Cian Fahey on Tom Brady

[DarrylS]

I am still trying to figure out DVOA, the product of idle minds trying to think of a way to make themselves somehow relevant in world of NFL stats..

Sounds good, but no idea wtf they are talking about..

When this lesson is complete then we can discuss: DYAR (Defense-adjusted Yards Above Replacement),
Adjusted Line Yards or Pythagorean Projection...

 

[bormio]

Exactly. "Interceptable passes" is subjective, and once someone adds their opinions into their statistics, it all becomes worthless

Kind of like "drops".

 

[fair catch fryar]

Cian Fahey's analysis of Brady's 2016 season is available at Pre-Snap Reads Quarterback Catalogue 2017 ($19.99)

Keep your $20 bucks. This is cheaper, softer and will last longer.

 

[QuantumMechanic]

I am still trying to figure out DVOA, the product of idle minds trying to think of a way to make themselves somehow relevant in world of NFL stats..

The concept is pretty simple.

You have a master chart of "success points" for a play outcome given down, distance, etc. So, for example, picking up 3 yards on 3rd-and-2 will be worth more than picking up 18 yards on 3rd-and-20.

With that chart in hand you go through every play for every team and assign the requisite number of points to it.

You adjust them for league averages (so if you do crappy on something, but you're less crappy at it than the average team, that's a positive and if you do good on something, but you're less good at it than the average team, that's a negative).

Then you adjust it that based on the defense of team you're facing (for offensive DVOA) or for the offense of the team you're facing (for defensive DVOA) to account for the DVOA analog of the fact that rolling up 400yd against a team that averages allowing 450yd isn't as good as getting 300yd against a team that averages allowing 250yd.

In the end you arrive at a per-play efficiency metric normalized to league average and taking into account the actual defenses (and offenses) you faced.

 

[PatsBoy12]

Slater doesn't count as a deep threat .

Doesn't mean the Pats won't send him on the fly route at least twice in 2017. It'll be incomplete. LOL

 

[slam]

The concept is pretty simple.

You have a master chart of "success points" for a play outcome given down, distance, etc. So, for example, picking up 3 yards on 3rd-and-2 will be worth more than picking up 18 yards on 3rd-and-20.

With that chart in hand you go through every play for every team and assign the requisite number of points to it.

You adjust them for league averages (so if you do crappy on something, but you're less crappy at it than the average team, that's a positive and if you do good on something, but you're less good at it than the average team, that's a negative).

Then you adjust it that based on the defense of team you're facing (for offensive DVOA) or for the offense of the team you're facing (for defensive DVOA) to account for the DVOA analog of the fact that rolling up 400yd against a team that averages allowing 450yd isn't as good as getting 300yd against a team that averages allowing 250yd.

In the end you arrive at a per-play efficiency metric normalized to league average and taking into account the actual defenses (and offenses) you faced.

One thing I've wondered that I've been too lazy to ever find out. How much is each situation isolated?

Is 1st and 10 from your 20 a different situation according to DVOA than 1st and 10 from the 50, for example?

Is 1st and 10 in the 1st quarter a different situation from 1st in 10 with 2 minutes left ahead or behind by two touchdowns?

I'm guessing "No" in each case, that every 1st and 10 is equal but that you could further inquire and say "Brady's DVOA while trailing in the 4th quarter is ...." but I'm not sure.

 

[PatsWSB47]

The concept is pretty simple.

You have a master chart of "success points" for a play outcome given down, distance, etc. So, for example, picking up 3 yards on 3rd-and-2 will be worth more than picking up 18 yards on 3rd-and-20.

With that chart in hand you go through every play for every team and assign the requisite number of points to it.

You adjust them for league averages (so if you do crappy on something, but you're less crappy at it than the average team, that's a positive and if you do good on something, but you're less good at it than the average team, that's a negative).

Then you adjust it that based on the defense of team you're facing (for offensive DVOA) or for the offense of the team you're facing (for defensive DVOA) to account for the DVOA analog of the fact that rolling up 400yd against a team that averages allowing 450yd isn't as good as getting 300yd against a team that averages allowing 250yd.

