Film Room: Ryan Tannehill

[fyo]

Great feature at FootballOutsiders.com on Ryan Tannehill.

http://www.footballoutsiders.com/film-room/2016/film-room-ryan-tannehill

The article includes a lot of nice gif "videos".

A couple quotes from the article:

Tannehill has always been a good quarterback, but he has never had the supporting cast to show it off consistently. Last year he had an accuracy rate of 80.8 percent, but his receivers cost him 63 completions on accurate passes for at least 717 yards and six touchdowns. Only Aaron Rodgers lost more completions to wide receiver error; only Cam Newton lost more yards to wide receiver error. Rodgers, Newton, and Andrew Luck were the only quarterbacks to lose more touchdowns.

Yet despite constantly being target practice for defenders, Tannehill's mechanics have never suffered. He has never refused to plant his feet or step into throws. He has always been willing to take contact to prioritize delivering the ball.

[Dolfanalyst]

Tannehill has always been a good quarterback, but he has never had the supporting cast to show it off consistently. Last year he had an accuracy rate of 80.8 percent, but his receivers cost him 63 completions on accurate passes for at least 717 yards and six touchdowns. Only Aaron Rodgers lost more completions to wide receiver error; only Cam Newton lost more yards to wide receiver error. Rodgers, Newton, and Andrew Luck were the only quarterbacks to lose more touchdowns.

Tannehill threw 586 passes, Rodgers 572, Newton, 495, and Luck 293.

I suspect that if you converted these "receiver errors" to a percentage of the passes thrown and not just an absolute number of them, Tannehill would be non-significantly different from the average QB in the league in that regard.

This is what can happen when "film" guys step into the world of numbers.  They can make very simple errors such as these that can then generate misleading information.

That said, Tannehill's game last week was great.

[Spider-Dan]

I suspect that if you converted these "receiver errors" to a percentage of the passes thrown and not just an absolute number of them, Tannehill would be non-significantly different from the average QB in the league in that regard.

...except that Tannehill was 8th in pass attempts last season, and none of the players with more pass attempts than him were cited.

[Dolfanalyst]

...except that Tannehill was 8th in pass attempts last season, and none of the players with more pass attempts than him were cited.

The lower the number of pass attempts, the greater the percentage the same number of receiver errors would equate to.

500 pass attempts with 50 errors equals 10% errors.  400 pass attempts with 50 errors equals 12.5% errors.

The fact that players with lower numbers of pass attempts were cited supports the point I made above.

Additionally, if you had to put money on it, and you didn't know whether Tannehill was any different from the average QB in this regard, you'd be wise to put money on his being in the average range, since the variable is likely normally distributed, and the average range contains the greatest number of QBs.

[Spider-Dan]

They didn't cite percentages; they cited occurrences.

You claim that Tannehill was probably average in the percentage of receptions/yards/TDs lost to receiver error.  If that's the case, then why weren't the QBs with more pass attempts than him cited in a list of league-leading occurrences?  Was every one of them far more fortunate than average?

The very fact that Tannehill, Rodgers, Newton, and Luck were cited as top victims in the number of occurrences of receiver errors - when none of them were close to leading the league in pass attempts - tells you that it's unlikely that their percentage of this event was around average.

[Dolfanalyst]

They didn't cite percentages; they cited occurrences.

You claim that Tannehill was probably average in the percentage of receptions/yards/TDs lost to receiver error.  If that's the case, then why weren't the QBs with more attempts than him cited in a list of league-leading occurrences?

The very fact that Tannehill, Rodgers, Newton, and Luck were cited as top victims in the number of occurrences of receiver errors - when none of them were close to leading the league in attempts - tells you that it's unlikely that their percentage of this event was around average.

The average team in the league attempted 571 passes in 2015 (standard deviation was 62 attempts).  Tannehill attempted 586 passes, which was a mere 0.23 standard deviations above the league average.

Only one QB in the league (Rivers) threw a number of passes that was greater than a standard deviation above the league average.  Therefore we shouldn't necessarily expect the seven QBs who passed the ball more than Tannehill to be mentioned with regard to the number receiver errors they experienced.  They could've hypothetically each had one fewer receiver error than Tannehill and not been mentioned for that reason, i.e., Tannehill had "more" than they did, and all of them would've nonetheless been victimized by roughly the same number of errors per pass attempt.

Again, I revert to the likelihood that receiver error is a variable that is likely normally distributed, and so in the absence of sufficient information, the best bet is to consider Tannehill to have been in the average range in that regard.  If the author would've presented more information, then we could make a definitive determination.

Now Andrew Luck on the other hand, whose number of pass attempts was a whopping 4.5 standard deviations below the league mean, is another story.  For him to be mentioned in the group of QBs victimized by receiver errors indeed suggests he was victimized by a non-average number of errors per pass attempt.

[Pappy13]

...except that Tannehill was 8th in pass attempts last season, and none of the players with more pass attempts than him were cited.

For a supposed numbers guy his personal bias toward Tannehill is astounding. You'll notice that he didn't bother to do the numbers and find out or give those stats, they aren't that important in this case.

[Dolfanalyst]

For a supposed numbers guy his personal bias toward Tannehill is astounding. You'll notice that he didn't bother to do the numbers and find out or give those stats, they aren't that important in this case.

Like I said, in the absence of that information, the best bet is on Tannehill's (and any QB's) being in the average range in that regard.

Now if you would like to find those statistics and provide them (I do enough of that myself), I'd be happy to stand corrected.

Or are they not important to you in this case?