DefEff
by tommy.otm on Feb 28, 2010 3:26 PM EST
Email
PrintIn Cafadaro's Sunday "Baseball Notes" column, he mentions the statistic "defensive efficiency" (DefEff) as one to watch. It "measures the percentage of balls put into play that are converted into outs." I googled it, and found the stat at Baseball Prospectus. Last year, the Sox were 28th out of 30 teams. Ouch.
Notably, the DefEff numbers aren't the direct inverse of team BABIP. DefEff includes errors, and BABIP by definition doesn't, but it is that the only difference?
I'd look it up, but I'm lazy. (Plus, I've got to hit the gym and watch the hockey game.) But I hadn't heard of the stat and thought I'd share. I know I'm curious how much our team D will improve this season, as measured by all these facy new metrics.
(Speaking of which, I found this Gasper piece on UZR to be amusing. I mean, count me as those who don't believe UZR is "teh best measurement evah," and I'm tired of over-reliance on it to say so-and-so sucks or is great or whatnot. But still, pieces like this make me pine for the days of "Fire Joe Morgan.")
38 comments
|
0 recs |
Do you like this story?
Comments
Oh lord.
meaning he is considered 5.7 percent better than an average first baseman.
If he just said he doesn’t understand it, why would he then try to explain it. That’s not what it means. It means he’s 5.7 RUNS better than an average first baseman, and that’s by UZR only. Even UZR claims that a good UZR is only a good UZR. Still, it’s the best freely available defensive metric, and actually correlates pretty well to things like ERA-FIP and BABIP on a team level (with some exceptions). On an individual level, small sample sizes remain pretty damn questionable. 3 seasons is considered a good minimum to get an average from.
Tommy, how do you feel about Bill James’ +/-? It maintains some of the biggest problems I have with UZR—generalizations in weighting, as the zones in UZR are pretty large:
It seems like not all balls in zone 3-4 are necessarily created equally. Ditto for plus-minus, which makes the very general distinction between “an average player at this position could” vs. “could not” make the play. On the other hand, it eliminates a lot of the distrust of computers by basically just being a huge work of scouting. It is, in a way, brilliant in its simplicity, and quite frankly almost perfectly correct in what it’s trying to measure. It doesn’t matter how many balls that are REALLY hard to get to that a player can save. It only matters how many more balls in play you can convert into outs than could an average defender at your position given the exact same distribution of balls (including things like velocity, which UZR is always sketch about). After all, there are no “bonus points” in baseball for truly amazing plays. All that matters is if it’s an out or not. Weighting for the huge play would give a slightly more accurate result, I expect, but if someone can make the huge play, then they can likely make the “above average” plays that count more often. Ditto that they will rarely miss the easy play.
Also, a small observation on Marco Scutaro: a .6, 7.8, and .9 UZRs these last 3 years. +/- puts him at +1, +9, and +12.
I had been thinking of him as more of an average defensive SS. Seems like he might be well above average.
USG
by Ben Buchanan on Feb 28, 2010 3:56 PM EST reply actions 3 recs
Oh lord...This thing right here needs to turn green
And I’m saying that even before reading it!!!
"That was a lot of fun… You just keep pounding balls into the gap. The one thing you don’t want to do is hit a home run. That’s a rally-killer." Jeff Francoeur
Haha
It’s not as great as the giant graph makes it out to be.
USG
by Ben Buchanan on Feb 28, 2010 4:52 PM EST up reply actions
After reading it, you're right....No I'm kidding.
I read that piece by Gasper last night, and that percentage thing ticked off, so I sent him an e-mail with a link to “The book” so he can have a better understanding about it…I hope he’ll read it!
"That was a lot of fun… You just keep pounding balls into the gap. The one thing you don’t want to do is hit a home run. That’s a rally-killer." Jeff Francoeur
WOW
Never knew there were that many “zones” on a baseball field…..would make for an interesting cow-patty bingo
by BobZupcic on Mar 1, 2010 5:59 PM EST up reply actions 1 recs
+1 and rec'd....for Yankee Stadium.
Galactus does as he pleases. Because Galactus is drunk.
