If you read the link refuting the study, it's about 99% attacks on the author and 1% against his actual statistics. 538 then adjusted for the few statistical problems and found that their results were actually pretty close.Huh? 538 was one of the ones that originally posted that Sharps data was mostly bullshit complete with links of other studies blasting the fuck out of it.
http://fivethirtyeight.com/datalab/y...ical-analyses/
I don't know where Benm (Benjamin Morris) got his data (and what it was) to plug into his calculations for that chart and I couldn't find where he said he got it, other thanherewhere he references Sharps original data even though 538 acknowledged in the past (topmost link) that Sharps not only cherry picked his data, but used incorrect data. 538 didn't tear the fuck out of Sharps themselves; they didn't have to. The linked studies they have at the bottom of their article that they tell you to read did that perfectly for them and they alluded to that in their own article.
In the first study that 538 links (http://regressing.deadspin.com/why-t...805710/+kylenw) done by statistics professor G. J. Matthews, he breaks down team fumble rates by position and concludes that
That's where I got my data analysis. Did Morris pull his data from Sharps? I've found literally nothing on him other than him alluding to Sharps original data that he may have used for his own calculations.
I actually believe quite the opposite. They don't need anything huge just "maybe" they did it and they can do what ever they want. It will be interesting to see what happens either way.I can't see much happening w/o hard proof here, the NFLPA isn't just gonna sit by and let Brady get railroaded over more likely than not he cheated. The NFL is gonna need something more than circumstantial.
The guy that you said is biased is from Baltimore and teaches statistics at a University there.If you read the link refuting the study, it's about 99% attacks on the author and 1% against his actual statistics. 538 then adjusted for the few statistical problems and found that their results were actually pretty close.
I really don't see how any non homer could read that link you posted and not think it's completely biased.
why would it be home only?Also keep in mind that any minute advantage to the Patriots from underinflated balls was home only.
Pats have a .661 road winning percentage under Belichick.
All footballs are handled by home team supplied staff after ref inspection. No opportunity at away games.why would it be home only?
Doesn't this go to the crux of the matter though? The balls being altered AFTER ref inspection? Away teams can provide their own balls, this was established in 2006 after Brady and Manning lobbied for it I believe.All footballs are handled by home team supplied staff after ref inspection. No opportunity at away games.
First, he's the article was written by 2 people who both went to college in New England and neither of them work in Baltimore. Secondly, if you even read the link you'd see that their actual data analysis was awful. Not only did they ignore the main point that NEs fumbles dropped dramatically starting in '07, but the best they could do was manipulate the data so that NE still went from middle of the pack to least amount of fumbles. Just by "a lot" instead of an insane amount.The guy that you said is biased is from Baltimore and teaches statistics at a University there.
Michael Lopez, the coauthor, is an assistant professor of statistics at a college at New York.
BenM, the guy who plugged in Sharps data into a graph, was hired by 538 as a sports writer. Before that he was a founder of a blog where he argued Dennis Rodman may be more valuable than Michael Jordan.
Can you link me the 538 article you're referring to that finds their results similar/close to Sharps?
I'll show you my super secret article the second you show me where I claimed there was a super secret articleDo you have the link to the 538 article you claim to have or not?