OT: More Deflategate information...its getting silly

Submitted by GoWings2008 on

I don't know if anyone else was listening to Dan Lebatard's show yesterday, but he had a guest on named Warren Sharp who brought up some interesting facts.  First he reminded us about the Brady and Peyton Manning led petition of the NFL to allow visiting teams to supply their own footballs for when their offense had the ball.  They were successful and we're left with the rule we have today.

Background on the rule change:  http://www.usatoday.com/story/sports/nfl/2015/01/26/tom-brady-deflategate-peyton-manning-rule-change-nfl/22372835/

Then, Mr Sharp did his due dilligence on the Pat's fumble rate since that rule change, which involved some solid statistics work (when I read the words 'distribution' and 'mean' ... I knew I was in the presence of greatness). 

Essentially, the result is that the Pats not only have the lowest fumble rate since 2006, the second least fumbling team isn't even close. The chances of the Pats having that low of a fumble rate are extremely slim.  He compares it as such:  “... in layman’s terms means that this result only being a coincidence, is like winning a raffle where you have a 0.0000616 probability to win.”

Link on a Boston.com story:  http://www.boston.com/sports/football/patriots/2015/01/25/patriots-fumble-nearly-impossible-rate/LCgrlUR9qgxDsIgcal9dUI/story.html

Finally, an interesting tidbit that ball boys need to go potty too.  And to conclude by quoting the article of this piece (http://www.msn.com/en-us/sports/nfl/patriots-ball-attendant-just-using-the-bathroom/ar-AA8CEwY)...

"The whole argument is grasping as straws at this point. Now it's just funny."

amaizenblue402

January 27th, 2015 at 9:10 AM ^

Wake me up when it's all over and we aren't taking about the words deflate or gate any longer. Hey guys, did you know they are playing the Super Bowl in 5 days?!

mgobleu

January 27th, 2015 at 9:15 AM ^

and that's for Seattle to win. I'm half heartedly pulling for the pats, although I mostly don't care. However, I know if they win, this shit NEVER goes away and will probably be brought up in every patriots game and every super bowl from now into eternity.

evenyoubrutus

January 27th, 2015 at 9:24 AM ^

Huh? If Seattle wins then deflategate is all people will talk about, as it will be seen as "proof" or "validation" that the Pats have been cheating for years and it's the only reason they have been so successful. The opposite is actually true- if the Pats win it will silence everyone.

ijohnb

January 27th, 2015 at 11:33 AM ^

is the precise issue.  Nobody really cares that the Super Bowl is being played.  Ratings have decreased the last two years, and they have propped this season up with WWF storylines released as "damning self-incriminating" information that is tarnishing the brand.  Tarnishing the brand the same way it did when Miss Elizabeth took off her skirt during Wrestlemania 5 or when Dibiasi bought the title from Andre the Giant.

If the NFL did not want this to be an issue it would not be an issue.  It is an issue because the more controversy there is, the more people watch.

DMill2782

January 27th, 2015 at 9:31 AM ^

the fact that the Patriots are doing it. This is one of the reasons why they are one of the best teams in the NFL year in and year out. 

I bet 30 years ago it was "impossible" for anyone to ever throw for 5000 yards in the NFL. 

Teams and players set new standards all of the time. It's called being great. 

HANCOCK

January 27th, 2015 at 3:06 PM ^

Well, to be fair, 30 years ago the rules were much different than the current NFL. WR's could be hit and re-routed. It was a much more physical game. QBs didnt get to doctor the ball to their liking. DBs were allowed to play physical coverage. Throwing for 4000 yards is pretty much expected at this point, but go back just 10-15 years and QBs were rarely hitting 4000, let alone 5000. Its a different game. Fumbles can be controlled to an extent, but for the most part they are pretty random. You can absolutely work on ball security and consistently post strong numbers, but to just be on a different level than the rest of the league for that long is a little strange

Heteroskedastic

January 27th, 2015 at 10:23 AM ^

If you read the original article, the rate that is statistically improbable is the fumble lost rate .  When he performs the same analysis on the overall fumble rate, the Patriot's run is no longer an outlier. He then asserts, that they are the least fumbling non-dome team since 2007 by a large margin, but that does not have the same punch as his original claim.