I have a five-tool player in my pants.
I'll have to research some about James' +/-
I haven’t read up on it too much.
Manny ain't the only bad man.
Can that be right?
7.8 UZR for +9 in 2008, followed by .9 UZR for +12? He lost 6.9 points of UZR but gained +3? That tells me that +- must be a count, not a rate — Scutaro played a lot more games at shortstop in 2009 than he did in 2008. And when it comes to measuring defensive skills I’m much more interested in rates than counts, otherwise we might as well just judge players strictly by errors.
For his career, Scutaro has a UZR/150 of negative 2.9 (I’d include the minus sign, but it crosses text out). If you accept that he improved significantly from working with John McDonald in Toronto — and all evidence seems to confirm that he did — he’s probably an above average defensive shortstop now, though not by a huge amount. I definitely think he’s going to be the weakest infield defender in the starting lineup this year, which means our infield should be one of the best in the majors.
I agree on all the points here
And I’ll add the fact that Scutaro Don’t need to have what a good fielding SS has to have in term of range because of the guy on his left side.
Dave Cameron (an M’s fan) on Beltre range:
When you watch Beltre play third base, you are amazed at the things he can do. He has perfected the charge on a bunt – no one in baseball comes in on the ball as well as he does. His lateral range is hilarious at times, as he regularly fields balls that are hit directly at the shortstop, just because he can.
UZR tend to confirm that: His Range Runs above average is ridiculous 14.3
Oh and the guy on his right isn’t bad either.
Boy, We need a petition so Orsillo could bless us with “like a Puma…” when he’ll makes one of his highlight reel plays at 3rd…or at ss.
"That was a lot of fun… You just keep pounding balls into the gap. The one thing you don’t want to do is hit a home run. That’s a rally-killer." Jeff Francoeur
UZR is also a count. UZR/150 is a rate, but you can just divide +/- by 150 innings and get a rate too.
Neither of them will take into account the number of chances a player gets either way. So neither is really a true rate. But saying that they’re equal to errors because they’re not rates is ridiculous. Errors take into account only “botched” plays, not if the play should be makeable or not, how much range the player shows, etc.
Obviously, a good player who gets a lot of innings at a position will have a higher +/- than an equal player who has fewer innings. But because +/- also subtracts for bad plays, a mediocre player is unlikely to manage a high +/- by volume alone, and a bad player won’t breach zero unless they’re having a particularly good year.
And re: Scutaro, a lot of it has to just be that he finally got consistent playing time at one position.
USG
by Ben Buchanan on Feb 28, 2010 7:36 PM EST up reply actions
A few things:
1) “UZR is also a count” — which is why it’s not a good measure for comparison. UZR/150, however, tells you the count per some set unit — that’s exactly what you need to make a good comparison. And count per time unit (e.g., UZR = count; 150 = time unit) is the very definition of a rate.
2) “But because +/- also subtracts for bad plays, a mediocre player is unlikely to manage a high +/- by volume alone…” - depends on how much volume you’re talking about. Show me two players, each with a +/ of +10, where one played 500 innings in the field and the other played 1200, and I know the former has performed roughly 140% better than the latter, though both are good fielders. If they played the same number of innings, the former player is going to have a +/- of 24. That’s why you really want to use rates rather than counts - the former player’s +/ per 150 innings is 3, which is clearly superior to the latter player’s 1.25. In any case, the latter player is basically playing 135 games and providing above average defense, but not by much.
3) “But saying that they’re equal to errors because they’re not rates is ridiculous.” — That’s not what I said. What I said was that simple counting numbers are poor indicators of defensive quality. Hey, Juan Samuel got on base 226 times as a rookie, and that sounds great until you realize he had 737 plate appearances (.307 OBP). By the same token, UZR isn’t really meaningful unless you know how much time a player has spent on the field, just as errors aren’t really meaningful unless you know about total chances, and even then fielding percentage is overrated. But a .9 UZR in 1250 innings — or a UZR/150 of 1.0, since that measure calculates UZR per 150 games — is above average, but not by much.
