think you are my Master now??
best defensive shortstop in baseball hahahahahahahahahah (omar visquel)
???
huh?
"Success is not final. Failure is not fatal. It is the courage to continue that counts." - Sir Winston Churchill -
by justincredubil02 on Jan 15, 2009 5:10 PM EST up reply actions
No worries, it is a Chris Rock skit...
What subject are you getting your Master’s in?
best defensive shortstop in baseball hahahahahahahahahah (omar visquel)
oh...gotcha
american military history – emphasis on the revolution.
"Success is not final. Failure is not fatal. It is the courage to continue that counts." - Sir Winston Churchill -
by justincredubil02 on Jan 15, 2009 6:05 PM EST up reply actions
I loved the military history classes I took in college (only ones I really paid any mind to, to be completely honest).
However, I preferred the WW2 era.
I guess I should be one to talk.
There's nights that I can't even walk.
There's days I couldn't give a fuck.
And in between is where I'm stuck.
by Smoltz's Beard on Jan 15, 2009 6:07 PM EST up reply actions
as do i...
but the revolution is an easier topic to cover – thus making it easier to write the epically long papers…
"Success is not final. Failure is not fatal. It is the courage to continue that counts." - Sir Winston Churchill -
by justincredubil02 on Jan 15, 2009 6:25 PM EST up reply actions
You know…if I had used that same line of thinking I could have averted many headaches. Well played.
I guess I should be one to talk.
There's nights that I can't even walk.
There's days I couldn't give a fuck.
And in between is where I'm stuck.
by Smoltz's Beard on Jan 15, 2009 6:59 PM EST up reply actions
I have a couple friends who got their PhDs in History and it always blows my mind how focused their dissertations are. One of my friends wrote his on Migrant Farm Workers in the Southwest named Hernandez at the Turn of the 20th Century, or something like that.
I am finishing up my Master’s in Psychology. Fun stuff.
best defensive shortstop in baseball hahahahahahahahahah (omar visquel)
“fookie and escobar were labeled that way bc of their errant throws…or how for some reason, known only to God, that Jeter is regarded as a great fielder?”
Yeah, exactly. Must be the calm eyes?! Or maybe because he knows how to win?
but,
anyone who REALLY watches Jeter with their eyes can see that he is not even above average defensively. and the whole “knowing how to win” argument in baseball just makes me want to puke! :)
how can ONE player win a baseball game? I know how to win too – score more runs than the other team! :)
"Success is not final. Failure is not fatal. It is the courage to continue that counts." - Sir Winston Churchill -
by justincredubil02 on Jan 15, 2009 4:07 PM EST up reply actions
Hahaha
Yeah it’s pretty funny some of the things they say. I can’t even believe an adult could formulate those arguments. Fire Joe Morgan was great for that.
“I’ve always said David [Eckstein] was the smartest guy on the field every night, and that included both coaching staffs,” Maddon said Friday. “And I’ve never seen the guy have a bad day. Even if he goes 0-for-4 and makes three errors, he helps you.”
just think
he gets paid exponentially more than we do to be that stupid.
"Success is not final. Failure is not fatal. It is the courage to continue that counts." - Sir Winston Churchill -
by justincredubil02 on Jan 15, 2009 4:30 PM EST up reply actions
Exactly.
I was just looking up The Fielding Bible when you posted this.
Here is a short description of how the +/- system is utilized, from John Dewan himself:
My book, The Fielding Bible, goes into great length (ad nauseum to some) describing the new fielding system we developed at Baseball Info Solutions, the Plus/Minus System. Video Scouts at BIS review video of every play of every major league game and record detailed information on each play, such as the location of each batted ball, the speed, the type of hit, etc. Using this in-depth data, we’re able to figure out how each player compares to his peers at his position. How often does Derek Jeter field that softly batted ball located 20 feet to the right of the normal shortstop position, for example, compared to all other major league shortstops?
