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

[Oldfan]

So I guess we should dump our entire criminal justice system.

Henceforth, all trials will go proceed according to this format:

prosecution - "your honor, he's guilty as hell"

defense - "Nu uh"

Guilt or innocence is a question of fact.

It's not a question of reasonableness.

[Califan007 The Constipated]

I would guess that since none of have time to do a 4000 page treatise on the subject, all facts could be considered to be "cherry-picked".

I would note that of all three redskin Super Bowls, Joe Thesimann comes closest to being an elite QB. Rypien and Williams would only be considered average, at best. But all three had excellent supporting casts. In fact, one could easily speculate that if offense were made up only of receving corps, running backs, OL and QB, that in each and every Super Bowl winning Redskin team, the QB was outclassed in each instance by the supporting cast.

I would also suggest looking at recent events, since many of the youngest fans on this board tend to discount teams prior to 2000 as "different than today's NFL". Look at Matt Cassel for my cherry-picked example.

Matt Cassel is the constant in my case in point.The variable in this case is the supporting cast - i.e. the receivers, the coaches, the offensive line. The results show a dramatic difference using the same constant - Matt Cassel. With the Patriots in 2008, he produces 11 wins. He throws for 21 touchdowns and 11 interceptions. He throws for 3693 yards. They miss the playoffs, but the team is obviously still a very good team.

Take the same QB, the same Matt Cassell. Change the supporting cast by moving him to Kansas City, running an offense very similar to New England's.. Does he have an 11 win season, a 3600+ passing attack? No. The stats are clear. In 2009, the same 11 game winning QB produces 4 wins with Kansas City, In 2009, with Kansas City, he has 2924 yards, 16 TD and 16 Interceptions.

The point is obvious. Same QB, different supporting cast - different results. I can't say that the 10% the QB brings to the table is perfectly accurate, but it appears close.

The Chiefs only had 4 different starters from 2009 to 2010. Yet they went from 4-12 to 10-6...150% win improvement. In your example you don't allow for Cassel slumping for one year, so we won't allow for any of the other players slumping in 2009, either, and assume that their level of play remained constant between 2009 and 2010. Do you really think those 4 new starters propelled the Chiefs to increase their wins by 150%?

I can tell you from looking at the stats, that the only person whose stats improved dramatically (from "Grade-C/D" to "Grade-A/B") from 2009 to 2010 was Cassel.

---------- Post added July-15th-2011 at 03:08 PM ----------

I'd consider 10% a significant number stated in this context. If anything' date=' I think it enforces the importance of the position.[/quote']

But adding a HOF level QB like Manning to an 8-8 team will only get the team one more win. How important can the position be?

[Oldfan]

In 2009, Cutler flew to Chicago from Denver and all his numbers went way down. Orton flew to Denver from Chicago and his numbers went way up. Cassel flew from Boston to Kansas City and his numbers went way down.

Must have been the plane rides, right, Cali?

-

But adding a HOF level QB like Manning to an 8-8 team will only get the team one more win. How important can the position be?

One more win at that level is very important. A good bounce or two and you're a playoff team.

[Shkspr]

Here, I began by assuming that we were a true six-win team quarterbacked at C minus level. Since our quarterbacking wasn’t bottom-grade, the full 10% improvement isn’t in play. So, I ballpark-estimated a one-win improvement (.375 to .438) if we lucked into a grade-A quarterback. So, let’s assume we would become truly a seven-win team with a grade-A quarterback. What are the chances of winning 10+ games and making the playoffs? Would you mind figuring that out for us?

Sure.

If a team had a true talent level of six wins, the odds of them finishing the season with a record of 10-6 or better, all other things being equal, would be about 27-1. If a player were strong enough to raise the talent level of the entire team by a win, to .438, the odds of winning 10+ games in a season drop to 9.5-1. Again, we see that the addition of one win to the true talent level will basically triple the chances of a "good" season.

However, I think that there's more parity amongst teams than that. I think that at any given time, the worst teams in the league run about .300 true talent level, and the best teams run .700. It'll take better aligned data than I have available to me here to be any more precise than that, but it feels right based on the number of outlier seasons it produces. I also think that in general Washington tends NOT to be one of the worst couple of teams in the league, but historically at about the third quartile point - ranking 20-24 in the league. That would peg us at a true talent level of .450 or so, which is why I started there. I'm kind of the opinion that we were closer to a .375 team under the Zorn administration and that 2010 was an example of a bad year under Shanahan.