In the end you arrive at a per-play efficiency metric normalized to league average and taking into account the actual defenses (and offenses) you faced.

That's all well and good but unless you account for the temperature and more importantly the wind chill factor along with the percentage of glare from the sunlight on the ball then your success points conclusions can be hopelessly skewed.......not to mention the better teams play a lot more night games and you'll have to account for the number of lumens each stadium's lights produce. I've been playing with this with so much time on my hands in the off season and though the study is not complete yet it looks like It's Ryan Fitzpatrick is actually the greatest of all time and Brady and Montana are somewhere in the middle. A good idea for the fans attending games would to have each play scored according to this metric on the scoreboard. Then they would know how loud they should cheer or boo based on how high or low the score came out.

 

[40yrpatsfan]

I had a hard time finding any bad throws from Brady in SB51. On a couple of overthrows to Edelman, JE was impeded by a step after the ball was thrown. On JE's miracle catch, Alford made an incredible play to get a hand on it as his back was to the play until the last instant. On the pick-6, you just tip your hat to the D for pulling off a trick coverage post-snap that NE had no way to anticipate.

On the other hand, TB tortured Atlanta in the 4th qtr with deep outs to Mitchell where his arm strength was more than satisfactory, and he sent a rifle shot to Amendola just after the miracle catch by JE. Not to mention a couple of beautiful touch passes he dropped in to JE and Marty.

 

[QuantumMechanic]

One thing I've wondered that I've been too lazy to ever find out. How much is each situation isolated?

Is 1st and 10 from your 20 a different situation according to DVOA than 1st and 10 from the 50, for example?

Is 1st and 10 in the 1st quarter a different situation from 1st in 10 with 2 minutes left ahead or behind by two touchdowns?

Yes.

From their website: "...adjusted for a number of variables. These include down and distance, field location, time remaining in game, and the team’s lead or deficit in the game score."

 

[Ivan]

It was in jest. The 2nd part is true - arm strength issues first show up on throw when defenders get at the QBs feet (the Falcons did not do that in OT). Not saying this is the case with Brady - but it is where to look for signs of trouble.

The Patriots should have known this when they drafted Brady, he threw more interceptable passes than any other QB at his Combine, and he was also last in the THROWING OFF YOUR BACK FOOT WHILE FALLING DOWN drill. Both clear signs of his future almost sucked but turned out great career.

 

[Bledsoehasaring]

I have watched every play for the last 4 years and my major concern about tb12 is he seems to be getting older year by year. In 2013 he was younger by quite a few days. The 2017 version worries me cause he will be older. Older than the 2016 version that won the superbowl. You can buy my more detailed analysis for 19.99.

 

[patsinthesnow]

NFL1000: Final Regular-Season Breakdown of the League's Top Players

This guy thought Bradford & Newton were top 5 QBs last year & Matt Ryan was 10th.

He blocked me for calling out "genius" Kyle Shanahan's absurd play calling in the SB.

He had Vernon Adams as his #1 QB in the 2015 draft. Guy was an atrocious prospect and didn't have an NFL skillset. He went undrafted and is in the CFL.

This guy is a QB analyst but simply doesn't understand the position or the sport enough. That's what happens when you start watching football in 2004 and pretend to know it.

Guy is the definition of a fraud.

 

[K. Dog]

One of the most interesting statistics that Fahey cites is that Brady is asked to throw more than other QBs. Specifically, "Brady threw 104 passes while playing with a two-touchdown lead during last year's regular season". Of the league's other quarterbacks, only Matt Ryan threw more than 60". That is, Fahey attributes Brady's declining arm strength through the 2016 season partly to overuse.

 

[Deus Irae]

That is, Fahey attributes Brady's declining arm strength through the 2016 season partly to overuse.

He claims there was declining arm strength. He doesn't come anywhere near to proving it. So looking for attribution is useless. It was a terrible piece, and I'd fire him for it.