Football Outsiders' take on Garoppolo and Brissett

[Ring 6]

FWIW. . . .



IOW, what they're saying is, all other things being equal, they'd still expect an average QB to do better than Brissett did on those nine passes.

No, they use a gauge that has nothing to do with his play, but penalizes him for the play call.
What could another QB have done better than complete 6 of the 9 throws he was told to make with 2 being dropped?

 

[PatsFan2]

Last game showed they are not afraid to make the Biscuit pass I hope he gets the green light alot Tonight. Biscuit is a unknown the Texans don't have an advantage as some think.

 

[primetime]

This is part of the problem with trying to judge football with statistics.
In the case Brisset sucks because he executed the plays that were called effectively, but the person analyzing it didn't like the play calls.
We have no clue what Brisset would have done if they called a 15 yard dig, but with a large lead, and a punt being better than a turnover, calling screen passes was wise. You cannot judge his performance by what play was called. Well, you can but like most attempts to reduce football to statistics, it fails miserably.

There's nothing to be upset about here. The sample size is so small in any case that it's not meaningful but the stat itself is an objective analysis with a transparent methodology, so it's not the 'person analyzing it' who's at fault but that the sample size is too small and non-representative. The reason the stat is called ALEX is tongue-in-cheek, because Alex Smith checks down on 3rd down a lot and fails to get the first down. Yeah, there's obviously a gameflow thing here, and like any stat it's important to analyze in context. If anything, the problem is presenting Brissett's score at all since he only had, what, 3 or 4 third down throws?

Note that on the stats page FO puts him in a "non-qualifying" section anyways because of this.

 

[Ring 6]

There's nothing to be upset about here. The sample size is so small in any case that it's not meaningful but the stat itself is an objective analysis with a transparent methodology. The reason the stat is called ALEX is tongue-in-cheek, because Alex Smith checks down on 3rd down a lot and fails to get the first down. Yeah, there's obviously a gameflow thing here, and like any stat it's important to analyze in context.

Who is upset?
Its a terrible method to judge football, and I am just pointing that out.
Statistics are great for baseball, and they are awful for football. Always have been, and making them more complex isn't going to change that.
By the way the sample size is, so far, his career. He did what he was asked to do, and did it well. You cannot make up a statistic and use it to judge his play. Well, you can, but it would be wrong.

 

[Ring 6]

Last game showed they are not afraid to make the Biscuit pass I hope he gets the green light alot Tonight. Biscuit is a unknown the Texans don't have an advantage as some think.

Of course they have an advantage if Brissett plays, because he isn't as good a QB as Jimmy G or Brady.
To say an unknown creates and advantage would be to say every rookie QB should be thrown in and expected to flourish.

 

[Clonamery]

....not meaningful ....

Exactly!

 

[primetime]

Who is upset?
Its a terrible method to judge football, and I am just pointing that out.
Statistics are great for baseball, and they are awful for football. Always have been, and making them more complex isn't going to change that.
By the way the sample size is, so far, his career. He did what he was asked to do, and did it well. You cannot make up a statistic and use it to judge his play. Well, you can, but it would be wrong.

Did you even check the page or are you judging this solely on the way it was presented in the OP? FO puts him in a "non-qualifiers" section because he doesn't have enough attempts to form a meaningful sample for their methodology to be useful. I agree that football statistics are fraught, but the ALEX score is about as straightforward as it gets. You do, however, need a meaningful sample of third down throws in representative game situations for it to hold any meaning and smooth out screen passes. Brissett, as you've noted, has only a tiny sample of third down throws in skewed situations.

 

[primetime]

Exactly!

That doesn't mean the stat itself isn't, just that you need a large and representative enough sample to be able to use it for inference. An admitted problem with football is that there just aren't enough plays and thus statistical events for this to happen except perhaps over the course of a full season, which makes these kind of statistics only useful in retrospect and in many cases only for players who were on the field for most plays in a given season. It really limits their usefulness, but doesn't mean they aren't useful.

That's an issue with non-"advanced" stats too, though. A player with 40 carries who gets 2 yards on each of 39 carries and then one where he breaks one for 90 has a yards per carry of 4.2, which might lead you to think he's a reliable four yards and a cloud of dust guy. Everything in context.

 

[I.M. Fletcher]

Funny that all the focus is on how they judged brissett poorly on 9 passes, but nothing about garoppolo having being so highly rated by the same metrics over a larger sample size.