A player gets credit (a “plus” number) if he makes a play that at least one other player at his position missed during the season, and he loses credit (a “minus” number) if he misses a play that at least one player made. The size of the credit is directly related to how often players make the play. Each play is looked at individually, and a score is given for each play. Sum up all the plays for each player at his position and you get his total plus/minus for the season. A total plus/minus score near zero means the player is average. A score above zero is above average and a negative score is below average. Adam Everett turned in the highest score we’ve had in four years of using the system with a +43 at shortstop in 2006. That means he made 43 more plays than the average MLB shortstop would make.
You’re going to tell me that some 60-year old scorekeeper calling the shots by himself is a better metric than what is described above?
I guess I should be one to talk.
There's nights that I can't even walk.
There's days I couldn't give a fuck.
And in between is where I'm stuck.
by Smoltz's Beard on Jan 15, 2009 3:43 PM EST up reply actions
one problem with using errors is Ryan Braun… he didnt make a single error all year… am I to beleive he is a great defender in LF, maybe…. but errors are not correctly accounted for most of the time. The rule of thumb (its almost a pun) is that you can’t be charged with an error if you don’t touch the ball. This isn’t always the case, but generally thats a rule scorekeepers use. SO if Raul Ibanez (for example) takes a horrible route to a ball and never touches it he isnt charged with an error. Meanwhile, ANdruw gets to the ball and just doesnt make the catch, we are to beleive in this isolated incident that what Andruw did was worse than what Raul did, which is hardly the case (in this isolalte incident they both suck ass). Also, scorekeepers are baised because they work in one stadium ( i beleive), whihc can tend to sway how they judge rulings.
"We win today, that's two in a row... if we win tomorrow, that's called a winning streak. It has happened before..."
This is a major reason
why i believe that errors should be charged either by a member of the officiating crew (as opposed to the home-town scorekeeper) or by MLB AFTER the game is played. Or maybe, MLB should institue a stats commission that is responsible for reviewing all games played publishing a correct box score after the games are completed.
"Success is not final. Failure is not fatal. It is the courage to continue that counts." - Sir Winston Churchill -
by justincredubil02 on Jan 15, 2009 3:37 PM EST up reply actions
OF errors are a joke.
Playing OF is all about running down fly balls. How often to outfielders have a ball bounce off their glove? Rarely. How often do they come up just short of catching the ball? ALL THE TIME. Measure plays made compared to plays that should have been made. PLEASE!
Beyond the Boxscore // Calling BJ Upton lazy is lazy.
I for one think you should delete any links to that crappy BtB blog.
Beyond the Boxscore // Calling BJ Upton lazy is lazy.
Pardon my ignorance...
Because I admittedly have no idea what I am talking about when it comes to defensive statistics, but it seems like some of these range measurements would be difficult to acquire. I see on saber-friendly blogging 101 they have a grid, so I suppose that is used, I suppose range is determined by which part of the grid a player can get to and field the ball. But doesn’t this seems awfully subjective, I mean you have to eye it right, you have to look at where the guy catches the ball and determine where that is on the grid. Doesn’t that seem difficult and is this really done by a stat keeper on every ball put in play? Or am I completely missing the point?
it would also make a difference
whether the player is right or left handed and whether the batter is right or left handed, whether the defense is pulled in, in double play depth, in a shift, etc…way too many variables to have a dogmatic metric. that is why, at least as far as defense is concerned (and to some extent, offense) i am just inclined to lean more towards results (like the generic stats) combined with what i see with my eyes.
"Success is not final. Failure is not fatal. It is the courage to continue that counts." - Sir Winston Churchill -
by justincredubil02 on Jan 15, 2009 3:30 PM EST up reply actions
i am just inclined to lean more towards results (like the generic stats) combined with what i see with my eyes.
I just find that way of thinking completely flawed. “These new metrics don’t entirely make sense to me, and because of this I’m going to ignore them and keep believeing in the old metrics.” I agree that the new ones are not perfect, but they are an improvement over what is already out there.
I guess I should be one to talk.
There's nights that I can't even walk.
There's days I couldn't give a fuck.
