PDA

View Full Version : Another Stat For Those of You who Flunked High School Algebra


SoxFan48
04-26-2005, 02:59 PM
Baseball Prospectus is pointing today that the White Sox success is due in large part to its Defensive efficiency rating which is tops in the American League and second in the majors.

http://www.baseballprospectus.com/statistics/def_eff2005.php

Chicago83
04-26-2005, 03:04 PM
What does this stat mean?

batmanZoSo
04-26-2005, 03:05 PM
Baseball Prospectus is pointing today that the White Sox success is due in large part to its Defensive efficiency rating which is tops in the American League and second in the majors.

http://www.baseballprospectus.com/statistics/def_eff2005.php

I didn't flunk algebra, but I'm more of an abstract thinker. How do they arrive at .7355? :?:

Flight #24
04-26-2005, 03:05 PM
What does this stat mean?

IIRC, it's percentage of balls put into play that are converted by the defense to outs.

SoxFan48
04-26-2005, 03:08 PM
Defensive Efficiency is a metric that measures a team's ability to turn balls-in-play into outs, using the formula...



(TotalOuts - Strikeouts)/(BIP-HR)

http://www.baseballprospectus.com/article.php?articleid=2387

Iwritecode
04-26-2005, 03:10 PM
IIRC, it's percentage of balls put into play that are converted by the defense to outs.

OK, how exactly does that differ from fielding pct?

SomebodyToldMe
04-26-2005, 03:11 PM
Am I the only person who thinks there are too.many.stats?!

batmanZoSo
04-26-2005, 03:12 PM
Defensive Efficiency is a metric that measures a team's ability to turn balls-in-play into outs, using the formula...





(TotalOuts - Strikeouts)/(BIP-HR)


http://www.baseballprospectus.com/article.php?articleid=2387



I see. Gotta give props to the pitching, too. Garland was dominant yesterday, but only had three strikeouts. Lots of easy outs. Teams aren't hitting the ball hard, it's all fly balls and two hoppers.

Flight #24
04-26-2005, 03:14 PM
OK, how exactly does that differ from fielding pct?

I believe fielding % is putouts+assists / total chances. So the major difference is in "total chances" versus all ball put into play. So if a ball is hit outside the range of a player, it wouldn't factor into FP, but it would factor into DE.

MRKARNO
04-26-2005, 03:14 PM
OK, how exactly does that differ from fielding pct?

Fielding percentage is a measure of the percentage of outs made on routine plays which should be outs.

Defensive efficiency is the percentage of non-homerun balls that are put into play which are turned into outs.

A double hit into the gap wouldnt lower your fielding percentage, but would lower your defensive efficiency

SoxFan48
04-26-2005, 03:15 PM
Pitching clearly factors into the equation heavily as the more outs the pitchers create, the higher Defensive Efficiency Rating.

Ol' No. 2
04-26-2005, 03:16 PM
What does this stat mean?It means that you can "explain" anything if you have enough statistics to choose from.

nebraskasox
04-26-2005, 03:20 PM
Am I the only person who thinks there are too.many.stats?!

For me, baseball can never have too many stats. It's part of the fun!

jabrch
04-26-2005, 03:22 PM
Lemme get this straight... They think a defense is more "efficient" if it gets more outs from balls in play? Um - that's the stupidest thing I have ever heard. First off, nobody will argue that more outs are better. And nobody will say that if you don't give up a HR or get a K that an out is the next best thing. But doesn't this directly conflict with DIPS - the stathead theory that you can take defense out of the equation in measuring performance of a pitcher? How the hell is a team D supposed to be blamed for not gettting a ball in the gap, or one down the line? Then you have DPs - where you get double credit despite...this is a joke.

There's a reason that Baseball Prospectus is such a pile of junk. "Stats" like this make no sense. They aren't measuring anything at all. I'd prefer tha AH/AP measurements...it is the ratio of Ahole size to armpit moisture. It's a very telling offensive stat...

I've said it a million times...these single combined stats that are supposed to aggreagte inputs and outcomes and rank players or teams based on a combination of skills and outcomes are a joke. Use basic, single variable stats and use a lot of them to evaluate a player against expectations - that's much more effective than trying to combine a bunch of different stats into one. And as far as defense goes - don't use stats to measure it - that's just too damn unreliable.