Trivia question: What's the next record in this sequence?

9-7

6-10

6-10

7-9

12-4

[anar-k21]

it would help alot obviously....we still need help with OL though of course

[Oldfan]

Sure.

If a team had a true talent level of six wins, the odds of them finishing the season with a record of 10-6 or better, all other things being equal, would be about 27-1. If a player were strong enough to raise the talent level of the entire team by a win, to .438, the odds of winning 10+ games in a season drop to 9.5-1. Again, we see that the addition of one win to the true talent level will basically triple the chances of a "good" season.

However, I think that there's more parity amongst teams than that. I think that at any given time, the worst teams in the league run about .300 true talent level, and the best teams run .700. It'll take better aligned data than I have available to me here to be any more precise than that, but it feels right based on the number of outlier seasons it produces. I also think that in general Washington tends NOT to be one of the worst couple of teams in the league, but historically at about the third quartile point - ranking 20-24 in the league. That would peg us at a true talent level of .450 or so, which is why I started there. I'm kind of the opinion that we were closer to a .375 team under the Zorn administration and that 2010 was an example of a bad year under Shanahan.

Trivia question: What's the next record in this sequence?

9-7

6-10

6-10

7-9

12-4

That’s good stuff.

Thank you.

Your peg of .450 true talent level sounds very close to our win percentage in the Snyder years. I figured it out once, but I have forgotten the exact number.

An article I read had the Skins as the most consistently mediocre team in the NFL. Our policy of trading draft picks for vets and depending more than most teams on free agent vets plausibly explains why we have not been very good or very bad.

The trivia question -- I’ll take a stab at 8-8

[Chump Bailey]

The trivia question -- I’ll take a stab at 8-8

My guess too OF.

[Est.1974]

But adding a HOF level QB like Manning to an 8-8 team will only get the team one more win. How important can the position be?

But I don't agree with that. It would add more. The stats in the OP are open to interprepation in my view.

In the OP, you have something like 5% protectors, 5% receivers & 10% QB against passing game.

At a minimum, you have 5 OL contribution to that 5% protection, and say 3 WR/TE contributing to the 5% receivers. Yet only 1 QB for the full 10%.

Therefore, the weighting should be altered to reflect the % over the number of players contributing to that number.

Or is that already built into the numbers here ?

[Mooka]

I don't necessarily agree with assigning % points and using football outsider numbers. They are better for noticing trends then actual rules of thumb or predictions.

But I like the pun in your thread title Oldfan.

[Oldfan]

But I don't agree with that. It would add more. The stats in the OP are open to interprepation in my view.

I'm not offering this idea as a precision instrument. So, let's not try to cut it too fine.

How much improvement Manning would make to a true 8-8 team would depend on the grade of the QB he replaced. My estimate would be 1-2 wins.

[Hitman21ST]

Guilt or innocence is a question of fact.

It's not a question of reasonableness.

So where does "beyond a reasonable doubt" come into play?

---------- Post added July-15th-2011 at 08:15 PM ----------

If you would learn to debate, you wouldn't need to resort to taking quotes out of context to distort meanings. The only thing you accomplished is showing your lack of regard for the truth.

If you would use my whole quote and not take it out of context, you would realize that what LKB posted is, in actuality, the exact answer to your question. Wins Above Replacement shows exactly how much difference an "A" quarterback would make.

[Kindred]

if we lucked into a grade A QB like Luck, we would be extremely lucky indeed.

[Long-Time-Fan]

Nothing would change

We need way more than just 1 player

[Gerald362]

Ok, I am going to try to be good tonite. I believe that 10% as a player may be close to accurate as to the percentage of strength for a QB to any average team with all else being equal. However, when you take into account different playing styles and schemes for the other players and the coaching staff, the value of that QB to that team could be significantly higher. I know that there will be the mention of Peyton Manning, but I believe that he has to be placed in a totally different class of QB since he is virtually a coach as well as a player, so we will not include him in the equations.

I think that the arguements offered here may not be as far apart as you might think. I also believe that when you see a QB come in and instantly see that there is a huge difference in wins, it is not just because the QB is that much better than his predecessor, but that he is a better fit for the talent and coaching that is in place.

Call me a fence-sitter, but I believe that both sides have valid arguements............if you have grade A QB on team A with fit in scheme A and replace him with grade C qb with fit in scheme A, there will be a small but significant drop in wins. This is where Oldfan's arguement may hold water. However, it rarely works like that because you do not usually see a replacement of identical scheme strength QB when a change is made.