And in between is where I'm stuck.
by Smoltz's Beard on Jan 15, 2009 3:37 PM EST up reply actions
it isnt that they dont make sense
i get what they are saying. I just think that, like your dislike of FP, there are way too many variables (as i have mentioned above) involved in even the new metrics, so that I feel that the best way to judge a player, is to actually watch him – and not just as he takes infield practice. you gotta watch his reactions in games. i almost am inclined to think that defense is more of an “intangible” that really cant be measured or taught. I was somewhat of a defensive specialist in highschool (i know, reliving the glory days here…lol) so, i know that players with the natural instincts are better than others. i just find the whole idea of trying to bottle up how good someone is defensively with a number is just over-doing it. just watch them…your eyes can tell you more than an arbitrary number can.
"Success is not final. Failure is not fatal. It is the courage to continue that counts." - Sir Winston Churchill -
by justincredubil02 on Jan 15, 2009 3:41 PM EST up reply actions
The problem with watching players is that it’s simply impossible to watch every player for the appropriate amount of time. It’s simple logic that you would spend most of your time watching the Braves, and if you were to try and compare Jeff Francoeur to Nick Markakis defensively…it simply wouldn’t be fair.
I guess I should be one to talk.
There's nights that I can't even walk.
There's days I couldn't give a fuck.
And in between is where I'm stuck.
by Smoltz's Beard on Jan 15, 2009 3:45 PM EST up reply actions
exactly
which is why the average fan should not be a credible source as to who is good and who isnt. the sports media is a large enough organization, and so is MLB, that they could have a team of employees whos sole job is to do what we are talking about…plus, educated fans would be able to eradicate the uneducated pretty quickly during a debate! :)
"Success is not final. Failure is not fatal. It is the courage to continue that counts." - Sir Winston Churchill -
by justincredubil02 on Jan 15, 2009 3:47 PM EST up reply actions
That’s exactly what these companies are doing, they just aren’t using errors. I feel like it’s ridiculous to ask, but would you be more comfortable with these metrics if they were using terminology that you were familiar with?
I guess I should be one to talk.
There's nights that I can't even walk.
There's days I couldn't give a fuck.
And in between is where I'm stuck.
by Smoltz's Beard on Jan 15, 2009 3:59 PM EST up reply actions
no. that's not the point.
the terminology isnt my issue. my issue is that the variables still exist, yet that is overlooked by the proponents of the new stats. yet, those same people will emphatically dismiss older metrics due to the variables present in them. I think, ideally, if you want to have a number assigned to a player to determine if he is good or not, you have to use a combination of everything, all the metrics, and what your eyes tell you.
"Success is not final. Failure is not fatal. It is the courage to continue that counts." - Sir Winston Churchill -
by justincredubil02 on Jan 15, 2009 4:34 PM EST up reply actions
Tom Tango did a study last year where fans voted 1-10 on a variety of different fielding traits for players, such as arm strength, accuracy, first step, speed, etc…but he requested that you only vote if you had seen enough of the player. Wisdom of crowds.
by mattdiaz4life on Jan 15, 2009 6:12 PM EST up reply actions
I would have to imagine that they have some kind of a computer program that is capable of laying the grid on top of the baseball diamond and using a freeze frame of some sort to pinpoint the exact sector on the grid in which the play is made. I’m sure the exact method can be found online somewhere.
And yes, they record every ball put in play over the course of the season and have been doing so for years now.
I guess I should be one to talk.
There's nights that I can't even walk.
There's days I couldn't give a fuck.
And in between is where I'm stuck.
by Smoltz's Beard on Jan 15, 2009 3:34 PM EST up reply actions
This is the issue with all defensive statistics.
Other than FP, which is flawed on its own, as discussed upthread, all these “zone” stats are eyeballed. It almost seems to go against the whole statistical analysis movement.
Defense still HAS to be eyeballed. You can’t just put one number over another number and then produce a percentage or something. It’s why so many, including myself are skeptical of defensive metrics.
If you go and look at a thread about KJ being the LF next year, there was a long discussion of whether or not KJ was a good defensive 2B. The different metrics had wildly different results, and no one can make a compelling argument for one over another.
Until there is a confluence or reduction in methodologies, the ambiguity of defensive metrics will continue to be the issue.
The thing is they can measure with accuracy where the ball was hit and how hard it was hit.
