SOSH regresses to the mean. Or does it?

Shane

lurker
Nov 26, 2014
110
NDame616 said:
 
The bolded is the most important part. He was arguably the team's best SP last season. People will try to parse it with "Buchholz is better but doesn't start enough" and "Kelly has nasty stuff when he's on" and "Miley is a horse" and "Porcello really turned it on at the end of the year" etc...but all of this fails to address the main point: Eduardo Rodriguez was one of, if not THE, best pitcher on our staff last year. 
 
Unless we pick up Price, Cueto and trade for guys like Gray/Sale etc, there is no reason for Edro to not  start the season in our rotation.
Don't you expect at least some regression from him next year? And I really don't like any of the pitchers in their rotation. What's stopping them from signing Price and then trading for Sale and Ross? Sale, Price, Ross, Buchholz, Porcello with Eduardo coming up when Buch inevitably gets hurt takes this team to the World Series with a solid lineup
 

Buzzkill Pauley

Member
SoSH Member
Jun 30, 2006
10,569
Shane said:
I actually agree with you. I think both Owens and Eduardo have been pretty good this year, but they'll both likely have some regression next year, and are not going to pitch like aces. Both also have things to work on (Eduardo: third pitch, Owens: control). We have enough guys to start the year without either of them. And it's almost certain that someone will get hurt, in which case they'll both be ready to join the rotation. I think starting both in AAA makes the most sense for all involved.
 
Do you even know what "regression" means?
 
If you meant it statistically, then I will point out that in 2015, the spread between Rodriguez's ERA and xFIP was only -0.15. The spread between his ERA and FIP was a mere +0.07. The spread between his BABIP and the MLB average was -0.006. The spread between his HR/FB rate and the MLB average was -1.0%.  How do you statistically regress his numbers and come up with any projection materially different than last season's results?
 
If you meant it like he would return to a prior performance baseline, then it bears reminding that Rodriguez just finished his age-22 season and has no prior experience at the MLB level. Those facts mean we can't assess whether the past 21 games set his baseline for future performance, or simply established a starting point from which his career will develop positively or negatively. 
 
If you meant it to smarten up an opinion that you think he's going to be worse next year than this year, then here's some lipstick for your pig.
 

DJnVa

Dorito Dawg
SoSH Member
Dec 16, 2010
39,687
Shane said:
I actually agree with you. I think both Owens and Eduardo have been pretty good this year, but they'll both likely have some regression next year,
 
Speaking of Eduardo here--regression to what? Regression doesn't mean "he'll get worse even though his FIP and xFIP was pretty close to his actual ERA".
 
 
 

Shane

lurker
Nov 26, 2014
110
Buzzkill Pauley said:
 
Do you even know what "regression" means?
 
If you meant it statistically, then I will point out that in 2015, the spread between Rodriguez's ERA and xFIP was only -0.15. The spread between his ERA and FIP was a mere +0.07. The spread between his BABIP and the MLB average was -0.006. The spread between his HR/FB rate and the MLB average was -1.0%.  How do you statistically regress his numbers and come up with any projection materially different than last season's results?
 
If you meant it like he would return to a prior performance baseline, then it bears reminding that Rodriguez just finished his age-22 season and has no prior experience at the MLB level. Those facts mean we can't assess whether the past 21 games set his baseline for future performance, or simply established a starting point from which his career will develop positively or negatively. 
 
If you meant it to smarten up an opinion that you think he's going to be worse next year than this year, then here's some lipstick for your pig.
Sorry...maybe a bad word choice. I meant that after teams have seen him for a whole year, won't they have better scouting reports on him? They become more likely to know about stuff (example: he relies heavily on his fastball, terrible breaking ball)
 

smastroyin

simpering whimperer
Dope
SoSH Member
Jul 31, 2002
20,684
People mis-use regression enough that I would say they will add "get worse" as a colloquial definition within a few years.  It does still drive me crazy though.
 
I think Rodriguez's September was a hopeful indication that he won't take a big step back.  I don't think he is going to be an ace, though it's still within the vast realm of possibility.
 

