NFL Super Bowl XLIX: Lick the Gun

Elki

Choose.

Also, don't get nasty. WE HAVE SPOKEN.

jungleroomx

Pats at 100%. Obviously an outlier and cheating is involved!

Elki

Sorry dudes, stealing the OP 'cause users can't make polls.

Elki

Anyways... I accidentally voted for the wrong team.

jungleroomx

Oh it's good Elki. I forgot about the poll.

jungleroomx

Oh, you shoulda went Legend of the Chronicles of Superbowl XLIX: World at War: Awakenings: Origins

Priest

I grudgingly want the Seahawks to win, but I am betting the Patriots will win.

ObiFett

I demand a third option of hoping both teams somehow lose.

Ilpala

Yea, I want 49ers/Ravens all over again, except the lights don't come back on.

ObiFett

Ilpala wrote: »

Yea, I want 49ers/Ravens all over again, except the lights don't come back on.

Yes

And I want it to happen while the Patriots are losing so everyone assumes that Belicheat somehow paid the electrical dudes to do it.

PantsB

Balls

ObiFett

@Elki

Allow the haters to hate

y2jake215

I am surprised the patriots are winning so far, with there being more Seahawks fans + everyone hates the pats

Elki

This poll is rigged! The corruption must run deep.

Doctor Detroit

y2jake215 wrote: »

I am surprised the patriots are winning so far, with there being more Seahawks fans + everyone hates the pats

Yes, but everyone else hates the Seahawks. And some hate both, as we've seen.

I look forward to Sunday night, when America finds out which team's fans will be more insufferable for the next year.

Carpy

On yesterday's (Wednesday's) Grantland nfl podcast they had a pretty interesting conversation about what sports reporting should be in 2014. While some of it was old media vs new bickering, they take a lot of jabs at the mindset that a column without a quote is a dead column, it was interesting to hear a group of professionals talk about the evolution of their field as traditional paradigms aren't necessarily relevant anymore.

The main thing they all said is that they approach a story on a game looking to highlight things you don't get on tv, whether the actual gamecast or the infinite repetition of sportscenter. As a consequence of this they found themselves asking more yes/no type questions of athletes because the player is more likely to answer and you can use them to pick out actual tactical decisions and that they use almost no quotes when writing about a game. It's a good listen and I'll post the link and timestamp when I get home, but it's towards then end when they start their interview with Brian Curtis.

Goumindong

And whomever posted the "did you read them" stuff yes. I did and the complaints were to put it shortly wrong. Deadspin is wrong on the epistemological aspects of what makes statistics believable. Additionally it's wrong on what data you should use (because we are explicitly wanting to look at the deflation issue it is not sufficient to say they use different balls for special teams and then throw that fumble data out. The fact that they're using balls controlled by the refs and inflated to the proper levels is what we are interested in)

All of them are wrong on the distribution (we know the distribution because we know the event, if it's not conforming there is a problem) issues. In this case 1/x is the avg of a geometric distribution, which is normal (though each one will have slightly different variances due to total test numbers).

The guy who looks at outdoor games is comparing events with significantly different variances and suggesting they have the same. Which negates the order issue of claims and also in this instance ntroduces significant omitted variable bias*. The guy who throws out the pass data is doing the same.

The people doing the normalcy tests don't actually understand the tests. If the pats are a legitimate outlier(ie not drawn from the same distribution) the tests should fail. So saying the data isn't normal when the question we are asking is "are the pats on the same distribution" actually answers the question in the other direction.

*technically there is ovb in the first data set too, but it should favor the patriots so it's not too much of an issue if you find that the pats are still out of bounds.

The first shot at it isn't perfect, because the question is not "are the pats anomalous" but "is it likely we would find one team so far out". And because there are data collection issues with how you handle inverting. But the answer to the second question is still an emphatic "yes" (at about a 1.5% p-value for the proper open air teams stat over five years not examining any other five year period [note that this is a five year period which includes the pats high fumble season]) and since we don't really have a reason to believe that the null of "the pats are on the same distribution as anyone else" we should definitely reject the hypothesis and the numbers do indeed suggest that it's worth investigating what type of effect this could have had.

davidsdurions

Ahh new thread smell.....oh wait, this stink like shit like the last one.

