Stats Article Suggesting QB Stafford Will Be a Bust

[maverick4]

Finding context in the Stafford debate - NFC North - ESPN

They used a math formula which took college production inputs (did not count yards thrown) and then came up with a number.

All of the QB's who had 0 or negative ratings from this formula turned out to be busts. Stafford had a -4 rating using this formula.

 

[Ochmed Jones]

Finding context in the Stafford debate - NFC North - ESPN

They used a math formula which took college production inputs (did not count yards thrown) and then came up with a number.

All of the QB's who had 0 or negative ratings from this formula turned out to be busts. Stafford had a -4 rating using this formula.

I do not know if he will be a bust per se, but I have a feeling he is going to need time, at least a year before he is productive. If Detriot does take him, I hope they start Culpepper all year and give Stafford time.

Of course if I GM in Detriot, I would pass on Stafford at #1 and hope Stafford or Sanchez is available at #20.

 

[maverick4]

This is a weak QB class. In my opinion no QB's are worth a day one pick.

That list had hits or misses based on the score, but anyone with a 0 or below was a complete bust. It isn't looking good for Stafford with a negative score.

 

[CanadianPat'sFan]

I do not know if he will be a bust per se, but I have a feeling he is going to need time, at least a year before he is productive. If Detriot does take him, I hope they start Culpepper all year and give Stafford time.

Of course if I GM in Detriot, I would pass on Stafford at #1 and hope Stafford or Sanchez is available at #20.

I totally agree.

Mayock was breaking down some tape of this kid on Total Access about a month ago and he made some throws which had me saying wow. Now I am not a QB guy or able to break down tape on this kid but what I saw impressed me. I do know that he is 2 years younger then Flaco and Ryan were when they started their rookie season so I think he should ride some pine before he gets his feet wet.

I agree that this is a very weak draft class for QB's so I really do not see any starting day one for any team. Heck it was due to Troy Smith's illness to get Flaco on the field so quick and in Atlanta well that was an easy choice

 

[ctpatsfan77]

This is a weak QB class. In my opinion no QB's are worth a day one pick.

Well, there is one QB worth a day one pick. . . . but he's already off the market.

 

[ctpatsfan77]

Finding context in the Stafford debate - NFC North - ESPN

They used a math formula which took college production inputs (did not count yards thrown) and then came up with a number.

All of the QB's who had 0 or negative ratings from this formula turned out to be busts. Stafford had a -4 rating using this formula.

And, BTW, showing that it's not a perfect correlation, Matt Cassel would get something like a -20 here.

 

[maverick4]

And, BTW, showing that it's not a perfect correlation, Matt Cassel would get something like a -20 here.

I'm not sure if that formula was made to project QB's who last started in high school!

 

[Nonentity]

Just worked it out for Tom Brady, he gets a rating of 19.95.

Formula:
(Career Starts x 0.5) + [(Career completion pct. - 60)x5] +[(Career touchdown-INT ratio - 2.25)x10]

25 career starts, 711 passing attempts, 443 completions (completion percentage of 62.3%), 35 TD, 19 INT (TD:INT ratio of 1.842)

stats.ath.umich.edu/football/cmaster.php?pkey=83 for his college stats.

 

[Ochmed Jones]

Just worked it out for Tom Brady, he gets a rating of 19.95.

Formula:
(Career Starts x 0.5) + [(Career completion pct. - 60)x5] +[(Career touchdown-INT ratio - 2.25)x10]

25 career starts, 711 passing attempts, 443 completions (completion percentage of 62.3%), 35 TD, 19 INT (TD:INT ratio of 1.842)

stats.ath.umich.edu/football/cmaster.php?pkey=83 for his college stats.

I still remembeer Brady in the bowl game against alabama when he was a senior at Michigan. He reminded me of Larry Bird that day, eyes of an assasin, heart of a lion.

But then after the game when we were doing interviews and I saw him without pads, I cringed. Talk about a stick man and stick men ususally break when they get hit in the NFL.

I am so glad BB listened to our QB coach that year.