Buzzkill Pauley

Member
SoSH Member
Jun 30, 2006
10,569
Devizier said:
It's a fine word choice; regression has two meanings and context is usually sufficient.
 
And yet, neither of those meanings is "become worse."
 
I'm with smas on this. It drives me nuts, especially when misused on a messageboard holding itself out as a stats-minded community.
 

Devizier

Member
SoSH Member
Jul 3, 2000
11,208
Somewhere
Buzzkill Pauley said:
 
And yet, neither of those meanings is "become worse."
 
I'm with smas on this. It drives me nuts, especially when misused on a messageboard holding itself out as a stats-minded community.
 
This is a "nauseous versus nauseated" scenario.
 
In this sense, regress is being used in place of retrogress.
 
In any event, the context makes the word use clear.
 

Minneapolis Millers

Member
SoSH Member
Jul 15, 2005
2,069
Twin Cities
Math geeks should mingle more with English majors.  The world would be better place generally.  Specifically, we would probably end up better contextualizing word usage.  When the directions toward the mean are generally perceived as either positive or negative (had a 5+ ERA but should "regress" back toward his usual 4.25, etc.), we could use "regress toward the mean" when the movement would be seen as a bad thing and "progress toward the mean" when it would represent an improvement.  ("Retrogress" doesn't improve the cocktail or message board conversation much.)  Just a suggestion.  Carry on.
 

smastroyin

simpering whimperer
Dope
SoSH Member
Jul 31, 2002
20,684
regress and progress are not antonyms within the context you present.
 
What you really mean is "math geeks and English majors should conspire to make new meanings for words based on what I think."
 
regress and progress are antonymish in the following usage:
 
"Eduardo has progressed from being a single pitch thrower into a 3 pitch pitcher."
"Eduardo has regressed to his bad habits of simply trying to overpower hitters."
 
That we can't have a distinct conversation about this without talking about statistical regression is kind of weird.
 

Reverend

for king and country
Lifetime Member
SoSH Member
Jan 20, 2007
45,086
Minneapolis Millers said:
Math geeks should mingle more with English majors.  The world would be better place generally.  Specifically, we would probably end up better contextualizing word usage.  When the directions toward the mean are generally perceived as either positive or negative (had a 5+ ERA but should "regress" back toward his usual 4.25, etc.), we could use "regress toward the mean" when the movement would be seen as a bad thing and "progress toward the mean" when it would represent an improvement.  ("Retrogress" doesn't improve the cocktail or message board conversation much.)  Just a suggestion.  Carry on.
 
Except regression to the mean is not necessarily a bad thing--it can denote a player who under-performed compared to his "true" mean talent level but then regressed, i.e. moved closer to that true mean the following year.
 
But most fans usually don't use it that way because they associate it with the bad connotation which is incorrect, which is precisely what smas and others are trying to explain here.
 
Here's is FanGraphs's explanation: regression to the mean.
 

tomdeplonty

Member
SoSH Member
Sep 23, 2013
585
Easiest for me: statistical "regress" = "to perform more normally" (neither unusually poorly, or well)
 

Minneapolis Millers

Member
SoSH Member
Jul 15, 2005
2,069
Twin Cities
smastroyin said:
regress and progress are not antonyms within the context you present.
 
What you really mean is "math geeks and English majors should conspire to make new meanings for words based on what I think."
 
regress and progress are antonymish in the following usage:
 
"Eduardo has progressed from being a single pitch thrower into a 3 pitch pitcher."
"Eduardo has regressed to his bad habits of simply trying to overpower hitters."
 
That we can't have a distinct conversation about this without talking about statistical regression is kind of weird.
Your linguistic rigidity proves my point, which, by the way, was mostly tongue in cheek. But whatever.
 

Reverend

for king and country
Lifetime Member
SoSH Member
Jan 20, 2007
45,086
Minneapolis Millers said:
Your linguistic rigidity proves my point, which, by the way, was mostly tongue in cheek. But whatever.
No, it doesn't. It proves he disagrees.