MeanFish
04-26-2005, 03:22 PM
OK, how exactly does that differ from fielding pct?

Fielding pct is different because it only counts balls that are gotten to and/or played by a defender. For instance, Carlos Lee had a fielding percentage of 1.000 last year. However, all of those balls he let drop in front of him would count against our team's Fielding Efficiency, since that particular stat includes every ball put into fair territory, period. Over a short period that stat can represent luck, but over time the distribution becomes more indicative of what's going on with a team. That stat can also be influenced by having a lot of factors, ranging from the defense's range and ability to field the ball to the pitching staff's ability to induce balls that are easily caught. Based on what I've seen this year, it's likely a combination of the two. The left side of our infield gets to everything, and the right side is doing pretty well too. The outfield doesn't let much of anything fall unless it's hit in just the right spot i.e. over the fence or a very shallow base-hit poke. We also have stocked up on ground ball pitchers, which helps that stat as well. Well, except for Freddy, anyway.

elrod
04-26-2005, 03:22 PM
DE clearly favors non-strikeout control pitchers and ground ball artists. This stat favors Buehrle, Garland and El Duque because they leave hitters off balance enough to hit the ball lightly somewhere.

Iwritecode
04-26-2005, 03:23 PM
OK, I wasn't quite sure how fielding % was calculated.

Thanks guys. :D:

ilsox7
04-26-2005, 03:24 PM
I feel bad for statheads. They spend so much time calculating stupid stats and don't have any time to enjoy their beer and hot dog and actually watch the game.

elrod
04-26-2005, 03:24 PM
Minnesota has given up a LOT of home runs. Second most in the AL after TB.

MRKARNO
04-26-2005, 03:36 PM
First off, nobody will argue that more outs are better. And nobody will say that if you don't give up a HR or get a K that an out is the next best thing. But doesn't this directly conflict with DIPS - the stathead theory that you can take defense out of the equation in measuring performance of a pitcher? How the hell is a team D supposed to be blamed for not gettting a ball in the gap, or one down the line? Then you have DPs - where you get double credit despite...this is a joke.


I will argue that more outs are better. Isnt it obvious that more outs are better? Isn't the point to get 27 outs from the opposition while trying to get them to score the fewest runs possible? You want outs and the more balls in play that are outs, the better.

DIPS stands for Defense Independent Pitching Stats. Strikeouts are defense independent, Homers are defense independent (with extremely rare exceptions), walks are defense independent. It's a way to measure what a pitcher would theoretically do with a neutral defense. It runs under the assumption that pitchers have no effect on defensive efficiency and on the percentage of balls hit into play that go for hits, something with which I do not agree and a large group of "statheads" do not agree with. It's a theory which Voros McCracken supports and not even Nate Silver, the inventer of VORP agrees with this. It is true for some pitchers, but not all pitchers in my view, but there are obvious exceptions to this rule.

DIPS is an attempt to measure a pitcher while defensive efficiency is the measure of the defense. It's not contradictory in any sense.

MeanFish
04-26-2005, 03:36 PM
It means that you can "explain" anything if you have enough statistics to choose from.

Is there anything wrong with that? Granted, there's a line where propellerheadism becomes really, really dumb. But there are also circumstances, such as this one, where a team performs in a way that was not previously foreseen. In that scenario, statistics can paint a useful, objective picture.

I know it can get annoying watching BP subjectively "bend" stats to favor teams whose GM's they like that day. However, that's not the fault of statistics, but rather the irresponsible use of them.

SoxFan48
04-26-2005, 03:38 PM
I've said it a million times...these single combined stats that are supposed to aggreagte inputs and outcomes and rank players or teams based on a combination of skills and outcomes are a joke. Use basic, single variable stats and use a lot of them to evaluate a player against expectations - that's much more effective than trying to combine a bunch of different stats into one. And as far as defense goes - don't use stats to measure it - that's just too damn unreliable.

...you are still wrong. Believe it or not Bill James and his stathead followers changed the way baseball is played. And the work of Baseball Prospectus and other baseball statistics writers will only get better. So, enjoy your beer and hot dog. I am going to continue to learn everything I can about the game I love--baseball.

JRIG
04-26-2005, 03:40 PM
I feel bad for statheads. They spend so much time calculating stupid stats and don't have any time to enjoy their beer and hot dog and actually watch the game.

