Mothman wrote:
People support flawed analysis and flawed conclusions all the time.
I don't see it as evidence that the flawed analysis was correct. Burke basically comes to the conclusion that the fumble rates are interesting but could just as easily be the result of good coaching as cheating.
I guess where I fall on this is I just don't know. I don't want to come off as saying you're being dismissive (of the other reports), but I personally I can't rule it out, either. And they're not necessarily supporting Sharp's analysis at face value; they're doing their own independent research and coming up with similar findings that support Sharp's conclusions. Now, I haven't researched these guys much but I do know sites like FiveThirtyEight are pretty reputable when it comes to this type of stuff.
I'm not a statistician but even I can see that Sharp has cherry-picked his way to a conclusion.
In what way? The years he chooses to analyze? He seems pretty clear why he chose those years.
He throws dome teams out to make it work.
He says in his latest article why they should be excluded and admits that even if they're included the conclusion doesn't change. He explained his reasoning here:
http://www.sharpfootballanalysis.com/bl ... -last-time
It's painfully obvious that he went sniffing around for evidence of cheating and found a way to present the data that makes it look as if he found something terribly significant.
Interesting. I was of the impression he went into it with no preconceived opinions, just as other independent researchers did that led them to their own conclusions.
Heck, just look at it simply and compare the Patriots fumble numbers in 2005 (when they fumbled 19 times and lost 9) to their numbers in 2008-2009. We can see they fumbled 17 times in each of those two seasons, losing 10 and 9 respectively.
Now who's cherry picking?
Are those numbers indicative of a big post-2006 spike?
What were the fumble rates of the other 31 teams during that same time time period? And can we dismiss Sharp's conclusion entirely if it doesn't work for
every year? And they did decrease their fumbles from 2008-2009 (by two each year), which seems to still favor the conclusion, even on a smaller scale. I mean, their fumble numbers didn't go
up during that period.
I don't know how effective it was. First of all, it's hard to place much stock in a report written by someone that admits he's a lifelong Patriots fan. Even if he tries to be unbiased, I have a hard time believing he can be and several remarks in the article show me as much. Though he makes a few interesting points.
I believe Sharp counters some of this in his latest response, though. Even though the Patriots aren't the BEST in league in fumbles each and every year, just look at the author's own revised chart.
It looks interesting that the Patriots were between 10th-19th from 2002-2006 and then in the top-5 every year save for 2013, which Sharp explains was the result of one bitterly cold Week 12 gave vs. Denver (I won't post his entire explanation but it makes sense when you read it). Fustin completely discounts this anomalous game, saying "and no, you can't just say 2013 doesn't count because of the 6 fumble Broncos game because you would then need to remove all bad games by all teams before making a comparison -- play fair!" I do agree that Sharp should also discount all bitterly cold games that resulted in high fumble rate anomalies. However I'm skeptical that even after doing this it would change the result.
I'm probably way off base but I'm starting to see this like the climate change argument: some politicians discount it completely because not every summer is the hottest on record and sometimes there are bad winters (how can there be climate change when we keep getting these December blizzards?!) when the cumulative data points to a shift in overall planet temperatures.
I also am disappointed that Fustin makes reference to the Vikings-Panthers game re: warming footballs when that has since been debunked.
Funsin also ends with " This team is not a team of cheaters, as far as I can tell. Numbers don't lie." which I think is pretty ironic seeing as though he went to all this effort to show that Sharp's numbers were lying.
And I found this part quite a bit of a stretch:
Anyhow, might I also suggest the following plausible explanations for drastic improvement?
Consider 2006: Reche Caldwell, Doug Gabriel. Now consider 2007: Randy Moss, Wes Welker. Same in 2008. You eventually get Rob Gronkowski, Julian Edelman, et cetera. Please don't make me compare Reche, Ben Watson, and Troy Brown (LOVE YOU, TROY!) to Gronk and Welker.
This is just lazy and dumb. Just because a player is better doesn't mean their grip wasn't improved by a slightly underinflated football.
The pre-2006 Patriots ran a completely different offensive set than the post-2007 Patriots. Remember the 2007 Patriots? The greatest offense ever assembled? They were the first Patriots spread offense. They still use that today.
I fail to see how scheme has anything to do with fumble rates. Were the players touching the balls at different rates?
Tom Brady is simply better now. I remember watching Brady dink high passes off of receivers on crossing routes leading to tipped interceptions. I remember him being in the pocket too long. I now see him getting rid of the ball at nearly the fastest pace in the game. I see him intentionally throwing the ball at the feet of receivers in traffic and on crossing routes causing them to catch the pass on the way to the turf (thus not allowing fumbles on tackles).
Well, we're not talking about interceptions here so...yeah. I get the release time thing as it relates to sacks/fumbles. This is one area I have the biggest question. I think all fumbles coming from the QB getting hit with the ball only in his throwing hand should be omitted as I assume even if the ball was under-inflated by 50% that these hits would've resulted in a fumble. The thought that Brady now throws at the feet of receivers in traffic on crossing routes as to minimize fumbles on tackles is pretty loose in my eyes. I'm not sure that's A) true and B) necessarily results in fewer fumbles. Someone would need to do a whole study on that before I put any stock in that. If anything, it could lead to more fumbles since it takes longer to "secure" the football bringing it up from the feet to the chest and cradling it in the bicep instead of taking it from chest-level to the bicep. If anything, you'd think throwing it low at the feet would cause more incompletions.
And why were they much more pedestrian in 2008 while still having the greatest offensive skill players in the league? Could it have had something to do with Bernard Pollard and Matt Cassel?
Again, I don't know if there's any correlation between a guy's overall talent level and lesser fumble rates. Heck, Peterson has been one of the best football players on the planet since he stepped on the field (and has the hardest grip of anyone outside of Robocop) and he still had quite a few fumbles (for a RB) from 2007-2009. Fustin can keep harping on this but it's a weak theory with absolutely no basis.
The bottom line for me is that this will never be answered. I'd want to dismiss all one-handed QB fumbles and also all fumbles where the ball-carrier is being careless with the ball (i.e. holding it wide, thinking he has a TD and not realizing someone was behind him who punched it out when he was relaxing, etc.). I think these fumbles would've happened no matter the inflation. The same is probably true for fumbles where a guy comes in full speed and puts his helmet on the ball. Frankly there are probably too many other variables. Ultimately I think it probably provided the Patriots with a marginal competitive advantage when it comes to fumble rates but not enough to really change their legacy from 2007-today as one of the best teams in the NFL.