If we lucked into a grade-A quarterback, how much difference would it make?

[Califan007 The Constipated]

A ten-per cent improvement in wins. A legit eight-win team would have a win pct of .500. A 10% improvement would raise it to .550 or 8.8 wins. With the luck factors canceling out, that’s probably 9 wins.

So by your analysis, The most a QB can effect the team is 10%. Right?

Which means the only thing that varies is whether or not the QB effects the team positively or negatively.

Put another way, a Grade-A franchise QB will effect the team in a positive way by 10%--and you said it would be 10% more wins. Likewise, a gawd-awful QB would effect the team in a negative way by no more than 10%--in this case, 10% more losses.

the difference between the absolute best QB and the absolute worst QB is a 20% swing in wins, all other things on the team being the same.

That means that the amount of wins that the absolute best QB can add to a team is 20% more wins than the absolute worst.

Which means that, all other things remaining exactly the same, the 2007 Patriots, who went 16-0 during the season, would have gone 13-3 if JaMarcus Russell had been their QB all season instead of Tom Brady.

Gotcha. :thumbsup:

[Oldfan]

So rather than applying the ten percent improvement on wins to the season win total apply it to the performance game by game. If almost half a teams losses are by 7 points or less a ten percent improvement game by game might result in more than a single game improvment. If we lose a game 10-9 a 10% improvement in our performance in that game would switch I from a loss to a win.

I wouldn’t know what stats to use to calculate a 10% improvement in our performnce.

Moreover, the scores of the games that might swing would simply be a matter of luck.

By the way you have not responded to the question I posed in an earlier post as below.

The question you pose in the OP is how much difference a grade A QB makes on wins and losses. I'm saying that it's not possible to answer this question simply by using the percentage contribution of the positions. There are too many other factors.

In an earlier post you stated that it was not possible to tell what the impact was that Ryan and Bradford had on their teams because there were too many other things happening for those franchises. I agree. Can you explain why that would be any different for any other QB or franchise?

Apples: 10 per-cent on the average is a reasonable estimate of the importance to wins of the QB position

Oranges: QB Bradford made a difference of five games in his team’s season.

[sportjunkie07]

Which means that, all other things remaining exactly the same, the 2007 Patriots, who went 16-0 during the season, would have gone 13-3 if JaMarcus Russell had been their QB all season instead of Tom Brady.

Gotcha. :thumbsup:

thank you for explaining oldfan's point perfectly.

[Conn]

I'm sorry, but I can't agree that a true Grade A QB only makes up 10% of a team's success.

I generally agree with your talent vs. Scheme outlook in regards to QB's, especially Brady.

But can I agree that Brees, Rivers, Peyton, and Warner only make up (or made up) ten percent of their team's success? No. And while you may argue that some (or all) of these are exceptions to the rule, I would argue that these are my standards for a Grade A QB. Flacco, Freeman, and the like may function and fit perfectly well as franchise QB's, but they aren't Grade A. At least, not yet.

So framed that way, i cant agree with your OP, though as I said, I do agree that there are many system QB's who reap the glory and the hype. I think that a true "Grade A" franchise QB has a much larger effect on their team's success than your average 10%.

[Oldfan]

SJ07 ~ 10% is way too low of an estimation. consider the fact that 100% of all offensive plays (since you are allowing a % for special teams) start at the qb; he has a direct affect on every play. he is worth much more than 10%. a qb even has a affect before the play with pre-snap reads and audibles.

Weak. YOU could be taught to hand the ball off on 40% of the plays. The QB only makes a difference on 60% on average.

bottom line, you cannot make an estimation for the level of importance for a qb. every scheme is different. every team relies on certain players more than others. the affect on each team would be different.

That’s why my estimate is the average value of the position based on a 60/40 pass/run ratio. Averages make those differences meaningless.

what is common sense though is that if you take manning off of the colts, they struggle, and by a lot more than 1 game. if you add manning to the bills, the bills would greatly improve. these things are obvious to nearly everyone except a crazy few that misguidingly attach a 10% number to all qb's.

So, your opinion is that I’m wrong, but you are out of arguments.

[sportjunkie07]

So, your opinion is that I’m wrong, but you are out of arguments.

common sense is a strong argument.

and we've had this discussion before in a similar thread. it boils down to common sense vs. trying to put a number on a qb's importance.

[Oldfan]

Califan ~ Likewise, a gawd-awful QB would effect the team in a negative way by no more than 10%--in this case, 10% more losses.