Furthermore, there are other factors that may influence the total fumble rate.  As one critic wrote in the comments, BenJarvis Green Ellis may be one of those factors.  He has an abnormally low fumble rate.  A more detailed analysis would include a comparison of Ellis' fumble rate to the fumble rate of all other Patriot's rushers.

Finally, I am wary of an analysis that makes its claim on a rolling five year average.  This is an old trick.  Stock Market guru's have been doing it for years.  If I find the right interval length, I can claim anything I want about an investment, good or bad and be correct.  In this case, what is the significance of the rolling 5 year period?  Why not look at the data by season.  I would assert that would be the most logical choice since most NFL team's undergo significant changes to player personnel annually after the draft and free agency.

ats

January 27th, 2015 at 9:37 PM ^

only if you have made the unsubstantiated and unsupported a priori determination that fubles follow a random statistical distribution. 

Unfortunately for that line of argument, no one has actually proven that, substantiated it, or has supporting data for it.  Yes, people will sometimes model it as a random statistical distribution but that is as a shorthand because we don't have a better way of modelling it.  Its very similar to using random statistical distribution to model yield in semiconductor manufacturing.  For a given process its generally reasonable, but its not actually correct.  We know that there are numerous things which will change yield and we know some fabs/companies are better than others. 

Likewise it is important not to confuse modelling and especially future predictive modelling with reality.  In the case on fumble data, there are numerous factors that contribute to an actual fumble, a great many of which (really the vast majority) that have nothing to do with randomness.  So the basic problem of the statement "Nearly statistically impossible" is that it is fundamentally incorrect.  The statement is assuming that a weakly correlating model is the underlying model for the process. 

Wolverine 73

January 27th, 2015 at 9:58 AM ^

It's been silly for days already.  Didn't the media see the final score?  Didn't they see how NE outscored the Colts in the second half after the footballs were "fixed"?  Didn't they see the Pats run the ball down the throats of the Colts' defense?  How anyone can think that whatever happened to the balls affected the outcome of the game is beyond me.  If someone fooled with the balls, fine, do something about it.  But that game was being won by NE if they had played with a volleyball.  Thank God for the snowstorm in the east so the media could carry on obsessively about that for the last couple of days.

LJ

January 27th, 2015 at 9:23 AM ^

I mean, no one else finds the fumble thing concerning?  Especially on this site, which champions the notion that fumbles lost is one of the most random stats in football, with very little year-to-year correlation?

I agree the story has been beat to death, but this is serious stuff we're talking about.

LJ

January 27th, 2015 at 9:26 AM ^

I agree it's over the top, but I think it's reasonable to want to know why a team's equipment was measured to be in violation of a clear rule, when violating that clear rule could potentially give you an advantage.  What's so controversial about that?

coldnjl

January 27th, 2015 at 9:38 AM ^

They should be punished IF it is proven that there was intention to break the rule...i.e. an act of intentionally deflating them. Without that, this is stupid. With that, fine them and move on. The NFL can't afford to come down hard on a team for slightly underinflating balls, especially after the domestic violence fiasco.

bsand2053

January 27th, 2015 at 12:51 PM ^

The rule is easy to enforce.  Let the teams supply their own balls to warm up with but only use league balls during play.  Not that hard and I can't understand why the NFL lets teams use their own balls.  Imagine what baseball teams would get up to if they could use their own baseballs.  

maizenbluenc

January 27th, 2015 at 9:51 AM ^

(So do Michigan State's) They are clearly violating of the rules (at the very least the intent), pushing the boundary, and daring the refs to call it. In their mind (and many TV talking heads' minds), they are just "playing the game".

How much does pass interference alter the score of the game over air pressure being 16% below tolerance? Is anyone talking about how Seattle's secondary cheats?