STAHHHHHPPPPP

Pellaeon

I hate the seahawks. But I expect them to win. Because of course they fucking will.

quovadis13

ObiFett wrote: »

I demand a third option of hoping both teams somehow lose.

Who are we rooting for??

davidsdurions

Patriots fans should root for the Seahawks so that there can be a reason for ol' Tommy boy to return for revenge next season. A rematch next year where the Patriots prevent the Seahawks from achieving a 3peat would be pretty good for that movie you guys were talking about before.

y2jake215

Goumindong wrote: »

And whomever posted the "did you read them" stuff yes. I did and the complaints were to put it shortly wrong. Deadspin is wrong on the epistemological aspects of what makes statistics believable. Additionally it's wrong on what data you should use (because we are explicitly wanting to look at the deflation issue it is not sufficient to say they use different balls for special teams and then throw that fumble data out. The fact that they're using balls controlled by the refs and inflated to the proper levels is what we are interested in)

All of them are wrong on the distribution (we know the distribution because we know the event, if it's not conforming there is a problem) issues. In this case 1/x is the avg of a geometric distribution, which is normal (though each one will have slightly different variances due to total test numbers).

The guy who looks at outdoor games is comparing events with significantly different variances and suggesting they have the same. Which negates the order issue of claims and also in this instance ntroduces significant omitted variable bias*. The guy who throws out the pass data is doing the same.

The people doing the normalcy tests don't actually understand the tests. If the pats are a legitimate outlier(ie not drawn from the same distribution) the tests should fail. So saying the data isn't normal when the question we are asking is "are the pats on the same distribution" actually answers the question in the other direction.

*technically there is ovb in the first data set too, but it should favor the patriots so it's not too much of an issue if you find that the pats are still out of bounds.

The first shot at it isn't perfect, because the question is not "are the pats anomalous" but "is it likely we would find one team so far out". And because there are data collection issues with how you handle inverting. But the answer to the second question is still an emphatic "yes" (at about a 1.5% p-value for the proper open air teams stat over five years not examining any other five year period [note that this is a five year period which includes the pats high fumble season]) and since we don't really have a reason to believe that the null of "the pats are on the same distribution as anyone else" we should definitely reject the hypothesis and the numbers do indeed suggest that it's worth investigating what type of effect this could have had.

i'm only responding so i don't get fined

So It Goes

lol Oregon is seahawks nation guys, we ain't got an NFL team

y2jake215

So It Goes wrote: »

lol Oregon is seahawks nation guys, we ain't got an NFL team

also there are more states in new england

Madpoet

Next year is the franchise sellout where Tom Brady defeats the Russians.

LoserForHireX

ObiFett wrote: »

I demand a third option of hoping both teams somehow lose.

Indeed, where is the option for the meteor?

Marathon

Voted Seahawks, the lesser of two evils

Element Brian

the best part of this is the underhanded burn on patriots fans

http://www.king5.com/story/news/local/2015/01/28/12th-pack-marijuana-cannabis-12000-joints/22497465/

LoserForHireX

y2jake215 wrote: »

I am surprised the patriots are winning so far, with there being more Seahawks fans + everyone hates the pats

I think that I might slightly be more okay with the Pats winning.

I mean, don't get me wrong, your city's sports fans are objectively terrible. Not as bad as Philly, but fucking awful none the less.

But. Tom Brady is good, and I'm not sure I mind him joining the ranks of Joe Montana for most winningest super bowl QB.