 

[VJCPatriot]

Finding context in the Stafford debate - NFC North - ESPN

nice article. wow, Stafford looks like an even riskier pick after reading that article. I'm not sure what the Lions are going to do with the #1 overall. This looks like a bad year to be drafting a QB.

 

[Synovia]

Finding context in the Stafford debate - NFC North - ESPN

They used a math formula which took college production inputs (did not count yards thrown) and then came up with a number.

All of the QB's who had 0 or negative ratings from this formula turned out to be busts. Stafford had a -4 rating using this formula.

Decent article, but Lewin's system (which doesn't use TD-INT, just completion percentage and college starts) correlates much better to success...

Stafford grades out poorly in both. Sanchez, on the other hand, grades out as a Leaftian bust in Lewin's system, and as a good player in this one.

 

[maverick4]

Decent article, but Lewin's system (which doesn't use TD-INT, just completion percentage and college starts) correlates much better to success...

Stafford grades out poorly in both. Sanchez, on the other hand, grades out as a Leaftian bust in Lewin's system, and as a good player in this one.

I'd say those two formulas are good filters not to pick great QB's, but to cross out high risk busts.

Looks like both Sanchez and Stafford are bad bets.

 

[Synovia]

I'd say those two formulas are good filters not to pick great QB's, but to cross out high risk busts.

Looks like both Sanchez and Stafford are bad bets.

While I agree, Lewin's system has been pretty good picking out high ceiling guys.


The highest predictions in the last 10 year or so under Lewin's system are P. Manning, Cutler and Rivers.


The only real knock at this point, is Leinart, who Lewin had a a decent pro, but not a probowler. I think the jury is still out though.

 

[maverick4]

Can you please give me a link of how quarterbacks rate under Lewin's system? I'll need some more proof about why his system is superior to the other one, his seems way too simplistic. His article only came out in 2006:
FOOTBALL OUTSIDERS: Football analysis and NFL stats for the Moneyball era - Authors of Pro Football Prospectus 2008

Also, it appears this guy Lewin wrote that article as a freshman in college, and is probably now just in his junior year.

 

[ctpatsfan77]

Just worked it out for Tom Brady, he gets a rating of 19.95.

Formula:
(Career Starts x 0.5) + [(Career completion pct. - 60)x5] +[(Career touchdown-INT ratio - 2.25)x10]

25 career starts, 711 passing attempts, 443 completions (completion percentage of 62.3%), 35 TD, 19 INT (TD:INT ratio of 1.842)

stats.ath.umich.edu/football/cmaster.php?pkey=83 for his college stats.

BTW--those numbers become 2.5 and 5 instead of 5 and 10 for non-BCS QBs. So, Kevin O'Connell would have a number of 27 x 0.5 + (57.4 - 60) x 2.5 + (1.36 - 2.25) x 5 = 1.55.

 

[Synovia]

Can you please give me a link of how quarterbacks rate under Lewin's system? I'll need some more proof about why his system is superior to the other one, his seems way too simplistic. His article only came out in 2006:
FOOTBALL OUTSIDERS: Football analysis and NFL stats for the Moneyball era - Authors of Pro Football Prospectus 2008

Also, it appears this guy Lewin wrote that article as a freshman in college, and is probably now just in his junior year.

You'd need to buy PFP 2006.


And yes, Lewin was a freshman. You don't need a PHD to find correlations in data samples. Its simple statistics.



His system is superior because it correlates to future performance (as measured by DVOA/DPAR) very well, whereas the ESPN article's metric doesn't seem to correlate to anything.

 

[maverick4]

You'd need to buy PFP 2006.

Well, that was helpful. Thanks. Sorry, but I'm not buying it and your one-line response does nothing to prove Lewin's formula does a good job with correlation or is better than the ESPN one.

The ESPN formula also identifies high performing QB's but more importantly screens out the busts. I fail to see how anything you've written so far proves Lewin's formula is better.

Is there any link you can give to show Lewin's formula applied to QB's over the past 10 years, or prove that it is highly correlated with future performance? If not, then don't bother giving a useless one-line answer again.