Discrepancy between the stats usually occur from different zones and while they don’t agree to the dot, they give you a good general picture. KJ is somewhere between a below average to average defensive player. Most people who say he’s absolutely awful at the position note his poor fielding (which I agree with), but forget about his awesome range.
except for that one time
he dropped a pop up to end the inning (and the game) against philly…that was just inexcusable! and i am a KJ supporter!
"Success is not final. Failure is not fatal. It is the courage to continue that counts." - Sir Winston Churchill -
by justincredubil02 on Jan 15, 2009 3:51 PM EST up reply actions
There’s actually was consensus on Kelly Johnson as a fielder in 2B using defensive metrics available. He’s seen as a bit below average and that’s good enough.
PMR had him 5.9, UZR had him -8.3, +/ had him 0.8. All three stats were stable from 2007 – (UZR -7.4, PMR -8.7, +/ 0). Taking the sum of those, he’s a -5 fielder in 2008 and -5.4 fielder in 2007.
The zone picked by zone based stats is completely objective. They have different ways for doing it, but here’s one way. They look at ALL line drives hit near or around position X and see how many shortstops make that play. If, for example, that number is greater than 50, then that’s considered in the player’s zone so players not making plays in that zone are punished. Now, let’s say a fly ball is hit to location Y, but only 10 of shortstops make that play. Then shortstops who don’t make that play are not punished, while shortstops who do make those plays are rewarded.
that sounds like
a good way to do it, but again, you have to look at the situational fielding. what teams are in what defensive positioning when that play is made? is the batter right or left handed (which affects the location of the fielder), is a hit and run on, is the D in double play depth, is the defender right or left handed?
Still way to many variables to simply tack a number to it and call it a day.
"Success is not final. Failure is not fatal. It is the courage to continue that counts." - Sir Winston Churchill -
by justincredubil02 on Jan 15, 2009 3:49 PM EST up reply actions
No one says defensive stats are perfect. Just that taken as a broad picture we can call certain players good, bad, average, great, or terrible. I mean, no one focuses on whether a hitter is a true talent 275/350/450 hitter or a true talent 270/340/440 hitter – we just focus on that player being a bit above average ML hitter. Same for defensive stats.
The thing is it’s up to the fielder to make those decisions. If a player has perfect instincts, then he always knows to take a couple steps in for Ichiro Suzuki, take a few steps to his right for a right handed hitter, a few steps in for the lefty, take some steps to the right when the catcher pitching a lefty down and away, etc. Their instincts, in this case, are rewarded by being able to make more plays.
Like 17843 said, stats aren’t perfect, but they give you a good idea of whether a player is good at what he’s doing.
Again…it seems like you’re going out of your way to find things wrong with these new metrics. I’m not saying to blindly accept whatever you hear, we all know that doesn’t work as well. But you just seem complacent with relying on old methods without giving new ones a chance.
I will give you one thing, though…having you around certainly is good for discussion.
I guess I should be one to talk.
There's nights that I can't even walk.
There's days I couldn't give a fuck.
And in between is where I'm stuck.
by Smoltz's Beard on Jan 15, 2009 4:07 PM EST up reply actions
lol...
i am not trying to argue against modern metrics. i just think that they are no better or no worse than the old ones. both have serious flaws that cant be ignored. i have no problem with modern metrics in and of themselves. but to dismiss older metrics like fp due to the variables and subjectivity of them and not do the same for the newer metrics is somewhat hypocritical. I am not going out of my way to find things wrong. in fact, i have never even really thought this much about this topic until right now.
"Success is not final. Failure is not fatal. It is the courage to continue that counts." - Sir Winston Churchill -
by justincredubil02 on Jan 15, 2009 4:10 PM EST up reply actions
Disagree
I am one of the bigger pessimists on here regarding defensive metrics, BUT they are FAR, FAR more reliable than the older models. Just because they are not 100% accurate, nor the fact you/we don’t fully understand them yet, should have no bearing on whether they are better/even/worse than older metrics. That’s a faulty argument.
"Debated ya right not one person agreed with me" by ATLsportsfrk on Dec 27, 2008 6:31 PM EST
Discussion’s a good thing!
The thing is newer metrics aren’t subjective and are very objective. They determine the zone that they base all players on through what all players at that position do.