I feel bad for people who go out of their way to make blanket statements in order to criticize others for the manner in which they enjoy watching and analyzing the game of baseball.

Ol' No. 2
04-26-2005, 03:40 PM
Is there anything wrong with that? Granted, there's a line where propellerheadism becomes really, really dumb. But there are also circumstances, such as this one, where a team performs in a way that was not previously foreseen. In that scenario, statistics can paint a useful, objective picture.

I know it can get annoying watching BP subjectively "bend" stats to favor teams whose GM's they like that day. However, that's not the fault of statistics, but rather the irresponsible use of them.Maybe I should explain what I meant. Most of the traditional stathead measures say the Sox should not be winning. If you sift through all the statistical measures you can always find one that fits the observations after the fact. Big deal. Explaining why the horse won after the race is easy. Predicting it before the race is harder.

MeanFish
04-26-2005, 03:42 PM
Maybe I should explain what I meant. Most of the traditional stathead measures say the Sox should not be winning. If you sift through all the statistical measures you can always find one that fits the observations after the fact. Big deal. Explaining why the horse won after the race is easy. Predicting it before the race is harder.

Gotcha :cool: That makes a lot more sense.

Hopefully the horse wins again tonight!

ilsox7
04-26-2005, 03:43 PM
...you are still wrong. Believe it or not Bill James and his stathead followers changed the way baseball is played. And the work of Baseball Prospectus and other baseball statistics writers will only get better. So, enjoy your beer and hot dog. I am going to continue to learn everything I can about the game I love--baseball.

I will continue to enjoy my beer and hot dog and I will consider stats relevant when you can explain how the Red Sox statistically came back from an 0-3 deficit. You cannot measure chemistry and luck, both of which go a lot farther in the game. Some stats are helpful, sure. But when you start trying to judge a player's range with numbers, you've crossed a line. When you tell me that a pitcher who goes 9 innings, gives up 5 hits and 1 run had a better outing than a guy who pitches a complete game shut-out and gives up 4 hits, you're crazy.

JRIG
04-26-2005, 03:43 PM
Maybe I should explain what I meant. Most of the traditional stathead measures say the Sox should not be winning. If you sift through all the statistical measures you can always find one that fits the observations after the fact. Big deal. Explaining why the horse won after the race is easy. Predicting it before the race is harder.

But if you do understand why the horse won the race, wouldn't it make sense that you can then use the information to either a) emulate success or b)"build a better horse" so to speak?

MRKARNO
04-26-2005, 03:44 PM
I feel bad for people who go out of their way to make blanket statements in order to criticize others for the manner in which they enjoy watching and analyzing the game of baseball.

I couldnt imagine going to a game and not thinking about what's actually taking place. I feel bad for those who go to the game and don't know things that are readily available to them that might greatly enhance their ability to know what's going on at a given time.

batmanZoSo
04-26-2005, 03:45 PM
Lemme get this straight... They think a defense is more "efficient" if it gets more outs from balls in play? Um - that's the stupidest thing I have ever heard. First off, nobody will argue that more outs are better. And nobody will say that if you don't give up a HR or get a K that an out is the next best thing. But doesn't this directly conflict with DIPS - the stathead theory that you can take defense out of the equation in measuring performance of a pitcher? How the hell is a team D supposed to be blamed for not gettting a ball in the gap, or one down the line? Then you have DPs - where you get double credit despite...this is a joke.

There's a reason that Baseball Prospectus is such a pile of junk. "Stats" like this make no sense. They aren't measuring anything at all. I'd prefer tha AH/AP measurements...it is the ratio of Ahole size to armpit moisture. It's a very telling offensive stat...

I've said it a million times...these single combined stats that are supposed to aggreagte inputs and outcomes and rank players or teams based on a combination of skills and outcomes are a joke. Use basic, single variable stats and use a lot of them to evaluate a player against expectations - that's much more effective than trying to combine a bunch of different stats into one. And as far as defense goes - don't use stats to measure it - that's just too damn unreliable.

I tend to agree, but observational analysis supports that we have been efficient in the field. Also, while we're number one in this particular stat, we're also number one in the standings. I wouldn't say it's completely useless but I've never been a proponent of defensive stats either.

ilsox7
04-26-2005, 03:49 PM
I feel bad for people who go out of their way to make blanket statements in order to criticize others for the manner in which they enjoy watching and analyzing the game of baseball.