No, I anticipated that someone would try to play this game. Earlier, I explained that I only use four QB grades A, B, C, D -- A grade-D QB would be one who had proven he can start in the NFL, but he’s not as good as the A, B, C grade QBs. John Kitna was a grade-D guy.

---------- Post added July-14th-2011 at 05:53 PM ----------

common sense is a strong argument.

and we've had this discussion before in a similar thread. it boils down to common sense vs. trying to put a number on a qb's importance.

Common sense isn't common. You have just proved that.

[Califan007 The Constipated]

No, I anticipated that someone would try to play this game. Earlier, I explained that I only use four QB grades A, B, C, D -- A grade-D QB would be one who had proven he can start in the NFL, but he’s not as good as the A, B, C grade QBs. John Kitna was a grade-D guy.

Fine. If John Kitna is the starting QB for the Patriots all year in 2007, they go 13-3. Correct?

And for the record, who else is a "Grade D" quarterback?

---------- Post added July-14th-2011 at 02:56 PM ----------

Common sense isn't common. You have just proved that.

In other words, you are out of arguments.

[Oldfan]

I'm sorry, but I can't agree that a true Grade A QB only makes up 10% of a team's success.

I generally agree with your talent vs. Scheme outlook in regards to QB's, especially Brady.

But can I agree that Brees, Rivers, Peyton, and Warner only make up (or made up) ten percent of their team's success? No. And while you may argue that some (or all) of these are exceptions to the rule, I would argue that these are my standards for a Grade A QB. Flacco, Freeman, and the like may function and fit perfectly well as franchise QB's, but they aren't Grade A. At least, not yet.

So framed that way, i cant agree with your OP, though as I said, I do agree that there are many system QB's who reap the glory and the hype. I think that a true "Grade A" franchise QB has a much larger effect on their team's success than your average 10%.

Well, you said you don't agree, but I didn't find a counter to the points I made or any specific challenge to my estimate.

If my estimate is seriously flawed someone should be able to do a comparable estimate showing a much larger number for the QB. So far, no one has.

[skinfan2k]

The quarterback is the only position to change a team's record from 3-13 to 9-7 in 1 year. 10% my ass lol. Also, Oldfan, QBs affect runnings plays. its just not simply handing off to the RB

[sportjunkie07]

[/color]Common sense isn't common. You have just proved that.

Common sense, as described by Merriam-Webster, is defined as beliefs or propositions that most people consider prudent and of sound judgment, without reliance on esoteric knowledge or study or research, but based upon what they see as knowledge held by people "in common"

myself, califan, skinfan2k, etc.. 90% of people in this thread agree, and they dont need statistics to prove it. thats common sense.

[Oldfan]

Fine. If John Kitna is the starting QB for the Patriots all year in 2007, they go 13-3. Correct?

I'd say 14-2 in all probably. There isn't that much spread between A and D grades.

And for the record, who else is a "Grade D" quarterback?

I gave you Kitna as an example. More would serve no purpose here.

In other words, you are out of arguments.

Without Googling it, give me a definition for "argument" in the logical context. I ask because you didn't seem to know in this thread and another.:evilg:

[Loyal2Washington]

Yeah, because the reason the Pats struggled year one of the the rebuild was Drew Bledsoe, not the fact that Belichick completely purged the roster when he took over as the team was millions over the salary cap.

Tris you're smarter then that. You can throw out excuse after excuse the only real concrete evidence is this

Belichick before Brady- 41-56 with one playoff appearance in 5 seasons

Belichick after Brady- 121-38 with 8 playoff appearances, 4 SB appearances, and 3 SB wins

He was a well below avg. HC in the NFL before the Bledsoe injury. Now I'll give you Belichick is a great DC but coordinators don't always translate into great HC's.

Also the Pats were 8-8 the year before Belichick got there so it was hardly a Mike Shanahan-Redskin type of scenario, so the rebuild argument is kind of a stretch unless you think the Pats were rebuilding off an 8-8 season. Oh and Belichick did not have control of the roster so he purged nothing without Pioli and Kraft checking off on it. If you look at there roster of who Belichick already had it was hardly an empty roster, he walked into a team that already had Bledsoe, Bruschi, Chris Slade, Lawyer Malloy, Ted Johnson, Bruce Armstrong and Willie McGinest just to name a few and he still managed to take an 8-8 team with a solid nucleus and go 5-12 his first 17 games, including going 3-8 the last 11 of those 17 games.

Insert franchise HOF quarterback and presto instant dynasty.