 

[denverpatsfan]

Why even suit up? The Pats are doomed.

 

[TBR]

Two bad Edelman drops, a lotta screens, some quick dump offs to White, and two very athletic runs... give him a gameplan tailored to his strengths and he'll be fine.

I can see why they're still outsiders.

Maybe it's just the way he gets up, but Bennett always seems to get up slowly off the ground.

 

[pwes]

Why are people pissed at the site for compiling stats? They do this for every player for every game. Football outsiders didn't compare JG and JB, the OP did. Some of you guys really over react to everything. It's interesting to see even if you shouldn't take much stock in it.

 

[primetime]

I don't expect Brissett to be a star tonight. Making your first start in a primetime game against an elite defense is tough. He has a good supporting cast. They can win. He's almost certainly a more talented football player than Trevor Siemian, for instance, but Brissett probably is not going to carry them to victory, and if he struggles we shouldn't declare him a bust, either.

 

[Ring 6]

Did you even check the page or are you judging this solely on the way it was presented in the OP? FO puts him in a "non-qualifiers" section because he doesn't have enough attempts to form a meaningful sample for their methodology to be useful. I agree that football statistics are fraught, but the ALEX score is about as straightforward as it gets. You do, however, need a meaningful sample of third down throws in representative game situations for it to hold any meaning and smooth out screen passes. Brissett, as you've noted, has only a tiny sample of third down throws in skewed situations.

Football statistics are extremely meaningless. Football is a game of context, a game of situation, a game of scheme and execution by 11 players. To try to boil that down to statistics is nice for the lazy who want someone to tell them what to think, but will never be accurate and meaningful. Its not baseball.
In this example a team with an excellent screen pass concept and execution would result in these people reducing their rating of the QB who carries out the well-designed and effective play. You are not a worse QB because you execute the play call instead of throwing into coverage deep downfield.

 

[ctpatsfan77]

Why are people pissed at the site for compiling stats? They do this for every player for every game. Football outsiders didn't compare JG and JB, the OP did. Some of you guys really over react to everything. It's interesting to see even if you shouldn't take much stock in it.

I will point out that I didn't draw any conclusions; I simply pointed out the wide disparity.

 

[patchick]

I don't expect Brissett to be a star tonight. Making your first start in a primetime game against an elite defense is tough. He has a good supporting cast. They can win. He's almost certainly a more talented football player than Trevor Siemian, for instance, but Brissett probably is not going to carry them to victory, and if he struggles we shouldn't declare him a bust, either.

The Bill O'Brien Texans are also a particularly tough draw for a first-time starter -- they're very much Pats-like in their game-by-game scheming and will do their best to overwhelm him.

On the flip side, BB's Patriots have sometimes struggled against blank-slate QBs they have no film on.

 

[Ring 6]

I don't expect Brissett to be a star tonight. Making your first start in a primetime game against an elite defense is tough. He has a good supporting cast. They can win. He's almost certainly a more talented football player than Trevor Siemian, for instance, but Brissett probably is not going to carry them to victory, and if he struggles we shouldn't declare him a bust, either.

If the fact that it is a prime time game causes him to poop his pants, then that is a big sign he isn't the right guy.

 

[Ring 6]

Why are people pissed at the site for compiling stats? They do this for every player for every game. Football outsiders didn't compare JG and JB, the OP did. Some of you guys really over react to everything. It's interesting to see even if you shouldn't take much stock in it.

They didn't? Then why are they ranked, and why is Brissett ranked horrendously based PRIMARILY on the play call?

 

[patchick]

They didn't? Then why are they ranked, and why is Brissett ranked horrendously based PRIMARILY on the play call?

It's just a stats page that happens to have an absurdly small sample size at this point in the season. It's up to users to understand that those stats aren't very meaningful yet -- just as it's up to users to understand that, say, a score posted 10 minutes into a game isn't very meaningful yet. That doesn't mean the score shouldn't be reported.

 

[ctpatsfan77]

They didn't? Then why are they ranked, and why is Brissett ranked horrendously based PRIMARILY on the play call?

Actually, re-watching Brissett's plays, I can tell you why he was rated so horribly: (1) the passing stats don't include the sack he took, but FO stats do. (2) He did cough up the football, and FO penalizes a QB for a kept fumble the same amount as for a lost fumble. (It doesn't penalize the QB if a WR/RB/TE fumbles, though.)