All I said was I feel bad for those people who try to analyze every single piece of data to look for something that show why a team is winning or losing. Of course there are some basic indicators that help teams out. But why have the White Sox underperformed the last 4 years? Why have the Twins seemingly beat all odds the last few years? Because they have something (or in the Sox case do not have something) that cannot be measured.

Baseball is a game. It is to be enjoyed. If your way to enjoy it is to find a bunch of stats to justify why you think a team is winning or losing, then so be it. I'd much rather watch a game and try to figure out what pitch is going to be called next. Or what pitcher is hot or cold and should be brought in next. Or whether or not we should be bunting to move the runner over.

Ol' No. 2
04-26-2005, 03:50 PM
But if you do understand why the horse won the race, wouldn't it make sense that you can then use the information to either a) emulate success or b)"build a better horse" so to speak?Absolutely. There's nothing wrong with using statistical measures to get a better understanding of the game and to fine tune your team. But statistical analysis is a lot more complex that is generally understood. And what never seems to be appreciated in these stats is the complexity of baseball. The team with the best OBP does not always win. Ditto for DIPS, DE, OPS or whatever stat you can come up with. They're all gross oversimplifications.

This is a classic example. The most "predictive" statistics all say the Sox shouldn't be winning like they have been. The pitching has been lights-out and they've been getting very timely hitting. So they've been winning because of DEFENSIVE EFFICIENCY???

MRKARNO
04-26-2005, 03:50 PM
The issue is making sure that you are aware of the delicate balance between observational and statistical analysis. Does what you see correlate with what the statistics say? Sometimes the answer is no, but a lot of times the answer is, in fact, yes. But when the answer is no, you have to look closely at both to see which makes more sense to you. Maybe the stats will illustrate the observational? Or maybe it's vice-versa. Either way, one who follows either stats or observation blindly is just that, blind. It's about finding the right balance. Even WSI hero Billy Beane would tell you that.

ilsox7
04-26-2005, 03:51 PM
Absolutely. There's nothing wrong with using statistical measures to get a better understanding of the game and to fine tune your team. But statistical analysis is a lot more complex that is generally understood. And what never seems to be appreciated in these stats is the complexity of baseball. The team with the best OBP does not always win. Ditto for DIPS, DE, OPS or whatever stat you can come up with. They're all gross oversimplifications.

This is a classic example. The most "predictive" statistics all say the Sox shouldn't be winning like they have been. The pitching has been lights-out and they've been getting very timely hitting. So they've been winning because of DEFENSIVE EFFICIENCY???

Thank you for summing up my sentiments better than I have been able to do.

I just find it hard to imagine going to a game, keeping score, enjoying it, then after the last out sitting back and calculating things such as:(TotalOuts - Strikeouts)/(BIP-HR).

JRIG
04-26-2005, 03:53 PM
Absolutely. There's nothing wrong with using statistical measures to get a better understanding of the game and to fine tune your team. But statistical analysis is a lot more complex that is generally understood. And what never seems to be appreciated in these stats is the complexity of baseball. The team with the best OBP does not always win. Ditto for DIPS, DE, OPS or whatever stat you can come up with. They're all gross oversimplifications.

This is a classic example. The most "predictive" statistics all say the Sox shouldn't be winning like they have been. The pitching has been lights-out and they've been getting very timely hitting. So they've been winning because of DEFENSIVE EFFICIENCY???

Well, I think it's actually a pretty decent arguement. Buehrle and Garland have obscenely low batting average against numbers right now. And I think at least a part of it is due to the outstanding defensive play thus far, especially from our SS. It's not the entire answer, but taking it as a part of an answer may help to explain things.

Flight #24
04-26-2005, 03:53 PM
Thank you for summing up my sentiments better than I have been able to do.

I just find it hard to imagine going to a game, keeping score, enjoying it, then after the last out sitting back and calculating things such as:(TotalOuts - Strikeouts)/(BIP-HR).

But if you didn't do that, how would you know who won?

Ol' No. 2
04-26-2005, 03:55 PM
Thank you for summing up my sentiments better than I have been able to do.