The problem with FP is that it looks at a smaller picture than Zone Based stats and is subjective. It’s basically worse in every way.
The fielding percentage metric is wholly based on subjectivity. The new metrics are restricted by information, not subjective. They certainly have flaws, but the data they use is not subjective – it simply measures plays made/expected plays made. Not mistakes/chances to make a mistake. It’s impossible to make an error on a ball you’re not close to, even though the end result of an error and a play simply not made is the same thing. PMR/UZR/etc. are certainly not the end all, be all, but they’re miles ahead of fielding percentage.
but they are subjective
in that they assume players are supposed to have a certain range. i mean, can you really say what the average range of a defensive player is? like i have said on here already, plays and range are all based on the situation of the game. I bet if someone were to go back and look at the defensive positioning of the players and redo whatever formula they did to take this into account, you will find drastic differences in the results.
I am not anti-stats or anti-new metrics by any stretch of the imagination. I just think that to use them as any more definative than some of the older metrics is a little irresponsible. both styles of metrics have serious flaws. That is why i said earlier that i am more inclined to take a little of what all of these metrics say, and use them combined with what I can see in the game, and then form my opinion. I dont think we need to label players defensive ability with a number.
"Success is not final. Failure is not fatal. It is the courage to continue that counts." - Sir Winston Churchill -
by justincredubil02 on Jan 15, 2009 4:39 PM EST up reply actions
“i mean, can you really say what the average range of a defensive player is?”
You take every single player at a position (let’s say CF) and you look at all line drives, fly balls, grounders, etc. hit to location X and you look at every ball hit to location Y, Z, etc. and see how many of those plays are turned into outs by the CF. If over 50% of that specific ball in that location is turned into an out (or 60% or whatever), then that counts as part of a position player’s zone. You can use that to determine what an average CF should be fielding.
There’s variances in how each metric scores, but the way the metrics work is far superior to something like FP.
All metrics have flaws, but don’t grade them in binary. Like Sky said earlier in the thread, something like FRAA is 3/10 while something like FP is 1/10. The zone based metrics are 6 or 7 out of 10 which is a much better judge of ability than flawed stats.
Is that any harder than determining error/hit?
And you have a rating of every play on a gray scale, instead of the ocassional black/white decisions. Which one seems better?
Beyond the Boxscore // Calling BJ Upton lazy is lazy.
I’m not saying that +/- or UZR are worse than FP. Of course they are better.
I’m just making the point that it is difficult for non-SABR or other non-stat-minded people to wrap their heads around these newfangled defensive metrics because they don’t see a point that makes it any better than anything else already being used.
I’m all for the new stuff. I do think it needs to be made a little more clear cut before it’s accepted on a more general level. That will happen in time. The beauty of statistics and baseball is that the whole thing is constantly updating itself and become more precise.
For now, however, count me a little bit skeptical.
I can’t see how anyone could see the new metrics as not an improvement if you really understand how they work. Judging based on an objective league-wide standard based on zones compared to judging based on subjective determination by one guy that doesn’t take range into account. I see a massive quality difference there.
So can I...
The issue was about people who DON’T get it.
I was just trying to articulate an argument why SOME would not appreciate the defensive metrics.
I am all for the new stuff. I mean, after all, numbers are the only friends I have in my mom’s basement.
Ok, I agree.
The whole key to getting people to accept new ideas is to explain them in a way where you can’t possibly disagree, which usually involves showing how they are actually the approach that does what they think the bad approach is doing. Without numbers. Or geeky language. And pictures.
Beyond the Boxscore // Calling BJ Upton lazy is lazy.
Love this stuff
I’ve been wanting to have this debate for a while, much thanks to BtB for starting the series (and props to the Beard for sharing).
My only problem with defensive metrics is when people use them to establish a firm value for players whether for a single year or over a historical period. Case in point was the argument (in another post) over whether Adam Kennedy had as much value as Adam Dunn last year. Saying that Adam Kennedy was a far superior defender to Dunn is completely accurate in my book, and is justified by defensive metrics. Saying that Adam Kennedy was as “valuable” as Dunn last year is faulty due to the “definitive” nature of the statement. As I said previously, the newer defensive metrics are far more accurate than their counterparts but they are not 100% accurate. So using them definitively in valuing players is also not 100% accurate, correct? I have no problem with someone making this type of statement as long as they qualify it – such as: using UZR/150 we can estimate X players worth to be $Y.