Right now, for the Redskins, even a grade A QB would not make much of a difference in the overall rating of the team unless a couple of things were to happen. First, he would have to be strong in the scheme that Shanny wants to run. Second, he would have to have a good enough line to allow him to run the plays. If those did not happen, there would not be much of an improvement. This is where the other side of the discussion would hold water. The QB is a 10% upgrade in talent over what was present before, but without the meshing of the other parts, the total addition would not be as significant. There is an additive effect on matching strengths that is not just explaned by total talent.

[88Comrade2000]

A grade A QB could make us a consistent contender if we surround him with players; see Peyton Manning and the Colts. A grade B QB could still get the job done.

[Shkspr]

The trivia question -- I’ll take a stab at 8-8

The official answer is that we won't know until Monday night when the Nationals complete their next set of sixteen games. Right now they're at 6-7, and I'm looking at matchups and thinking we're going to win one out of the next three, so that would make the next record 7-9 again.

The point of this, of course, is that a big part of the reason true talent level is so hard to figure out is because in baseball, those records were generated with the same group of players over three months...but for the Redskins, that kind of performance record takes six years to generate, and 80% of the players on the roster are going to turn over during that time. So look at the way we tend to treat Pollyannas who think we can go 12-4 a year after three consecutive losing seasons, and for that matter look at the way we can figure we've turned the corner after we do go 12-4 and win the division one year, only to slip back to 7-9 the next. Finally, think about the fact that going into the baseball season, the Nationals were pegged as having the talent of a 70 win team, and think that over the next six iterations of sixteen game blocks, they've won, on average, eight games per block rather than seven. But if you look at the players individually over their career and not just the half year played so far, they still shake out for the last two months of the season as about a .450 team. Even over a period of years, I think that a team's record can provide an inaccurate picture of their talent level. Not all the breaks even out. I think that's a huge issue in a league where the average number of games played by a player in his entire career is less than a season's worth in any other professional sport.

You can't just say, "We would have won these three games with X player that we lost with Y player." You don't get to choose which plays have better results under X player. All you can do is shift the context. Where we might be a .450 team with lousy quarterbacking, maybe we are a .480 team with competent quarterbacking, and maybe a .520 team with great quarterbacking. And at that point we go from having seasons of 5-9 wins to having seasons of 7-11 wins. And maybe that means we're 8-8 in a good year with a bad QB and then 7-9 in a bad year with a good QB. That gives us a much different view of the QB's worth than if you go 6-10 in a bad year with a bad QB followed by 10-6 in a good year with a good QB. But it's still the same QB underneath, and the tricky part is that the noise inherent in chance's role in a 16 game season ends up dwarfing the signal that comes from finding an individual player's true ability level.

[Est.1974]

I'm not offering this idea as a precision instrument. So, let's not try to cut it too fine.

How much improvement Manning would make to a true 8-8 team would depend on the grade of the QB he replaced. My estimate would be 1-2 wins.

You have to cut it fine OF, the devil is always in the detail. :evilg:

Based on the percentage, I'd say the QB influences winning three-fold compared to the other positions, and as such the number of additional wins should be increased to reflect that.

In my view, I would say 1-2 wins, as quoted above, is too conservative an estimate.

[Oldfan]

Gerald ~ This is where the other side of the discussion would hold water. The QB is a 10% upgrade in talent over what was present before, but without the meshing of the other parts, the total addition would not be as significant.

You have misread my opponents’ argument in this thread. They, like 80% of fans and media, think the QB is worth far more than 10%

As for your point about the various playing styles and schemes, the 60/40 pass/run ratio that I worked into the estimate isolates the biggest set of scheme differences. As for the rest, remember that my numbers represent averages, so some deviation is not a problem.

The fit of the QB to the scheme is an important factor as you say. However, one expects all NFL coaches to know that and to select QBs who can run their scheme. They do miss now and then, but again, we’re looking for an average of the 32 teams.

[Oldfan]

Shkspr ...it's still the same QB underneath, and the tricky part is that the noise inherent in chance's role in a 16 game season ends up dwarfing the signal that comes from finding an individual player's true ability level.

When posters debate QBs in this forum, they will often cite wins as a measure of his ability. In my view, they are offering a stat with 90% noise. When they offer the QB rating or yards-per-attempt, I hear 50% noise.