Terms evolve, but it's often spurred by confusion, not beneficial innovations in usage. As such, it the process can reduce clarity and undermins some of the purposes of language.

And I mean that literally.
 

iayork

Member
SoSH Member
Apr 6, 2006
639
Minneapolis Millers said:
Your linguistic rigidity proves my point, which, by the way, was mostly tongue in cheek. But whatever.
 
And by "rigidity", I mean "fluidity".
 

Savin Hillbilly

loves the secret sauce
SoSH Member
Jul 10, 2007
18,739
The wrong side of the bridge....
Buzzkill Pauley said:
 
And yet, neither of those meanings is "become worse."
 
I'm with smas on this. It drives me nuts, especially when misused on a messageboard holding itself out as a stats-minded community.
 
On the contrary, as a quick look at some common dictionaries will show.
 
Encarta (print): "A going backwards or a backward movement or progress."
 
Merriam-Webster: "a trend or shift toward a lower or less perfect state"
 
Dictionary.com: "retrogradation" (the definitions of which include "decline or deterioration")
 
The root meaning is simply "moving backward", which contains the idea of getting worse. It doesn't have to mean going back toward any specific target or standard.
 
I run into this stuff all the time on my job (working with scientists). When people get used to using a word in a special sense, they assume that that's the only correct sense, and start complaining when people use it in another, perfectly valid non-technical sense.
 
And Minneapolis Millers' suggestion is excellent: when you're using it statistically, say "regression to ______" to make that clear. 
 

Buzzkill Pauley

Member
SoSH Member
Jun 30, 2006
10,569
Maybe English majors should mingle more with Math geeks. The world would be a better place generally. Specifically, we would probably end up better distinguishing word usage.
 

smastroyin

simpering whimperer
Dope
SoSH Member
Jul 31, 2002
20,684
You guys talk about context matters but the only context you care about is "this is what *I* think."  
 
Decline or deterioration implies a return to a previous state.  Since Eduardo Rodriguez is the case in point for this discussion, there is nothing for him to "regress" to because this year he was ex-fucking-actly what you would expect, except for maybe that his adjustment period to the majors featured an amazing hot start.  But if we are going to use progression and regression for every hot streak or slump we may as well give up now because we've made the words meaningless.
 
All we are asking is that people stop using regression as a synonym for "Gets worse."  Some guys do perform over their heads for many different reasons, and talk about regression to mean may be appropriate (key: may be) so if you use it every single time you want to talk about worries a player will get worse, it confounds the discussion instead of adding to it.
 

Savin Hillbilly

loves the secret sauce
SoSH Member
Jul 10, 2007
18,739
The wrong side of the bridge....
smastroyin said:
You guys talk about context matters but the only context you care about is "this is what *I* think."
I think this is pretty funny, considering that I'm the only person in this argument who has tried to provide any objective evidence.
 
Decline or deterioration implies a return to a previous state.
 
This is not true, of course (my teeth are declining and deteriorating to a state they've never been in before, sadly)--but I agree that "regression" should be used for declines or deteriorations that involve backward movement, whether toward an actual previous state or a state that's perceived as inherently less advanced. And athletic declines always fall into this category, since athletic skill is an advance over the default human condition.
 
Look, if you want to lay it down as a SoSH best practice that "regression" should only be used in its statistical sense, fine by me (probably a good idea to add this to a sticky if it isn't already there). But it's simply wrong to claim that this is necessary to conform to correct English usage, and the point of the dictionary references was to show that.
 

BroodsSexton

Member
SoSH Member
Feb 4, 2006
9,372
guam
Savin Hillbilly said:
I think this is pretty funny, considering that I'm the only person in this argument who has tried to provide any objective evidence.
 
 
This is not true, of course (my teeth are declining and deteriorating to a state they've never been in before, sadly)--but I agree that "regression" should be used for declines or deteriorations that involve backward movement, whether toward an actual previous state or a state that's perceived as inherently less advanced. And athletic declines always fall into this category, since athletic skill is an advance over the default human condition.
 