I just find it hard to imagine going to a game, keeping score, enjoying it, then after the last out sitting back and calculating things such as:(TotalOuts - Strikeouts)/(BIP-HR).If people want to do that, that's great. They provide information that, when used correctly, is invaluable for understanding the game. I don't have the patience for that, but hey...whatever floats your boat.

Flight #24
04-26-2005, 03:56 PM
Well, I think it's actually a pretty decent arguement. Buehrle and Garland have obscenely low batting average against numbers right now. And I think at least a part of it is due to the outstanding defensive play thus far, especially from our SS. It's not the entire answer, but taking it as a part of an answer may help to explain things.

This is exactly the point. IMO it's not that statistics arenot useful, it's that they're improperly used. There are tons of reports that basically say "Sox can't keep winning because they have a league-trailing OBP". That's an incorrect usage of that stat because winning is NOT correlated with high OBP.

The Sox are winning because they have excellent pitching, strong D, and specifically, a very good bullpen. Thus games are likely to be close. And in close games, it's teams that execute and don't make mistakes that tend to win. So yes, it's unlikely that they'll play .800 ball throughout the season, but it's completely likely that they'll continue to win more than they lose.

ilsox7
04-26-2005, 03:56 PM
Well, I think it's actually a pretty decent arguement. Buehrle and Garland have obscenely low batting average against numbers right now. And I think at least a part of it is due to the outstanding defensive play thus far, especially from our SS. It's not the entire answer, but taking it as a part of an answer may help to explain things.

OK, but here's the thing: how many truly AWESOME plays has Uribe made? Sure, he has made some very good big league plays. And he has done so without making many errors. But one would argue that most other shortstops in the game would be making the mostly routine plays he has made. So how has having Juan Uribe there as opposed to someone else made a significant difference. Also, can you tell me if we had another SS there if MB's BA against owuld be better or worse?

I'd argue that their BA against is much more a product of them getting ahead in the count and throwing better pitches than anything our defense has done. Sure, the defense has contributed, but I'd put more of the credit with the guys takign the bump.

MRKARNO
04-26-2005, 03:59 PM
Thank you for summing up my sentiments better than I have been able to do.

I just find it hard to imagine going to a game, keeping score, enjoying it, then after the last out sitting back and calculating things such as:(TotalOuts - Strikeouts)/(BIP-HR).

When I go to a game, I don't think about all of those abstract numbers because the sample size is so small as to be irrelevent. I doubt many people sit at games and calculate such abstract stats. Most of what is called performance analysis is done in between games and is used to identify trends and to form opinions about specific players. Do I think about the likelihood of scoring runs in certain situations? Yes as it's pretty critical to the game itself. But does anyone really think about the Reliever Expected Wins while watching a game? Maybe a few, but probably not even most "statheads" or "Friends of Billy Beane." We're normal fans, we just look at the performance a little further beyond "His fastball was at 96 MPH, he's a solid pitcher."

MRKARNO
04-26-2005, 04:01 PM
This is exactly the point. IMO it's not that statistics arenot useful, it's that they're improperly used. There are tons of reports that basically say "Sox can't keep winning because they have a league-trailing OBP". That's an incorrect usage of that stat because winning is NOT correlated with high OBP.


I agree. The White Sox might be winning in spite of their low OBP, but that doesnt mean that they cannot win with a low OBP. It does mean that they have to play good defense, have good pitching and make sure that the fewer runners you do get on make it to home plate more efficiently than normal, but it doesn't mean that they cannot win with this.

ilsox7
04-26-2005, 04:03 PM
When I go to a game, I don't think about all of those abstract numbers because the sample size is so small as to be irrelevent. I doubt many people sit at games and calculate such abstract stats. Most of what is called performance analysis is done in between games and is used to identify trends and to form opinions about specific players. Do I think about the likelihood of scoring runs in certain situations? Yes as it's pretty critical to the game itself. But does anyone really think about the Reliever Expected Wins while watching a game? Maybe a few, but probably not even most "statheads" or "Friends of Billy Beane." We're normal fans, we just look at the performance a little further beyond "His fastball was at 96 MPH, he's a solid pitcher."