4) Scutaro’s +9 in 2008 was in 56 games played at shortstop, while his +12 in 2009 was in 143 games. Convert those to /150 games and you get +24.1 in 2008 and +12.6 in 2009. So again, if you just look at the count, 2009 looks better than 2008, but the reality is that Scutaro played way over his head in limited time at short in 2008. And UZR/150 confirms it — Scutaro’s numbers for 2008 would be elite if he played a full season at short, while his numbers for 2009 were above average, but not by much.
All of which points to exactly what I said — we can expect Scutaro to be a slightly above average fielder. Given that Beltre, Pedroia, and Youk are all Gold Glove-quality at their positions, that would make Scutaro the weakest defender in the infield. But again, as I’ve already said, if a slightly above average fielder is our weakest infielder and the other three are Gold Glove-quality, we’re going to have one of the best defensive infields in the majors.
A few things of my own:
1) “And count per time unit (e.g., UZR = count; 150 = time unit) is the very definition of a rate.” But innings is a bad time unit. Realistically it should be per chances. Otherwise a good shortstop on a team with groundball pitchers will be given priority over a good shortstop on a team with flyball pitchers.
2) “What I said was that simple counting numbers are poor indicators of defensive quality. Hey, Juan Samuel got on base 226 times as a rookie, and that sounds great until you realize he had 737 plate appearances (.307 OBP).” Not really a viable comparison. Both UZR and +/- are based around a system where a 0 is perfectly average. A guy who fielded poorly cannot appear to field well because of volume alone, small or otherwise. He can appear closer to average, but that is all. It’s because of this that the rates, while still nice to have, are not entirely necessary. Especially since basic connections can be made by anyone who sees the inning numbers.
3) “All of which points to exactly what I said — we can expect Scutaro to be a slightly above average fielder.” I think your judgement of slightly above average and mine are slightly different. First, I should say that the /- I gave was actually the “Runs saved” stat (they are a +12 and +15, which convert to 9 runs saved and 12 runs saved with a few other factors tossed in). While UZR doesn’t think he’s as good as +/- does, there are a number of reasons that could be the case—likely due to him making a lot of plays within a certain zone and then a very few plays outside of it as compared to a more general drop off. So he gets to most anything X feet away, and then very little X1 feet away (but that’s just an educated guess based on how the systems work). The real upshot of the +/- number, though, is that is has Scutaro pegged as being 1 win above average (average, not replacement). I’d say that’s pretty good.
USG
by Ben Buchanan on Feb 28, 2010 9:25 PM EST up reply actions
For reference, in 2008, the top defensive SS was Jimmy Rollins with +23.
Scutaro slotted in at 8th with +12 (Jed Lowrie was a +8, go figure!). While I agree that he was way above his head that year and likely would not have kept up nearly that pace, his +15 in 2009 would have put him in the top 20% of shortstops. Pretty damn good.
USG
by Ben Buchanan on Feb 28, 2010 9:28 PM EST up reply actions
Wow, I made a huge mistake.
+/- does weight individual plays by how likely they are to be made.
I also take back my speculation about why Scutaro’s UZR and +/- were so different, then, as well as my biggest complaint against +/-
USG
by Ben Buchanan on Feb 28, 2010 9:30 PM EST up reply actions
since when is +/- weighing plays?
i thought you get +1 when you make a play that somebody else hasnt and 1 when you dont make a play that somebody else playing the same position has. Thats why +/ has only full numbers and no decimals. Thats the biggest knock on +/-. On the other hand it takes into account trajectory and velocity of the ball.
A merged version of UZR and +/- would be great.
by German Red Sox Fan on Mar 3, 2010 12:29 PM EST up reply actions
And I quote from the Fielding Bible
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.
USG
by Ben Buchanan on Mar 3, 2010 12:31 PM EST up reply actions
if thats so why are there no decimals?
by German Red Sox Fan on Mar 3, 2010 2:41 PM EST up reply actions
Maybe they only deal in whole numbers? Or they have a decimal number tucked away and round it for the website.
But that’s what they say, so I believe they weight them.
USG
how 'often'? vs how 'likely'?