[Califan007 The Constipated]

Well, you said you don't agree, but I didn't find a counter to the points I made or any specific challenge to my estimate.

If my estimate is seriously flawed someone should be able to do a comparable estimate showing a much larger number for the QB. So far, no one has.

What's there for us to counter? Where is your detailed explaination of why a QB will only play a 10% role in their team's success? It sure wasn't in the OP.

"The 10% value for the QB is the average value of the position based on a 60/40 pass to run ratio."

That is flawed right from the get-go, because it makes the assumption that the QB only has an effect on the 60% of the offense that is the passing game. It also ignores those QBs who are legitimate running threats.

"The value of the position can change somewhat with the scheme, but it does not change with the quality of the QB."

This makes zero sense. If the importance of the QB position is inflated by an offensive scheme, then the quality of the quarterback IN that system becomes even more important, and thus his value becomes even more inflated. A system that requires the QB to play a larger role in the offense than the average offense will, by default, make the quality OF that QB play a more valuable role in the team's fortunes.

---------- Post added July-14th-2011 at 03:12 PM ----------

I'd say 14-2 in all probably. There isn't that much spread between A and D grades.

I think that tells everyone all they need to know.

I gave you Kitna as an example. More would serve no purpose here.

How about leaving that decision up to me, eh sport? lol

Who are the other "Grade D" quarterbacks in your opinion? We can't test your hypothesis unless you give us data we can plug into the equation. So give us some in the form of "Grade D" quarterbacks.

Without Googling it, give me a definition for "argument" in the logical context. I ask because you didn't seem to know in this thread and another.:evilg:

Going by your definition earlier, NOT having an argument is when someone doesn't say anything concrete or factual to refute your point.

[Oldfan]

Common sense, as described by Merriam-Webster, is defined as beliefs or propositions that most people consider prudent and of sound judgment, without reliance on esoteric knowledge or study or research, but based upon what they see as knowledge held by people "in common"

myself, califan, skinfan2k, etc.. 90% of people in this thread agree, and they dont need statistics to prove it. thats common sense.

I see your problem. You think that the dictionary's description of "propositions that most people consider prudent and of sound judgment" applies to opinions on such thing as quarterbacks. It does not. The applicable term that applies to such opinions:

The Bandwagon Fallacy is committed whenever one argues for an idea based upon an irrelevant appeal to its popularity.

http://www.fallacyfiles.org/bandwagn.html

---------- Post added July-14th-2011 at 06:33 PM ----------

Califan007 ~ What's there for us to counter? Where is your detailed explaination of why a QB will only play a 10% role in their team's success? It sure wasn't in the OP.

I’ve explained four-five times how a challenger can prove my estimate bogus. Read the thread.

That is flawed right from the get-go, because it makes the assumption that the QB only has an effect on the 60% of the offense that is the passing game. It also ignores those QBs who are legitimate running threats.

The 10% is an average. Averages take differences into account.

This makes zero sense. If the importance of the QB position is inflated by an offensive scheme, then the quality of the quarterback IN that system becomes even more important..

I don’t have the patience to explain this again. Read the thread.

I think that tells everyone all they need to know.

You mean "everyone" on your bandwagon?

Going by your definition earlier, it's when someone doesn't say anything concrete or factual to refute your point.

How about when you want to know whether it’s an argument or not, you ask me?

---------- Post added July-14th-2011 at 06:35 PM ----------

The quarterback is the only position to change a team's record from 3-13 to 9-7 in 1 year.

There is no way to tell how much of a factor the QB change influenced such a disparity. Lots of factors were in play.

[Hitman21ST]

If my estimate is seriously flawed someone should be able to do a comparable estimate showing a much larger number for the QB. So far, no one has.

I actually did. I compared the Zorn roster with Campbell vs. the Zorn roster with an "A" QB, and then a much more complete roster with the same two QBs.

I gave you numbers. First year estimate on the complete roster for Campbell would yield 6-7 wins, and first year estimate on the complete roster for "A" QB would yield 9-10 wins. Conservative estimate that's a two game improvement, at the high end that would be a 4 game improvement. Considering we agreed that Campbell is a "C" QB, the range of improvement for an "A" QB is 28.5% - 66% (7 wins up to 9 wins or 6 wins up to 10). Considering that you and I had a civil discussion on this and we came to an agreement on it, I would say we've come up with a much larger number

[Oldfan]

I... Considering that you and I had a civil discussion on this and we came to an agreement on it, I would say we've come up with a much larger number

You can say it, but it's not true.