But I'd like him to retire, and Belly-Check to finally be dragged back to hell (from which he sprung)

Element Brian

also @Preacher this ksk link is just for you

http://kissingsuzykolber.uproxx.com/2015/01/celebrity-super-bowl-pickkake-voxs-13-charts-and-maps-that-explain-the-super-bowl.html

HappylilElf

Madpoet wrote: »

Next year is the franchise sellout where Tom Brady defeats the Russians.

excuse me did you just refer to the epic finale of the rocky films as a "franchise sellout"

y2jake215

Element Brian wrote: »

the best part of this is the underhanded burn on patriots fans

http://www.king5.com/story/news/local/2015/01/28/12th-pack-marijuana-cannabis-12000-joints/22497465/

washington state follows their teams example and treats drugs as legal tsk tsk

y2jake215

HappylilElf wrote: »

Madpoet wrote: »

Next year is the franchise sellout where Tom Brady defeats the Russians.

excuse me did you just refer to the epic finale of the rocky films as a "franchise sellout"

ooh ooh does that make 2030 when all of these players are stumbling around suffering from CTE Rocky 5???

Shadowhope

y2jake215 wrote: »

Goumindong wrote: »

And whomever posted the "did you read them" stuff yes. I did and the complaints were to put it shortly wrong. Deadspin is wrong on the epistemological aspects of what makes statistics believable. Additionally it's wrong on what data you should use (because we are explicitly wanting to look at the deflation issue it is not sufficient to say they use different balls for special teams and then throw that fumble data out. The fact that they're using balls controlled by the refs and inflated to the proper levels is what we are interested in)

All of them are wrong on the distribution (we know the distribution because we know the event, if it's not conforming there is a problem) issues. In this case 1/x is the avg of a geometric distribution, which is normal (though each one will have slightly different variances due to total test numbers).

The guy who looks at outdoor games is comparing events with significantly different variances and suggesting they have the same. Which negates the order issue of claims and also in this instance ntroduces significant omitted variable bias*. The guy who throws out the pass data is doing the same.

The people doing the normalcy tests don't actually understand the tests. If the pats are a legitimate outlier(ie not drawn from the same distribution) the tests should fail. So saying the data isn't normal when the question we are asking is "are the pats on the same distribution" actually answers the question in the other direction.

*technically there is ovb in the first data set too, but it should favor the patriots so it's not too much of an issue if you find that the pats are still out of bounds.

The first shot at it isn't perfect, because the question is not "are the pats anomalous" but "is it likely we would find one team so far out". And because there are data collection issues with how you handle inverting. But the answer to the second question is still an emphatic "yes" (at about a 1.5% p-value for the proper open air teams stat over five years not examining any other five year period [note that this is a five year period which includes the pats high fumble season]) and since we don't really have a reason to believe that the null of "the pats are on the same distribution as anyone else" we should definitely reject the hypothesis and the numbers do indeed suggest that it's worth investigating what type of effect this could have had.

i'm only responding so i don't get fined

Several thousand plays between 2007 and now the refs handling the ball between every play haven't noticed a problem. In dozens of plays since 2007, the other team has been able to recover the ball without noticing a problem until the Colts game. And perhaps most noticeably, despite the huge degree of turnover among Patriots players over the past seven years, not one of them has come forward to say that the balls were deflated.

HappylilElf

y2jake215 wrote: »

HappylilElf wrote: »

Madpoet wrote: »

Next year is the franchise sellout where Tom Brady defeats the Russians.

excuse me did you just refer to the epic finale of the rocky films as a "franchise sellout"

ooh ooh does that make 2030 when all of these players are stumbling around suffering from CTE Rocky 5???

THERE WAS NO ROCKY V

ROCKY V DID NOT HAPPEN

GAWD

jungleroomx

Is Goumindong still on this shit?

C'mon man. Move along.

TexiKen

I vote for Patriots as I hold my nose, and I hope Katy Perry bombs horribly to the point where they scale back the halftime show next year.

Wishful thinking.

HappylilElf

however there was an epilogue that came out years later called Rocky Balboa

...

it was ok

Zombie Hero

Goumindong wrote: »

And whomever posted the "did you read them" stuff yes. I did and the complaints were to put it shortly wrong. Deadspin is wrong on the epistemological aspects of what makes statistics believable. Additionally it's wrong on what data you should use (because we are explicitly wanting to look at the deflation issue it is not sufficient to say they use different balls for special teams and then throw that fumble data out. The fact that they're using balls controlled by the refs and inflated to the proper levels is what we are interested in)

All of them are wrong on the distribution (we know the distribution because we know the event, if it's not conforming there is a problem) issues. In this case 1/x is the avg of a geometric distribution, which is normal (though each one will have slightly different variances due to total test numbers).