Not sure I totally grock the language here. If a play is really hard, but really frequent, players will still make the play ‘more often’ than if a play is really easy, but rarely occurs.
I'm not even a stat guy
But as soon as I read “5.7 percent better,” I closed the window…
SB Nation's Boston Red Sox community:
OverTheMonster.com
Follow this guy on Twitter: @RCBooth
DefEff
Sounds like a rapper.
“Soul City Monsta” by Rhianna, featuring DefEff and Missy Elliot.
Galactus does as he pleases. Because Galactus is drunk.
I have a five-tool player in my pants.
I love how CHB does a 180 at the end there.
"These stats suck. OH BUT OF COURSE THE RED SOX USED THEM TO PERFECTION IN 2004"
But that must’ve been a lucky correlation, right CHB?
Look, players don’t need to know about these stats. They’re the ones playing this game. Generally, they need to know that taking a walk isn’t a failure while giving up a walk is. That’s about it. But the people who analyze baseball and write about baseball need to understand them, because they’re often going to be analyzing moves made by front offices. Moves that are made based on these stats (of course the Red Sox have better metrics because of their resources, but UZR is the best publically available metric).
You can, like many others, think there are flaws in the system. Fine, account for them, but the flaws are not a good reason to completely dismiss a metric which shows a great deal of correlation with actual results. Teams with high team UZRs have low BABIPs and negative ERA-FIPs. They also tend to just plain give up fewer runs. They cite Jacoby, and he was probably hurt by a particularly bad fluctuation. That’s why when you talk about his defense you say he was probably hurt by having a bad year that was exacerbated by the fickleness of a single season sample size for UZR.
But don’t go off on some tangent about how you know what you see with your eyes. For the most part, you don’t even watch the whole play. You watch the fielder when they cut to him about half-a-second after the contact. Or you track the ball if you’re at the game. The eyes of announcers have told them that Derek Jeter is one of the best defensive shortstops of all time. But for the most part it’s becoming accepted, even by Yankee fans, that he is far, far from it. It was the numbers that outed him.
As Steve Slowinski of DRays Bay quoted in his pretty fantastic Sherlock Holmes, the Sabermatrician piece:
"It is a capital mistake to theorize before one has data. Insensibly one begins to twist facts to suit theories, instead of theories to suit facts."
The Dan Shaugnessys of the world would have you believe that UZR cannot possibly be correct because it disagrees with his longheld beliefs. I would have you believe that he cannot possibly be correct because he provides no evidence in the face of a great deal.
He’s doing a disservice to his readers and the people who pay his bills by not only ignoring but attempting to discredit the most informative (freely available) statistic on one of the 3 major components of the game, and forming opinions without it. It’s pathetic that he’s paid to put on his farce.
USG
Oh
And when his knee-jerk reaction to a much misinterpreted buzz word inspires him to call for fans to cancel NESN subscriptions and boycott ticket purchaces it borders on what I would call “treasonous”.
F’ing hilarious that the Lackey signing followed less than a week later. I haven’t looked into this, but I’m willing to bet all the money in my pocket that he never issued a ‘mea culpa’. I will if I’m wrong. But I doubt that I am.
USG
CHB is such a tool!
I gotta go 'cause I'm probably definitely gonna nod out again.
by Drugs Delaney on Mar 3, 2010 8:50 PM EST up reply actions
Shaugnessy == failure to learn.
Jeeze after his hatchet job on Randy Moss this fall, he’s now proven he knows nothing about two whole sports that he gets paid to comment at length on.
Comments For This Post Are Closed
Secondary Sidebar
User Tools
FanPosts
Community blog posts and discussion.
Recommended FanPosts
Recent FanPosts
In Association With
Big League Stew - MLB - Yahoo! Sports
- Photo: Padres fan regrets letting mother capture moment with Trevor Hoffman
- Marlins agree to unretire No. 5 for Logan Morrison, who will honor his father’s favorite player
- Weatherman Derek Holland does Dallas TV
- Ask Alex: Will Julio Teheran or Matt Moore win rookie of the year?
- Ku Klux Klan, umpires and bowling: Why the minor leagues tossed out Jim Evans academy
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.