[gutlead74]

Well the equation the op invented supported his argument completely so he must be right.

I think we will win the super bowl in 11 because

Monkey farts= 20%

leprechaun condoms = 30%

and panda dung = 50%

That clearly equals 100% so skins win in 11!!

[Califan007 The Constipated]

I’ve explained four-five times how a challenger can prove my estimate bogus. Read the thread.

How about you first prove your estimate accurate? lol...How exactly did you come up with the QB having no more than a 10% effect on the team's success or failure? Once you do that, we will all gladly give you multiple valid reasons why your explanation is flawed.

Are you able to do that? Have you done it already? (I mean beyond the "QBs only have a 10% effect on the team's success because I say so" argument, that is)...If you have, just point me to the appropriate post.

And interestingly enough, you decided to just skip over my other request for a list of "Grade D" quarterbacks so that we may more thoroughly test your hypothesis and understand your "logic" (such as it is)...

For someone who seems so sure of their opinion, this should be a piece o' cake to do. :thumbsup:...and you should be EAGER to do so, what with your methodology being as infallible as it is and all.

[PokerPacker]

percentages don't work for this, I don't think. there are too many intertwining parts. for example, if you have a great runningback but no offensive line, the runningback will be ineffective. If you have a highly accurate QB behind that awful line, he'll be pummeled to the ground and won't have a chance to show off that accuracy. I think it would take a lot of thought to come up with a reliable system of how important the parts are.

On the other hand, if you have a great O-line and a great QB, and great WRs, the importance of the runningback diminishes. A poor runningback will affect the team, but after, say "good", you're dealing with diminishing returns, so perhaps his value would be measured as an inverted negative exponential function plus a constant where as t goes to infinity it approaches said constant which would be his maximum individual value. You'd have another similar function for the o-line. But then the change-up would be a function factoring in both the O-line and the runningback because the two of them are greater than the sum of their parts.

so for example, you'd have V = x(-e^-t) + c : where V = value, e = euler's number, t = talent, and c = the constant we decide is the greatest value of the individual position of runningback, x = the rate of growth in value (that would be a constant like c that we decide would determine how the law of diminishing returns affects the position). And to see one that takes two positions into account, youd have V = x(-e^-(t1*t2)) + c : where t1 is the talent of position 1 and t2 is the value of position two. So get the total value of the runningback taking his individual value into consideration and the talent of the O-line in front of him, it'd be V = x1(-e^-t1) + x2(-e^-(t1*t2)) + c1 + c2

any math geeks out there want to help turn this value system into a reality?

also, I'm considering the possibility of it being a logarithmic function of sorts, too, but I'm not sure how I'd make that one work.

[Laxpunk2006]

It amazes me that anyone still attempts to have logical, intelligent discussions with Oldfan.

[Califan007 The Constipated]

By the way, a HUGE reason why using percentages the way OF has done here is flawed is because of teams like the Colts.

Simple math shows why this is:

The percentages are supposed to show how much having a "Grade A" person at each position will increase the team's wins...

Me: "A 10% improvement of what, exactly?"

OF: "A ten-per cent improvement in wins. A legit eight-win team would have a win pct of .500. A 10% improvement would raise it to .550 or 8.8 wins. With the luck factors canceling out, that’s probably 9 wins."

The argument, based on the "logic" in the OP and what Oldfan said later in the thread, clearly states that even having a top of the line "Grade A" quarterback will give you AT MOST 10% more wins.

So let's take that logic to all facets of the team, using OF's own percentages, and apply them to the 3-13 Colts:

20% -- Coaching (0.6 more wins)

10% -- Special Teams (0.3 more wins)

35% -- Defense (1 more win)

35% -- Offense (1 more win)

A 3-13 team, having all starters suddenly playing at a "Grade A" level the next season, should only produce 3 more wins. That's it.

If the entire roster of starters were playing at a "Grade D" level when the team went 3-13, then at MOST the team would improve to is 9-7 (a 200% swing, from Grade D to Grade A performances all around).

Yet the Colts went from 3-13 to 13-3...they gained 10 more wins...which means they inproved their wins by over 330%.

But that can't be...because OF's hypothesis does NOT allow for any player or coach to be responsible for more than their allotted percentage of extra wins.

They only added in 6 new starters from their 3-13 season to their 13-3 season. A 330% improvement in wins would require a complete overhaul of all three units, along with a new head coach...not just 6 new players. And even THEN it would only result in 6 more wins, not 10.