The guy who looks at outdoor games is comparing events with significantly different variances and suggesting they have the same. Which negates the order issue of claims and also in this instance ntroduces significant omitted variable bias*. The guy who throws out the pass data is doing the same.

The people doing the normalcy tests don't actually understand the tests. If the pats are a legitimate outlier(ie not drawn from the same distribution) the tests should fail. So saying the data isn't normal when the question we are asking is "are the pats on the same distribution" actually answers the question in the other direction.

*technically there is ovb in the first data set too, but it should favor the patriots so it's not too much of an issue if you find that the pats are still out of bounds.

The first shot at it isn't perfect, because the question is not "are the pats anomalous" but "is it likely we would find one team so far out". And because there are data collection issues with how you handle inverting. But the answer to the second question is still an emphatic "yes" (at about a 1.5% p-value for the proper open air teams stat over five years not examining any other five year period [note that this is a five year period which includes the pats high fumble season]) and since we don't really have a reason to believe that the null of "the pats are on the same distribution as anyone else" we should definitely reject the hypothesis and the numbers do indeed suggest that it's worth investigating what type of effect this could have had.

I'm not sure I follow with your endgame here. Is it: "The pats fumble rate being significantly lower than every other team -> they've been deflating balls the whole time"?

jungleroomx

Zombie Hero wrote: »

Goumindong wrote: »

And whomever posted the "did you read them" stuff yes. I did and the complaints were to put it shortly wrong. Deadspin is wrong on the epistemological aspects of what makes statistics believable. Additionally it's wrong on what data you should use (because we are explicitly wanting to look at the deflation issue it is not sufficient to say they use different balls for special teams and then throw that fumble data out. The fact that they're using balls controlled by the refs and inflated to the proper levels is what we are interested in)

All of them are wrong on the distribution (we know the distribution because we know the event, if it's not conforming there is a problem) issues. In this case 1/x is the avg of a geometric distribution, which is normal (though each one will have slightly different variances due to total test numbers).

The guy who looks at outdoor games is comparing events with significantly different variances and suggesting they have the same. Which negates the order issue of claims and also in this instance ntroduces significant omitted variable bias*. The guy who throws out the pass data is doing the same.

The people doing the normalcy tests don't actually understand the tests. If the pats are a legitimate outlier(ie not drawn from the same distribution) the tests should fail. So saying the data isn't normal when the question we are asking is "are the pats on the same distribution" actually answers the question in the other direction.

*technically there is ovb in the first data set too, but it should favor the patriots so it's not too much of an issue if you find that the pats are still out of bounds.

The first shot at it isn't perfect, because the question is not "are the pats anomalous" but "is it likely we would find one team so far out". And because there are data collection issues with how you handle inverting. But the answer to the second question is still an emphatic "yes" (at about a 1.5% p-value for the proper open air teams stat over five years not examining any other five year period [note that this is a five year period which includes the pats high fumble season]) and since we don't really have a reason to believe that the null of "the pats are on the same distribution as anyone else" we should definitely reject the hypothesis and the numbers do indeed suggest that it's worth investigating what type of effect this could have had.

I'm not sure I follow with your endgame here. Is it: "The pats fumble rate being significantly lower than every other team -> they've been deflating balls the whole time"?

Short answer: Yes. He's invalidating statistical analysis by many people (and 538) because he's got a grudge.

y2jake215

HappylilElf wrote: »

y2jake215 wrote: »

HappylilElf wrote: »

Madpoet wrote: »

Next year is the franchise sellout where Tom Brady defeats the Russians.

excuse me did you just refer to the epic finale of the rocky films as a "franchise sellout"

ooh ooh does that make 2030 when all of these players are stumbling around suffering from CTE Rocky 5???

THERE WAS NO ROCKY V

ROCKY V DID NOT HAPPEN

GAWD

Yeah none of these nfl players will remember it either