Anybody check any other teams' footballs at halftime? How many have been altered in some way? (I am guessing someone ought to check Peyton Manning's game footballs.) Does the entire nation think just New England pushed the boundary on these minor rules?

New England had one turnover in the game: an interception in the 1st half when the balls were deflated, no fumbles in either the deflated half or the inflated half, and twice as many (4 vs 2) touchdowns in the inflated half.

This issue was silly the moment it came out.

LJ

January 27th, 2015 at 10:09 AM ^

Every time a referee thinks Seattle has committed pass interference, they impart the predetermined, appropriate punishment.  Sometimes, there is disagreement of whether the rule has been violated.  Here, there is no disagreement that a rule was violated.  Thus, some punishment is appropriate.

I have no idea if the rule is uniformly enforced.  But it appears that it could lead to at least some advantage (I agree that how much of an advantage is debatable) and thus it warrants some punishment.  Whether it aided them in that particular game is totally irrelevant, just the way whether defensive holding aided you on a particular play is totally irrelevant to it being flagged, and punished.

it's Science

January 27th, 2015 at 9:39 AM ^

Considering how corrupt the league is in general, it's really not that serious. Murderers, wife beaters, PED addled monsters roam the field. Grown men bashing their heads in and shortening their life span. A commissioner who will do anything to protect the NFL brand. I think I can live with few balls deflated by 1PSI and go about my day just fine. It's still just a game, and we have to remember that. Reporters need a story, and those with "scientific facts" proving something positive for either side will get the spotlight. Why do stress ourselves over this nonsense? I need a meditation break so I can get back to reality.




Sent from MGoBlog HD for iPhone & iPad

GoWings2008

January 27th, 2015 at 9:46 AM ^

yes, I can deal with this story.  But those are extreme cases and, imo, more of a reflection on society than on football.  I think that when it comes to the competitive nature of the game, regardless of the NFL's leadership, having some statistical support to a story of a professional team possibly getting an unfair advantage...well, to me that seems like something to talk about.  Especailly if that team has a history of that sort of behavior. 

GoWings2008

January 27th, 2015 at 10:06 AM ^

that the statistics are being cherry picked. But thats sort of why they exist...to support.  Statistics aren't facts.  If certain facts are being cherry picked, then I think people should have a problem with all of this.  Which, I'm sure, there is a certain amount of that happening on both sides of the argument.

Heteroskedastic

January 27th, 2015 at 2:05 PM ^

I beg to differ.  Statistics are facts.  They are a numerical summary of historical occurences.  How a statistic is presented or framed is open to interpretation and subject to bias.  IIRC, Brian et al has asserted numerous times that turnover margin is random, especially, fumble recoveries/fumbles lost because they depend largely on which direction an oblong ball bounces.  However, total fumbles is less subject to randomness.  The statistic that Sharp was refering to was total fumbles lost per running play over a five year period.  The total fumbles per running play number is much less damning.  If the fact you are referring to is the underinflating of balls, that fact has been cherry picked as well.  This is a fact that has also been framed with the intent of putting one team on the defensive.  We have not been privy to the ball inflation records of all teams for all games since 2006.  If that data comes out and it shows that Belichek's game balls trend below the NFL imposed limits and significantly below the mean (-3 standard deviations, assuming normal distribution, ceteris paribus), that might support the existence of an underinflation trend with the intent to provide a competitive advantage.  A sample size of one should not be enough data to accomplish this objective.

bronxblue

January 27th, 2015 at 9:30 PM ^

Statistics, as mathematical represenations of the data provided, are absolutely facts.  You can take issue with the inputs, but if I run a regression on a data set it is going to give me a result that is based on cold, unbiased math, not subjective reasoning applied to 1's and 0's.  But the issue here is as much the inputs and data sets used, and if you look around on the internet a number of people  have poked many holes in how these studies were performed.  Yes, there is always data manipulation to an extent in any analysis, but the number of questionable decisions made to tease out the findings in these "IMPOSSIBLE TO BELIEVE" reports shouldn't be ignored as merely par for the course.