Or maybe the entire team was full of "Grade F" players who magically became "Grade A" players in one offseason lol...yeah, that makes sense...the entire team--including the head coach--all improved astronomically in 9 months. Makes FAR more sense than saying the percentage model in the OP is flawed.

But what about the other 4 wins? Where did they come from?...Maybe "luck" gave them 4 extra wins. No, can't be that, because he said "luck factors" cancel themselves out and play no real role in a team's improvement.

Hmmm...those extra wins are damn near impossible to explain.

Or maybe, just maybe, Oldfan got his logic backwards. Maybe its not that a "Grade A" quarterback will only give a team 10% more wins...maybe it's that, of the extra wins a team does gain in one season, the "Grade A" QB was responsible for 10% of that additional win total. Wouldn't that make more sense?

Of course it would ...and if OF was more intent on finding an accurate measure of a "Grade A" quarterback's effect on a team than merely defending his belief that he already HAS found that measure, he would have seen the error of his logic earlier.

[MartinC]

As for Ryan and Bradford, there's no way to tell how much of a factor they were because there were a number of other changes within the organization at the same time.

There is no way to tell how much of a factor the QB change influenced such a disparity. Lots of factors were in play.

My point in this thread has been that it's not possible through the isolation of the position percentages you provide (which I have no problem with) to calculate the difference a grade A QB makes. There are too many other factors and it's not possible to isolate cause/effect.

The two statements from you above from your own mouth confirm you agree.

You seem to keep dragging any dispute back to "show me why my stats are wrong". It's not that the stats you provide are wrong but that you are trying to use them to show something they can't.

Good chatting but thats me out of this discussion. I think we have taken it as far as we can.

[MartinC]

so for example, you'd have V = x(-e^-t) + c : where V = value, e = euler's number, t = talent, and c = the constant we decide is the greatest value of the individual position of runningback, x = the rate of growth in value (that would be a constant like c that we decide would determine how the law of diminishing returns affects the position). And to see one that takes two positions into account, youd have V = x(-e^-(t1*t2)) + c : where t1 is the talent of position 1 and t2 is the value of position two. So get the total value of the runningback taking his individual value into consideration and the talent of the O-line in front of him, it'd be V = x1(-e^-t1) + x2(-e^-(t1*t2)) + c1 + c2

See this is exactly what I have been saying. Well sort of.

[Oldfan]

How about you first prove your estimate accurate? lol...How exactly did you come up with the QB having no more than a 10% effect on the team's success or failure? Once you do that, we will all gladly give you multiple valid reasons why your explanation is flawed.

Are you able to do that? Have you done it already? (I mean beyond the "QBs only have a 10% effect on the team's success because I say so" argument, that is)...If you have, just point me to the appropriate post.

And interestingly enough, you decided to just skip over my other request for a list of "Grade D" quarterbacks so that we may more thoroughly test your hypothesis and understand your "logic" (such as it is)...

For someone who seems so sure of their opinion, this should be a piece o' cake to do. :thumbsup:...and you should be EAGER to do so, what with your methodology being as infallible as it is and all.

You’re making a dumb request. If I had offered a logical argument, would you ask me to prove that it’s logical?

I gave you what I’m calling a reasonable estimate and I explained my method. It’s not up to me to prove that it’s reasonable. If you think it isn’t, proving it should be a snap if you can handle seventh grade math.

Use the 3 - 3 - 1 ratio for offense, defense, special teams. Use the 60/40 pass/run ratio. Have your estimate total 100%... and see if you can come up with a much higher estimate for the QB without reducing the other positions to unreasonable levels.

In the two threads where I’ve used this estimate, I’ve had 20 -25 posters call my numbers “arbitrary,” but only one made a sincere effort to post his own. Before I could get to it, another poster showed that, when adjusted to the same basis as mine, the challenger had come up with 9% for the QB.

And no, I’m not going off-topic to give you a list of grade D QBs.

---------- Post added July-15th-2011 at 07:04 AM ----------

percentages don't work for this, I don't think. there are too many intertwining parts. for example, if you have a great runningback but no offensive line, the runningback will be ineffective. If you have a highly accurate QB behind that awful line, he'll be pummeled to the ground and won't have a chance to show off that accuracy. I think it would take a lot of thought to come up with a reliable system of how important the parts are.