Look, if you want to lay it down as a SoSH best practice that "regression" should only be used in its statistical sense, fine by me (probably a good idea to add this to a sticky if it isn't already there). But it's simply wrong to claim that this is necessary to conform to correct English usage, and the point of the dictionary references was to show that.
 
Yes, clearly what this site needs is a sticky note about how to use the words "regress" and "regression."  That, my friends, is progress behind which we can all get. 
 
The offseason is on.
 

metaprosthesis

Member
SoSH Member
May 22, 2008
199
Central NJ via Western Mass
When math geeks and English majors mingle, you end up with linguists.  We don't make anything clearer or better.  Also, we don't try to shoehorn language into some "perfect" state, nor do we regard language change as any kind of loss of communicative capacity.

I think that the issue here is partially one of assuming that dictionaries are sources of complete usage information.  In actuality most non-OED dictionaries simply aim to provide a word sense (and this is not a knock on lexicographers, who face a daunting task).  While I don't think that "decline" or "deterioration" require any idea of return to a previous state, "regress" does have this requirement.  In baseball terms, think of career trajectory.  As a player gets older, his skills may decline or deteriorate, but that does not mean that they pass through former skill levels.  Some skills or abilities are simply lost, and a new less-productive state is created.  When a player regresses, however, this is understood as returning to a previous performance level.  Something cannot regress to a state that has not existed.  In the statistical sense, the state may not have been actualized, but is shown through analysis-- an empirically expected state.  A decline in Eduardo Rodriguez's production is not expected based on any observable grounds.  For some it may be expected based on some pessimistic subjective grounds (e.g. "young pitchers are never as good as they look in small rookie samples.").  I think that the willingness to accept an uninformed (no offense) opinion as a possible state is the source of the offending usage of regress/regression.  And if that is true, then smas' point about people's context being "what *I* think" is on point.

tl;dr: "Regression" entails a return to a previous or expected state.  Traditionally, "expected" entails "based on empirical evidence".  Some people accept their impressions of things as empirical evidence.
 
 

mauf

Anderson Cooper x Mr. Rogers
Staff member
Dope
According to Dictionary.com, the first definition of regression is "a return to a former or less developed state." This is the sense in which the word is used in medicine (unless you're talking about statistical analysis in the context of medical research).
 
The second definition is "a measure of the relation between the mean value of one variable and corresponding values of other variables." The phrase "regression to the mean" uses regression in this sense.
 
I don't think the OP's use of the word was necessarily wrong -- saying he expects E-Rod to "regress" is the same as saying he expects him to "return to a former or less developed state." I think he was wrong, however, to express that opinion without explaining why he expects a player in his age-23 season to get worse. Players in their early 20s aren't normally expected to decline, and there isn't something obvious in E-Rod's statistics or pedigree that would make us think he's an exception.
 

Buzzkill Pauley

Member
SoSH Member
Jun 30, 2006
10,569
I don't understand the new thread title.  Without a clear definition for "regress," how can "mean" be understood in context? 
 
"mean" like "spiteful"...
 
"mean" like "average"...
 
"mean" like "signifying"...
 
"mean" like "intending"...
 
"mean" like "shabby"...
 
"mean" like "greedy"...
 
I mean, unless we know what mean means, mean's mean. And that's just being mean.
 

metaprosthesis

Member
SoSH Member
May 22, 2008
199
Central NJ via Western Mass
Buzzkill Pauley said:
I don't understand the new thread title.  Without a clear definition for "regress," how can "mean" be understood in context? 
 
"mean" like "spiteful"...
 
"mean" like "average"...
 
"mean" like "signifying"...
 
"mean" like "intending"...
 
"mean" like "shabby"...
 
"mean" like "greedy"...
 
