Dead$pin claims that they are going to put an article together to refute Sharp's work that they got together with a bunch of statisticians.
As a statistician myself, I don't see any glaring flaws in Sharp's work, or at least any I could think of. My cousin has a PhD in Mathematics, I'll probably call him and see what he thinks (he doesn't know jack about football though).
YR
To keep every team on a more level playing field, I eliminated dome teams from the analysis, grabbed only regular season games, and defined plays as pass attempts+rushes+times sacked
I think this is a pretty big one, particularly because it's just an assumption. I guess maybe if he had stated to what degree and whether or not it was significant, but just saying you tossed indoor teams is kind of questionable.
I wanted to know the difference so I looked it up, using SportingCharts.com. Lots of copy and pasting in google docs (never used before but computer just got back up) going on so I'm willing to accept I may have made a mistake or two along the way. Seasons where teams played indoors saw teams fumble on average 1 time less per year. Over this guy's time period, indoor teams fumbled 22.59 times per season while outdoor teams fumbled 23.62 times per season. That's 1.02 fumbles worth of difference, and I'm not entirely sure but doesn't look to be significant. Excluded teams Indy and Atlanta have the two best totals at 259 and 264 fumbles, respectively. Patriots have 269, followed by Chargers at 278 and KC at 281. Indy alone drops the average for indoor teams by 0.6 fumbles because they had such a good run with Peyton, and because there are way fewer teams that play indoors so the impact of the Colts numbers is more pronounced. Since he's left they've fumbled 90 times in only 4 of the 14 years, which is a 5.5 fumble increase over his period from 2001 to 2010.
Atlanta is just the opposite. They started the time period really quite poor, but around 2006 things got much better. They actually have the same number of fumbles post-2006 that the Patriots do, 121. What changed for Atlanta? Well, getting rid of Mike Vick's 12 or so fumbles per season helped, and finding Matt Ryan was also very good for them. Ryan has 33 fumbles in his 7 seasons since 2008, and in the 7 seasons since 2007 that Brady has been able to play, Brady has 37 fumbles.
Point being, whatever benefit indoor teams have doesn't seem to be nearly as important as what kind of QB they have so excluding indoor teams is very questionable.
Which in layman’s terms means that this result only being a coincidence, is like winning a raffle where you have a 0.0001711874 probability to win. In other words, it’s very unlikely that results this abnormal are only due to the endogenous nature of the game
The entire basis for his claim of how unlikely their low fumble rate is revolves around the assumption that each team has an equal likelihood of fumbling on any given play.
This is not true. Over the past 14 seasons (time span he looked at), there have been no more than 2 players in any given season that played a position other than QB and finished in the top 20 in fumbles. There are 21 non-QB players in the last 14 years out of the 280 total players. QBs fumble at a higher rate than other positions, and some QBs fumble a hell of a lot more than others. In fact, if your QB's name is Jay Cutler (75 fumbles in 119 games) your QB is twice as likely as Peyton Manning (70 fumbles in 256 games) to fumble in a game.
Which brings me to another issue I have. In addition to comparing varying probabilities of fumbling, he's simply not comparing like situations. Not every team has a Tom Brady on their roster, and most of the teams with comparable QBs were left (Indy, New Orleans, maybe Atlanta) out because they play indoors. Furthermore there are only 3 QBs who played in both the 2001 season and the 2014 season in Brady, Manning, and Brees. No other team than the Patriots can say that the guy who started the season in 2001 also started the season in 2014. The rest of the NFL shuffled through QB after QB and the Patriots had Brady the whole time. His period of observation is far too broad because 14 seasons is an exceptionally long QB career and many of the numbers for the rest of the league were built on the backs of awful QBs.
Is there any question that the Saints would be better in fumbles per play if they didn't have Aaron Brooks' awful fumbling habits included in their numbers? The Saints have 23 more fumbles than New England since 2007, but Brees has 54 fumbles whereas Brady (plus Cassel's 1 year) has 44.
As for outdoor teams, Baltimore has 139 fumbles since Flacco joined the team. He personally has 62 fumbles. In that same time span from 2008, the Patriots have 107 fumbles with only 38 of them attributable to QBs. Excluding the QB positions, the teams differ by 8 fumbles over the 7 year span. The difference is less for the Chargers when looking at 2007 and beyond. Rivers has 67 of the Chargers 148 fumbles. Patriot QBs have 44 of 121, putting the score at 77 to 81 if you exclude QBs.
QB's usually contribute a hefty amount to fumbles. New England has the most stable QB position since 2001, and their QB is one of the best in the game and he doesn't fumble all that often. Part of the reason they had as many fumbles between 2001 and 2006 was because Brady was fumbling more. He had 59 through 2006, and has 37 since. Rather than deflating footballs, maybe the rule change itself helped him. Would be a shocker considering all teams have improved in fumbling since the rule was implemented. Pretty easy to see, the distribution in the guys charts shifts to the right as a result of more touches per fumble. Or maybe he made an effort to improve as all younger players do. Romo fumbled a hell of a lot more before 2009 than he has since. 32 in 2-2/3 seasons versus 32 in 5-1/3. Coincidentally that was the year that Dallas moved into a retractable roof stadium. I guess being "indoors" really does make a difference, or maybe Romo is just playing better.
I have no idea either way if they deflated the footballs. Even though I think the math and science behind the pressure loss is interesting, I'm inclined to believe they did let air out or possibly inflated them under conditions that would yield a lesser PSI. I don't think fumble numbers really say much. Most QB's don't have 13 or 14 year runs of great play and some suffer through horrid QB play for years on end. Nobody should expect a team led by Mike Vick to rule the world in ball security.
One more thing. People have cited both both control and receiving ability as the perks of deflation. The Patriots are near the top of the league in terms of % of passes dropped since 2007. 6th highest percentage, and 3rd most drops. For whatever reason the Colts and Lions have more overall and a higher percentage, which someone may not expect if they think weather influences one's ability to hold onto the ball.