So am I wrong in my thinking?
"Debated ya right not one person agreed with me" by ATLsportsfrk on Dec 27, 2008 6:31 PM EST
Not sure if I’m answering your question, but I’ll try.
Basically, player values are summed up with wins above replacement (WAR or just wins). To determine the number of wins a player is worth, you take what his bat produces, glove produces, and a positional adjustment.
You can convert these things to runs and then use 10 runs = 1 win. UZR/150, for example, compares a player’s defense to average and gives them a run value for it. Dunn, is probably around -10 runs or so which would be -1 wins for his defense.
If you add all that up, you can come up with a win value for a player and then you assign a dollar figure for a win. I think it’s around $4.5M on the free agent market right now. That’s how you end up with Player A = $10M, Player B = $12M, etc.
Not really my point
but I appreciate the feedback. I’ve been trying to get a better grasp on WAR and how it can be translated to player values and your summary is a big help!
I guess my point is this, if I may borrow from your analogy. Considering UZR/150 is used in the value determination, and we know it is not 100% accurate, is it fair to say that Player A is “exactly” worth $10M. Isn’t this merely an estimation of his worth? If so, is it not also fair to assume that Player A’s value can range, to a certain degree, above or below this stated value?
If this answer to both my questions are true, and depending upon the statistical accuracy of UZR/150, it is also possible that both players could be valued at $11M per, correct? This is my mental block over player values which is rooted in defensive metrics. It’s very hard to explain my issue due to lack of understanding but hopefully this helps.
"Debated ya right not one person agreed with me" by ATLsportsfrk on Dec 27, 2008 6:31 PM EST
You’re exactly right, which is why no one should take statements like “Player A is worth $10 million” as definitive stats with no margin of error. It’s a nice estimation, but I imagine errors in the defensive component are worth something like $1-2 million each way. That’s part of the problem with only using UZR; I think you can get a pretty close estimate of defensive value from a combination of metrics and arm ratings that wouldn’t have an error of more than 1-2 runs ($500K).
can you imagine
trying to quantify the defensive ability of a player like we are doing here in any other sport? basketball, football, hockey – it would be rediculous. in those sports, we look at the results (stats), the intangibles (aggressiveness, desire, etc) and what our eyes show us. why does baseball have to be any different? why do we have to put a number next to a players name to indicate his defensive ability?
"Success is not final. Failure is not fatal. It is the courage to continue that counts." - Sir Winston Churchill -
by justincredubil02 on Jan 15, 2009 4:42 PM EST reply actions
Because we can
The uniqueness of baseball allows for advanced analysis.
There’s advanced analysis in basketball too that I’m not too familiar with like adjusted +/-, defensive efficiency, and other stuff.
but those
are usually (i am not saying always, bc i just dont know) team stats, not individual.
"Success is not final. Failure is not fatal. It is the courage to continue that counts." - Sir Winston Churchill -
by justincredubil02 on Jan 15, 2009 5:01 PM EST up reply actions
I’m not too familiar with the advanced statistics of those other sports either, but here’s one example of an adjusted +/- for basketball that rates and individual players’ defense. Link.
interesting
that the current top players are all the ones that our eyes tell us are the best. :)
"Success is not final. Failure is not fatal. It is the courage to continue that counts." - Sir Winston Churchill -
by justincredubil02 on Jan 15, 2009 5:12 PM EST up reply actions
Yeah, it’s pretty nifty. Some of it is kinda surprising though. Like Nash is considered a defensive cripple, but according to that stat, he’s worth about the same as Garnett on defense.
i guess that is a flaw here
comparing teh defense of a guard to the defense of a C/PF is just wrong…oh well. Maybe I am just too much of a “what i see” kind of guy to like these stats at all. If i saw a good one, i would approve it, but so far, i am just skeptical at best.