I mean, unless we know what mean means, mean's mean. And that's just being mean.
The fact that it follows a determiner tells us we can eliminate all non-nominal options.  That only leaves one choice.  (Yes, I know that you are making a funny.)
 

metaprosthesis

Member
SoSH Member
May 22, 2008
199
Central NJ via Western Mass
maufman said:
 
I don't think the OP's use of the word was necessarily wrong -- saying he expects E-Rod to "regress" is the same as saying he expects him to "return to a former or less developed state." I think he was wrong, however, to express that opinion without explaining why he expects a player in his age-23 season to get worse. Players in their early 20s aren't normally expected to decline, and there isn't something obvious in E-Rod's statistics or pedigree that would make us think he's an exception.
 
I think this is the problem, though.  There is no former, less developed state.  He's certainly not going to get younger.  Explaining why he expects EdRo to get worse would require reference to a non-existent prior state.
 

DLew On Roids

guilty of being sex
Lifetime Member
SoSH Member
Aug 30, 2001
13,825
The Pine Street Inn
metaprosthesis said:
When math geeks and English majors mingle, you end up with linguists.  We don't make anything clearer or better.  Also, we don't try to shoehorn language into some "perfect" state, nor do we regard language change as any kind of loss of communicative capacity.
 
Are you trying to claim that there are no prescriptivist linguists?
 
I've shaken my head quietly at this conversation so far, but this claim MUST NOT STAND.
 

metaprosthesis

Member
SoSH Member
May 22, 2008
199
Central NJ via Western Mass
DLew On Roids said:
 
Are you trying to claim that there are no prescriptivist linguists?
 
I've shaken my head quietly at this conversation so far, but this claim MUST NOT STAND.
 
Well, my experience is entirely with generativists and prescriptivism is pretty much counter to the basic tenets of generative linguistics.  That's not to say that people don't have their personal style preferences, but I have not encountered linguists who are grammarian preachers.
 

alwyn96

Member
SoSH Member
Aug 24, 2005
1,351
My opinion, which certainly no one has asked for, is that we try to interpret each other charitably and address the intention of people's arguments rather than get worked up about whether someone has used a word incorrectly or not. I don't think this site is improved by more name-calling getting bogged down in grumpy semantic debates. I have a dream where we can discuss whether we expect Rodriguez to be better or worse next year than he was this year without taking an INTERNET STAND about using the formal statistical definition of 'regression' or a more colloquial version.
 
I mean, I'm not going to tell people how to live their lives. If you want to feel enraged by someone using a word in a nonstandard way, go for it. I certainly have plenty of peccadillos that get me worked up. But it's a discussion that rarely seems to go anywhere interesting or new, and to me a lot of it seems to be more about internet point-scoring than considered discussion.
 

benhogan

Baynes Hogan (pending trade)
SoSH Member
Nov 2, 2007
8,598
Santa Monica
DLew On Roids said:
 
Are you trying to claim that there are no prescriptivist linguists?
 
I've shaken my head quietly at this conversation so far, but this claim MUST NOT STAND.
I've come across connie linguists many times, but have not encountered prescriptivist linguists as of yet.  I look forward to it.
 
Yes, the off-season has begun.
 

dbn

Member
SoSH Member
Feb 10, 2007
7,781
La Mancha.
I actually pulled out my print OED (Vol II, Q-Z) and was squinting at the tiny little print, reading the various definitions, when I took a hard look at myself and realized how sad this all was. Then I went back to my day-time drinking and watching American Dad reruns to feel better about myself.
 

Devizier

Member
SoSH Member
Jul 3, 2000
11,208
Somewhere
My opinion, which certainly no one has asked for, is that we try to interpret each other charitably and address the intention of people's arguments
That was my initial point. Pedantry sucks.
 

Savin Hillbilly

loves the secret sauce
SoSH Member
Jul 10, 2007
18,739
The wrong side of the bridge....
terrisus said:
 
"Encarta" and "Common?"
What year is this?
 
Hey, it was common when we got it as a Christmas present!
 
I agree that we don't need more sticky notes. I'd rather we correct each other only when we're making actual errors that cause actual confusion. But if we're going to invent unique-to-SoSH usage rules they should be posted somewhere so people who want to avoid a knuckle-whacking for doing it wrong know where to look.