Started By
Message

re: Donald Trump: It's ridiculous the way the NFL has mistreated my friend Tom Brady

Posted on 8/24/15 at 9:18 am to
Posted by MrFreakinMiyagi
Reseda
Member since Feb 2007
18958 posts
Posted on 8/24/15 at 9:18 am to
They haven't been debunked to the tune that a reasonable person could write off the success that the Pats have had in holding onto the ball.

Posted by wildtigercat93
Member since Jul 2011
112311 posts
Posted on 8/24/15 at 9:43 am to
They have been debunked in the way to they are completely twisted
statistics.

LINK
This post was edited on 8/24/15 at 9:47 am
Posted by MrFreakinMiyagi
Reseda
Member since Feb 2007
18958 posts
Posted on 8/24/15 at 10:19 am to
The article you linked links Brian Burke's article, which states:
quote:

They may vary slightly because I'm including playoff games, but overall the numbers look correct.


It also fails to do anything to explain why the Pats fumble numbers improved so much since 2007.
Posted by wildtigercat93
Member since Jul 2011
112311 posts
Posted on 8/24/15 at 10:35 am to
They are exaggerated to paint a picture.


quote:

It's not a great strategy to penalize Brady and the Patriots for a lack of fumbles when there was a lower chance of fumbling to begin with, based on the team's play-calling and personnel that yield quick throws and incompletions, as well as fewer sacks. So what's a better metric? Well, one idea would be to compare running backs on fumbles per rush, or receivers on fumbles per reception. To look at quarterbacks is a much more difficult task, given the different speeds at which they get rid of the ball and how often they take sacks, but you might look at something like fumbles per sack or fumbles per QB hit. But the point remains: when quarterbacks throw quickly and avoid sacks, it's seemingly much less likely for the play to end in a fumble. More on this later.


quote:

Could it be that the Patriots have one of the best quarterbacks of all time? Or perhaps it's because they have the best coaches of all time? Or perhaps it's because Gisele taught Tom Brady and his receivers how to fake hanging onto the ball using secret supermodel makeup tips—she and Brady started dating in 2007, after all. Or perhaps it's Robert Kraft's blue shirt/white collar combo. It could be any combination of the above, this line of inquisition would argue. If we're really worried about quarterbacks playing the balls the way they like, why aren't we talking about Aaron Rodgers, who led the league with the fewest interceptions and has admitted that he likes his ball over-inflated?



quote:

And back to the caps. And the fact remains, it's difficult to learn anything from this analysis. There are a few other problems with the SharpFootball investigation that we didn't even bring up—issues of multiple testing and the post-hoc fallacy, for example—but we figured it was more important to provide a different look at New England's fumble rates than to continue rambling about best practices.



There's a reason no one is using this analysis in any of the arguments of deflategate. It's bogus statistical reporting


More

quote:

New England running backs have the fewest fumbles per 100 rushes (about 0.6) of any NFL team going back to 2007, but the difference between New England and the rest of the league isn't as extreme as previous authors suggested—its about 1.5 to 2 fumbles per season. And this puts a point on the whole indictment of these posts: There are some broad and relatively minor observations to be made, if you're conspiratorially-minded enough to make the effort, they just don't have the statistical muscle that you'd expect given how widely they've traveled.




quote:

New England finished with the third-lowest fumble rate on receptions, looking at all regular season plays between 2007 and 2014. Interestingly enough, the Ravens, who have been cited as one of the possible whistle blowers for Ballghazi, post the lowest per-reception fumble rate. Given the number of teams in the league, it is not far-fetched to see one with fumble rates this low. Neither of these results appear all that extreme or insane, and, most certainly, they are not "extremely abnormal."
This post was edited on 8/24/15 at 10:37 am
Posted by wildtigercat93
Member since Jul 2011
112311 posts
Posted on 8/24/15 at 10:38 am to
quote:

Look in the bottom row. Brandon Tate is listed as having 11 fumbles in 35 touches during his time playing for "Non NE". That's absurd. How is that possible? Is Tate a fumbling machine? Of course not. According to NFL.com, all 11 of Tate's fumbles during his four years with the Bengals came on kick or punt returns, and because special teams fumbles use different balls, there is no reason for them to be included in this analysis. In other words, not only was the author using inappropriate methods in all three of his posts, but there is a reason to believe he also wasn't using the correct data, either.



quote:

Gregory J. Matthews is an assistant professor of statistics at Loyola University Chicago. He completed his Ph. D. at the University of Connecticut in 2011. Matthews is a co-author of openWAR and actively blogs at statsinthewild.com. You can follow him on Twitter @statsinthewild. Michael Lopez (@statsbylopez) is an assistant professor of statistics at Skidmore College.
This post was edited on 8/24/15 at 10:39 am
Posted by Dr RC
The Money Pit
Member since Aug 2011
58058 posts
Posted on 8/24/15 at 11:46 am to
quote:

all 11 of Tate's fumbles during his four years with the Bengals came on kick or punt returns, and because special teams fumbles use different balls, there is no reason for them to be included in this analysis.


all I see is a clear admittance that the type of ball used does influence the odds of a fumble happening.
This post was edited on 8/24/15 at 11:47 am
Posted by MrFreakinMiyagi
Reseda
Member since Feb 2007
18958 posts
Posted on 8/24/15 at 11:56 am to
Except they aren't all exaggerated to paint a picture.

Again, from the Burke analysis linked in the deadspin article:
quote:

In this case NE is at the top of the list, and the next best team is a distant second. Notice how the second team (BLT) through the second to last team (PHI) have rates that are within 1 or 2 plays of each other. NE, however, is better than the next best team by 20 plays per fumble.


Posted by wildtigercat93
Member since Jul 2011
112311 posts
Posted on 8/24/15 at 1:13 pm to
quote:

It's arrived at by using plays per fumble, rather than fumbles per play. If you use plays per fumble and calculate the same Z-score for the Patriots you get 3.84. We can then calculate that a team will do better than this (i.e. have a higher Z-score; higher is better for plays per fumble) about 1 in 16256. So we're guessing this is where that number comes from. The problem is that this calculation also relies on the assumption of normality, which holds for fumbles per play, but NOT for plays per fumble (see the Q-Q plot)!


quote:

The "data scientist" and the author validly assumed that fumbles per play followed a normal and then went ahead and calculated the Z-score based on plays per fumble, a variable that is strongly skewed to the right. Formally, X being normal does not imply that 1/X is also normal. If this is how that happened, it would be a sloppy misstatement of statistics stemming from an arbitrary preference in diction. Either that, or the author just went for the most improbable result that he could find.

Posted by wildtigercat93
Member since Jul 2011
112311 posts
Posted on 8/24/15 at 1:17 pm to
quote:

all I see is a clear admittance that the type of ball used does influence the odds of a fumble happening.


Ask Tony romo how he feels about K balls.

By your statement, do you think any type of ball management is against the rules? Might want to tell NFL films that who just filmed the Houston texans rubbing down the K Balls into regular game balls on hard knocks last week.
Posted by MrFreakinMiyagi
Reseda
Member since Feb 2007
18958 posts
Posted on 8/24/15 at 1:40 pm to
Again, from the Burke article:

quote:

As I mentioned, sometimes outliers can be exaggerated depending on which way you look at a rate. Should we look at fumbles per play or plays per fumble? Mathematically, both measures contain the same information. I prefer fumbles per play, because I conceive of each play having an independent probability of fumble, and fumbles per play is an unbiased estimator of that probability. So this is what it looks like from the other direction. Not much difference.
Posted by wildtigercat93
Member since Jul 2011
112311 posts
Posted on 8/24/15 at 1:47 pm to
good god I'm done with this. You keep citing an data set with clear bias and falsities that you have even admitted, and just keep thumping the same book with it despite the flaws already pointed out. There's a reason no one has even mentioned that bogus interpretation of the data in the past 4 months. Even lsupride87 and goldrush25 (the biggest NFL defenders on here) gave up on that crap.
Posted by Dr RC
The Money Pit
Member since Aug 2011
58058 posts
Posted on 8/24/15 at 2:38 pm to
quote:

By your statement, do you think any type of ball management is against the rules?


If ball management in question is against the rules in the rule book yes.

If it isn't no.

Your boy cheated and he doesn't get excused from getting caught just b/c other guys might have done it.
This post was edited on 8/24/15 at 2:39 pm
Posted by LeonPhelps
Member since May 2008
8185 posts
Posted on 8/24/15 at 2:47 pm to
Such a shame. I liked Trump before this. Any friend of my enemy is my enemy. frick Tom Brady's cheating arse.
Posted by MrFreakinMiyagi
Reseda
Member since Feb 2007
18958 posts
Posted on 8/24/15 at 4:07 pm to
quote:

You keep citing an data set with clear bias and falsities that you have even admitted, and just keep thumping the same book with it despite the flaws already
pointed out.

I cited analysis LINKED IN THE ARTICLE YOU POSTED.

first pageprev pagePage 3 of 3Next pagelast page
refresh

Back to top
logoFollow TigerDroppings for LSU Football News
Follow us on Twitter, Facebook and Instagram to get the latest updates on LSU Football and Recruiting.

FacebookTwitterInstagram