On the other hand, if you have a great O-line and a great QB, and great WRs, the importance of the runningback diminishes. A poor runningback will affect the team, but after, say "good", you're dealing with diminishing returns, so perhaps his value would be measured as an inverted negative exponential function plus a constant where as t goes to infinity it approaches said constant which would be his maximum individual value. You'd have another similar function for the o-line. But then the change-up would be a function factoring in both the O-line and the runningback because the two of them are greater than the sum of their parts.

so for example, you'd have V = x(-e^-t) + c : where V = value, e = euler's number, t = talent, and c = the constant we decide is the greatest value of the individual position of runningback, x = the rate of growth in value (that would be a constant like c that we decide would determine how the law of diminishing returns affects the position). And to see one that takes two positions into account, youd have V = x(-e^-(t1*t2)) + c : where t1 is the talent of position 1 and t2 is the value of position two. So get the total value of the runningback taking his individual value into consideration and the talent of the O-line in front of him, it'd be V = x1(-e^-t1) + x2(-e^-(t1*t2)) + c1 + c2

any math geeks out there want to help turn this value system into a reality?

also, I'm considering the possibility of it being a logarithmic function of sorts, too, but I'm not sure how I'd make that one work.

You are trying to go beyond the purpose of my estimate. I am not trying to come up with a formula to weigh the efficiency of a football team. My estimate is of the relative importance positions have to wins. It estimates just one number useful in trying to weigh the efficiency of a team.

The quality of the running back, the quality of the O-line, and the quality of the QB -- these factors that concern you -- have no bearing on the average importance of those positions. In other words, if y represents the quality-value of Tom Brady, then maybe .10y could become part of a grid analysis to weigh Tom Brady’s contribution to his team’s final grade.

---------- Post added July-15th-2011 at 07:47 AM ----------

By the way, a HUGE reason why using percentages the way OF has done here is flawed is because of teams like the Colts.

Simple math shows why this is:

The percentages are supposed to show how much having a "Grade A" person at each position will increase the team's wins...

The argument, based on the "logic" in the OP and what Oldfan said later in the thread, clearly states that even having a top of the line "Grade A" quarterback will give you AT MOST 10% more wins.

So let's take that logic to all facets of the team, using OF's own percentages, and apply them to the 3-13 Colts:

20% -- Coaching (0.6 more wins)

10% -- Special Teams (0.3 more wins)

35% -- Defense (1 more win)

35% -- Offense (1 more win)

A 3-13 team, having all starters suddenly playing at a "Grade A" level the next season, should only produce 3 more wins. That's it.

If the entire roster of starters were playing at a "Grade D" level when the team went 3-13, then at MOST the team would improve to is 9-7 (a 200% swing, from Grade D to Grade A performances all around).

Yet the Colts went from 3-13 to 13-3...they gained 10 more wins...which means they inproved their wins by over 330%.

But that can't be...because OF's hypothesis does NOT allow for any player or coach to be responsible for more than their allotted percentage of extra wins.

They only added in 6 new starters from their 3-13 season to their 13-3 season. A 330% improvement in wins would require a complete overhaul of all three units, along with a new head coach...not just 6 new players. And even THEN it would only result in 6 more wins, not 10.

Or maybe the entire team was full of "Grade F" players who magically became "Grade A" players in one offseason lol...yeah, that makes sense...the entire team--including the head coach--all improved astronomically in 9 months. Makes FAR more sense than saying the percentage model in the OP is flawed.

But what about the other 4 wins? Where did they come from?...Maybe "luck" gave them 4 extra wins. No, can't be that, because he said "luck factors" cancel themselves out and play no real role in a team's improvement.

Hmmm...those extra wins are damn near impossible to explain.

Or maybe, just maybe, Oldfan got his logic backwards. Maybe its not that a "Grade A" quarterback will only give a team 10% more wins...maybe it's that, of the extra wins a team does gain in one season, the "Grade A" QB was responsible for 10% of that additional win total. Wouldn't that make more sense?

Of course it would ...and if OF was more intent on finding an accurate measure of a "Grade A" quarterback's effect on a team than merely defending his belief that he already HAS found that measure, he would have seen the error of his logic earlier.

Cali, you can cherry-pick an exception but it won’t disprove a general rule. And, you can cherry-pick an exceptional jump from 3-13 to 13-3 but it won’t disprove an estimate based on Probability.

Did the Colts actually improve by ten games in one year? Surely they didn’t, because they were back to six wins two years later. They didn’t get back to the 13 game level for six years. The 3-13 to 13-3 was an anomaly. Some improvement is likely, but there was most likely a combination of bad luck in the 3-13 season and good luck in the 13-3 season added in.