OK, that's cool. I guess what sets me off are these defensive stats that try to measure a player's range. Or something that tries to measure how "quality" of an outting a pitcher had. Some things, IMO, need to be left for obersvational analysis. I just think that if you looked at all of the stats out there right now, the vast, vast majority would show that the Sox should be a .500 team right now. Again, it's a small sample size. But Pennant Races and playoffs come down to small sample sizes, and that's what I care about.

JRIG
04-26-2005, 04:04 PM
OK, but here's the thing: how many truly AWESOME plays has Uribe made? Sure, he has made some very good big league plays. And he has done so without making many errors. But one would argue that most other shortstops in the game would be making the mostly routine plays he has made. So how has having Juan Uribe there as opposed to someone else made a significant difference. Also, can you tell me if we had another SS there if MB's BA against owuld be better or worse?


Throwing aside all stats (incredible, I know) and just going on observations, Uribe has been brilliant defensively this year. Last night alone, he made two, if not three, plays that I don't think 99% of shortstops make. He's been an incredibly valuable defensive player this year and has made a difference.

ilsox7
04-26-2005, 04:09 PM
Throwing aside all stats (incredible, I know) and just going on observations, Uribe has been brilliant defensively this year. Last night alone, he made two, if not three, plays that I don't think 99% of shortstops make. He's been an incredibly valuable defensive player this year and has made a difference.

I agree that he has been very good and has probably made some difference. My argument is that the main driving force behind our pitchers' success this year is their pitching, not Juan Uribe. You will probably dispute this, but I think you've proven my point somewhat by saying that just by watching Uribe this year, he's been great. I agree. But there is no statistical way to measure that. We cannot possibly say that Uribe has prevented X amount of runs. Or another shortstop would have meant we would have allowed X amount of hits more. That's my point.

jabrch
04-26-2005, 04:12 PM
Stats are great when they make sense. I have a stats minor - so I am fairly literate in understanding how to use stats. These defensive stats, including DIPS and DE are ridiculous. There are just too many variables that they fail to consider and there are too many scenarios that don't fit.

There are a lot more than one ways to win a baseball game. When you try and use one measure of goodness or efficiency, you are concluding that there is one way to win. This just isn't true. Stats are fine. Statisticians are great. Statheads are complete and total morons with no clue about what they are talking about - except what they spew back that came from some calculatorhead somewhere.

Ol' No. 2
04-26-2005, 04:13 PM
Throwing aside all stats (incredible, I know) and just going on observations, Uribe has been brilliant defensively this year. Last night alone, he made two, if not three, plays that I don't think 99% of shortstops make. He's been an incredibly valuable defensive player this year and has made a difference.No doubt about it. Just counting the number of great plays is misleading, since they may be coming at critical times. Same for hits, SO, or anything else you measure. Baseball is such a complex game that no one stat or even a handful can hope to capture it. And stats don't say anything about balance. Scott Podsednik has the highest OBP of any of the regulars on the team, but would you want 9 Scott Podsedniks?

ilsox7
04-26-2005, 04:15 PM
No doubt about it. Just counting the number of great plays is misleading, since they may be coming at critical times. Same for hits, SO, or anything else you measure. Baseball is such a complex game that no one stat or even a handful can hope to capture it. And stats don't say anything about balance. Scott Podsednik has the highest OBP of any of the regulars on the team, but would you want 9 Scott Podsedniks?

Exactly!

JRIG
04-26-2005, 04:18 PM
No doubt about it. Just counting the number of great plays is misleading, since they may be coming at critical times. Same for hits, SO, or anything else you measure. Baseball is such a complex game that no one stat or even a handful can hope to capture it. And stats don't say anything about balance. Scott Podsednik has the highest OBP of any of the regulars on the team, but would you want 9 Scott Podsedniks?

Not sure about that, but I'd win a fair number of games with 9 Frank Thomas clones. :cool:

ExpoPuddingHead
04-26-2005, 04:21 PM
You're being sarcastic i hope, a team of Big hurts would end up like the bears offensive line

jabrch
04-26-2005, 04:26 PM
Not sure about that, but I'd win a fair number of games with 9 Frank Thomas clones. :cool:

I want to see Big Hurt play SS, CF and 3B. How bout running on the Hurt in RF? I wanna see the Hurt-Hurt-Hurt 6,4,3. 9 Hurts would score a lot of runs - but 9 hurts in the field....That would hurt.