"Success is not final. Failure is not fatal. It is the courage to continue that counts." - Sir Winston Churchill -
by justincredubil02 on Jan 15, 2009 5:20 PM EST up reply actions
Baseball is unique in that it can largely be broken down into individual moments – whether a batter hits a ball and what happens, whether the base runner takes the extra base, whether the fielder completes the play. There are so many more variables in other sports that don’t lend themselves to be judged in that format; ie, a left tackle can do his job completely, but if the left guard misses a linebacker, the QB is sacked, and the play is a failure. It’s unfair to dock the left tackle, but perhaps if he had blocked differently the play would’ve turned out differently.
To answer the last question, because we can and because it’s essential to appropriately valuing a player in the business. A hitter who produces 30 runs above replacement with the bat is by most definitions a solidly above-average player. However, if he plays -15 defense, he’s below average. If he plays +15 defense, he’s an all-star and possible MVP candidate. That’s a vast difference in value (30 runs or ~$13-14 million in value). Without valuing defense and having an objective standard to do it by, Adam Dunn might be getting $20 million a year for $10 million worth of preformance and Franklin Guittierez might get 400 K for $10 million worth of preformance.
i know what you are saying
but i disagree…basketball is comprised of players rising up and making one play too. that one steal, or that one block, or that one 3 pointer can stand out just like that one hit, the xtra base, etc.
"Success is not final. Failure is not fatal. It is the courage to continue that counts." - Sir Winston Churchill -
by justincredubil02 on Jan 15, 2009 11:20 PM EST up reply actions
That’s not what I’m saying. In basketball, a player goes one-on-one with a defender, but can be double teamed and if he, for example, drives to the basket he’ll engage numerous defenders and blame can’t really be applied exactly to each defender. In baseball, you can place blame on the batter for not getting a hit, you can give credit to a fielder for fielding a ball, etc. There’s really no way to apply blame to the defense if a player scores or give credit for a guy scoring because there’s tangible impact of others on the play. That doesn’t really exist in baseball at all.
what about
if a pitcher makes a mistake with a pitch, say, leave a curve hanging, or something like that and the batter crushes it (as most would). what about a hit-and-run play where the runner takes off, but the batter strikes out and the runner is subsequently thrown out. should that count against his base-stealing ability? what about a slug of a runner on 1b, a guy mashes a ball, but due to the slug (or a baserunning mistake by the runner) the batter only gets credited with a single. that would factor against his slg%. there are examples that work both ways – maybe not as much in baseball as in basketball (ok, obviously not as much) but still, the point remains.
"Success is not final. Failure is not fatal. It is the courage to continue that counts." - Sir Winston Churchill -
by justincredubil02 on Jan 16, 2009 12:34 AM EST up reply actions
The first example is not the same thing. The chances of a batter seeing an enough mistake pitches to distort his batting line away from his true talent level are very low.
The second example is true.
The third example always isn’t the same thing. We still can credit the baserunner for advancing a base successfully and the hitter for the run value of his single. We cannot adequately give credit to a point guard for making an excellent pass his forward lets go through his hands.
The point is not that baseball always can be broken down to individuals confrontations, but that it can almost always can be and other sports cannot to anywhere near the level baseball can be.
I’ll throw this out there since its a saber-nerd thread
I wanna buy one of the annuals, but have no idea which to buy. Is any one better than another?
BIG JOE SUCK ONE
Are you looking for anything in particular in an annual?
by mattdiaz4life on Jan 15, 2009 6:14 PM EST up reply actions
Well, I’ve heard two are good values – both BP and THT put out good ones. I’m not speaking from experience, though. BP’s had some editing issues that you had to wade through a few years ago, but the last couple have been good.
by mattdiaz4life on Jan 15, 2009 6:54 PM EST up reply actions
hahahah
i had to read this twice…lol…me and my gutter-mind.
"Success is not final. Failure is not fatal. It is the courage to continue that counts." - Sir Winston Churchill -
by justincredubil02 on Jan 15, 2009 6:26 PM EST up reply actions
Comments For This Post Are Closed
Secondary Sidebar
User Tools
FanPosts
Community blog posts and discussion.
Certain photos copyright © 2012 by Associated Press or Getty Images. Any commercial use or distribution without the express written consent of Associated Press and Getty Images is strictly prohibited.