Posters will sometimes say things like, “I believe what I see, and Tom Brady is much better than Marc Bulger.” When I look at those two, I see two QBs with comparable skillsets. If they had exchanged teams, I have no doubt that Bulger would be thought of as the superior QB with HOF credentials.

When it comes to performance, we can’t believe what we see because we can’t tell whether we are seeing the QB play at 70% of his potential or 95%. His supporting cast will determine that.

---------- Post added July-16th-2011 at 06:25 AM ----------

...Based on the percentage' date=' I'd say the QB influences winning three-fold compared to the other positions, and as such the number of additional wins should be increased to reflect that...[/quote']I can't agree with that.

I think you may be underestimating the value of one win. As Shkspr said in 291, going from a true level of six wins to seven drops the odds of winning 10+ games (playoff level) down from 27-1 to 9-1. Moving from eight to nine would have a similar effect at much lower odds.

[Est.1974]

I can't agree with that.

I think you may be underestimating the value of one win. As Shkspr said in 291, going from a true level of six wins to seven drops the odds of winning 10+ games (playoff level) down from 27-1 to 9-1. Moving from eight to nine would have a similar effect at much lower odds.

I didn't think you would

I must be reading that wrong, because it reads to me that if you win that extra game, to move from 6 to 7, it then becomes more likely that you could go further and win more. Thus, the shorter odds.

Therefore, it one player is strong enough to win that one game, the probability of winning 10+ games increases.

[Shkspr]

This is weird. I feel like I'm arguing both sides. Basically, I agree with the magnitude that Oldfan suggests, that moving from a lousy to an elite quarterback is worth a fraction of a win per season to a team's fortunes, but I think what's being missed is that that's a hell of a lot. The spread between a "good" team and a "bad" one over 16 games is only about four or five wins, and responsibility for that is divvied up amongst 25 players or so. If the quarterback position is worth one eighth to one tenth of that himself, then it's a seismic change. Using the Football Outsiders post in the OP, moving from Stephon Heyer to Jonathan Ogden at LT is the difference between say, Philip Rivers and Drew Brees. We've spent what, five first or second rounders on quarterbacks in the last decade or so? Hitting the jackpot on just one of them not only would have let us spend those picks on other building blocks, but upgrading the QB position is in itself the equivalent of hitting on three or four first round picks.

So breathless hyperbole that one player will guarantee three to five wins a season is silly and shallow, but that doesn't take away from the fact that a grade A quarterback would go a long way towards rehabilitating this franchise.

[Chump Bailey]

So breathless hyperbole that one player will guarantee three to five wins a season is silly and shallow, but that doesn't take away from the fact that a grade A quarterback would go a long way towards rehabilitating this franchise.

I agree, but am of the opinion that Sam Bradford, and SB only, could realistically propel us/anyone to more than the proposed 10% win differential. I can't offer anything to support that. It's just my gut feeling that Sam Bradford is a special QB capable of playing exceptional football at his position.

[Oldfan]

I didn't think you would

I must be reading that wrong' date=' because it reads to me that if you win that extra game, to move from 6 to 7, it then becomes more likely that you could go further and win more. Thus, the shorter odds.

Therefore, it one player is strong enough to win that one game, the probability of winning 10+ games increases.[/quote']It sounds like what you need to do is get a better grip on how the luck factor plays havoc with the outcomes of 16-game seasons. With a team at the six-win quality level, the odds against getting enough good breaks to bounce up to playoff level (10+) in any given year is 27-1. Another way of saying that is that it should happen once every 27 years. Once you understand that idea, the rest should follow.

[Est.1974]

that moving from a lousy to an elite quarterback is worth a fraction of a win per season to a team's fortunes

Using the Redskins as an example, over the past 3 seasons we have lost something like 17 games by a touchdown or less.

Are we saying here that our record would not have changed that much, even with an elite QB ?

---------- Post added July-16th-2011 at 04:09 PM ----------

It sounds like what you need to do is get a better grip on how the luck factor plays havoc with the outcomes of 16-game seasons.

Luck, damn that old chestnut again.

[Oldfan]

Using the Redskins as an example' date=' over the past 3 seasons we have lost something like 17 games by a touchdown or less.

Are we saying here that our record would not have changed that much, even with an elite QB ?[/quote']We are saying that, if you had given Peyton the Magnificent or Tom Terrific the same support given Jason or Donovan, you would not now be thinking of them as elite QBs.