I am a huge fan of facts and statistics. I am also a student of human behavior and how the general populace, can be significantly driven towards a specific conclusion based on creatively written headlines and articles. The information landscape has definitively changed as we now live in a world where social media drives what is deemed as important and those that get the most “traction” are the ones who make the boldest statements.
That said, I am very disturbed by the overblown situation with the current called #DeflateGate, which has taken over all forms of media from last week and continues to be a subject discussed this week leading up to the Super Bowl.
The NFL, from what I have researched, has a specified fine of $20,000 for anyone caught under/over inflating the football; all this talk of extreme punishments from draft picks forfeiture to suspensions, to the ridiculous “…could the Patriots be removed from the Super Bowl” (I believe it was the Today show that lead with that statement one day last week, but, cannot confirm) has caused me to go back and try to make some sense of a situation that has gotten totally out of control.
Truly I have not taken either “side” of this discussion; however, once again we have an overabundance of people from all walks of life chiming in for one side or the other and coming up with all kinds of facts, guesses, and just flat out lies.
I have had some conversation about one particular article that has apparently compiled some “clear” evidence (speculation at best) that something has been going on in the Patriots organization for the past 5+ years and that #DeflateGate is simply the tip of the proverbial iceberg!
Let’s begin with this portion here (taken from the article) as an example:
“I spoke with a data scientist who I know from work on the NFLproject.com website, and sent him the data. He said:
Based on the assumption that fumbles per play follow a normal distribution, you’d expect to see, according to random fluctuation, the results that the Patriots have gotten over this period, once in 16,233.77 instances”.
Which 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. Which in other words, it’s very unlikely that it’s a coincidence. “ 1
Although the NFL data scientist spoken of in the original article ultimately stated “…Based on the assumption that fumbles …” (Walsh, 2014), the fact remains that his first line, assumes that no other data was considered in determining that probability. In other words, their statement is simply based on the statistical data provided and not the multitude of additional data that truly needs to be evaluated to draw any valid conclusions relevant to the ball inflation rate. He also states “…it’s very unlikely that it’s a coincidence” (Walsh, 2014); true, it is unlikely. I contend that it is absolutely not a coincidence, but a product of the personnel and organizational structure.
Anyone can creatively select statistical evidence which can be utilized to create the illusion of “proving” a concept or idea; this is, incidentally, what author of the referenced article has done.
So let’s get to the crux of the issue here.
Here are my concerns with the stated conclusions:
- A low number of fumbles(actually protecting the football) is a basic product of fundamental football – goes to coaching and training
- Why not look at completions/pass attempts, fumbles after the catch, special teams fumbles
- Look at FG % (as that has is also relevant to ball inflation amount).
- Coaching – tenure of head coach and coaching staff and their tendencies (beliefs)
- Propensity for pass first offense or run first
- Availability of a top tier quarterback and/or running back on the team
Let’s also look at the following items for other statistical similarities that although relevant, are not directly related to ball inflation:
- Total turnovers
- % of pass plays to running plays
- Total fumbles, including special teams
Other possible areas to consider and explain:
- Are “cold weather teams” more likely to spend more time on ball security?
- If this is directed towards a particular quarterback (T. Brady), then why were the Patriots still in the top 10 of the chart below when Brady was injured in game 1 of 2008 and missed the entire season?
- All teams, including dome teams, must be included for statistical relevance
The chart that follows refers to most plays/fumble in the last 25 years, earliest period begins in 2001? Okay, I am assuming, since I don’t have the data, that in the last 25 years no team has had a better number of fumbles to number of plays statistics.
Here is the referenced chart from the original article:
Now, let’s consider the success of those teams with low number fumbles per number plays just since 2010.
The top 10 teams (from 2010-2014) are as follows (chart from original article):
The top 10 teams in terms of number of fumbles per number of plays seems to have some obvious (or should I use “clearly shows” for effect) characteristics i.e., coach, quarterback, and running back.
In other words, those teams that have 1) a good, h igh quality coach and/or 2) a top tier quarterback and/or 3) a top tier running back; happen to be the best teams at, not only holding on to the football, but, arguably, record, playoff appearances, and overall general success.
So, based on some basic research along with my own general knowledge of each of these teams, I would say that they have some very obvious common denominators including:
- A Good Quarterback, in most cases (8/10), a top tier quality quarterback that is, doubtless, the strength of the team.
- The teams with more of a 50/50 mix of pass/run have a top tier running back (3/4); Arian Foster, Ray Rice, Adrian Peterson
- Generally have (7/10 including Indianapolis) a long term, top tier head coach
General Offensive Tendencies
Are they a run first or pass first offense, do they have a stud running back and/or receiver?
Atlanta: generally pass first offense: Matt Ryan , 2008
N.E.: Generally pass first offense: Tom Brady, 2000
Saints: Generally pass first offense: Drew Brees, 2010
Texans: probably closer to 50/50, Arien Foster… Who Knows
Baltimore: probably closer to 50/50, Ray Rice, Joe Flacco, 2008
Chargers: Generally pass first offense, no real great, consistent backs, Philip Rivers, 2006
Vikings: probably closer to 50/50, Adrian Peterson, T. Bridgewater, 2014
Bengals: probably closer to 50/50, a few different, decent backs, it seems, every year, A. Dalton, 2011
Colts: generally pass first offense, A. Luck, 2012 – P. Manning previous
Packers: generally pass first offense, some decent backs in some years, A. Rodgers, 2008
The article mentions the Patriots being far and above any other team, but, according to the chart the Falcons are better. The dome teams are completely discounted. Does that mean that dome teams have some kind of advantage(s) in regards to fumbling? If that were the case, why wouldn’t all teams be in a dome if it were, in fact, that critical?
Just because a team is an indoor team does that mean they have a huge (or any) advantage over all other teams? Rams, Cowboys, and Cardinals are all listed as dome teams and yet they are all at or below the league average for fumbles/number of plays.
Top 10 Teams (since 2010) records, playoff appearances, Super Bowl appearances, and Super Bowl wins
Does that not reflect how critical it is to hold on to the football (we hear that preached all of the time) and how it relates to the success of the team?
How about that, successful teams hold onto the football – how relevant it is depends on the coaching staff and, if deemed relevant, how well they ingrain that into the heads of the players and teach them to make sure, above everything else, that they hold on to the football.
Coaching Staff
Look at the coaching staff (not just the head coach) in terms of who they are, tenure, and what they (each member) are “known for” (their strengths).
Of course, these statements only begin to touch the surface of the relevant statistics that pertain to this, ultimately, non-issue. Call it what it is; for one half of a football game, the N.E. Patriots were using footballs that were, by rule, underinflated. The NFL has a fine amount in place, from what I have read, for this specific violation. It does not involve anything more than just that, a fine, and a minimal one at that.
All of this over-sensationalism of a particular issue is indicative of the society that we currently live in today where anyone with internet access (or a cellphone) can say anything in any way that they like on any subject by simply getting them out to the right places. If, added to that, the headline is set with the proper hook to get people to read it, then, so much the better (for whom I know not).
The author of the article, Warren Sharp, should be both proud and ashamed. Proud in the fact that he got a whole lot of folks interested in, and convinced of, his overall conclusion. Ashamed that he failed to provide adequate facts and statistics to formally draw any valid conclusion. In other words, he got people to read, and believe, his information, simply because he wrote a convincing article and provided just enough statistics to seemingly validate what he was trying to say.
1 – Walsh, January 2014, http://webcache.googleusercontent.com/search?q=cache:http://www.sharpfootballanalysis.com/blog/%3Fp%3D2932