## Russell Wilson, and Other Quarterbacks Through 48 Starts

Through 48 starts, Russell Wilson has been pretty darn good. Since Wilson entered the league, he ranks 5th in ANY/A and tied for 2nd in wins. The idea of measuring quarterback by wins is a curious one, especially in the case of Wilson, who has benefited from a historically dominant defense. No, Wilson is not a great quarterback because his team has won 36 games over the last three years.

On the other hand, analysts can go too far in the other direction. For example, among the 31 quarterbacks to play in 24 games and throw 500 passes since 2012, Wilson ranks just 24th in passing yards per game. Some could use that statistic, or another variation, to argue that Wilson hasn’t been responsible for much of his team’s success. But that ignores that Wilson leads all quarterbacks in rushing yards since 2012, and ranks 3rd among those 31 quarterbacks in rushing yards per game. It also ignores the fact that Seattle has run the 5th fewest plays of any team over the last three years.

So how about this stat: over the last three years, Wilson has been responsible for two-thirds of all Seahawks yards. How does that compare to other quarterbacks through 48 starts? And how many wins did those quarterbacks have?

I looked at all quarterbacks who started at least 48 games since 1960. I’ve plotted those quarterbacks in the graph below, with the X-Axis representing percentage of team yards and the Y-Axis displaying wins. As you might suspect, Wilson winds up in the upper right corner:

That graph is a bit too crowded, though, with 159 quarterbacks. So let’s group things by era. First, all quarterbacks since 2000. For all graphs, I used the same X- and Y-Axes. As you’ll see, this also helps us see some ways in which the league has shifted. For example, as passing became more prevalent, that meant quarterbacks have become responsible for a larger percentage of team yards. So this graph looks skewed quite a bit to the right:

From a Pareto Efficiency standpoint, three quarterbacks stand out: Matthew Stafford, Andrew Luck, and Tom Brady (Wilson is tied with Brady for 36 wins, but is a bit lower in percentage of team yards). For those — and only those — three players, there is no quarterback with the same or more wins who was responsible for the same or a larger percentage of their team’s yards through 48 starts.

Stafford through 48 games was responsible for a larger percentage of his team’s yards than any other quarterback in history. That’s not too surprising, since we all have watched his career unfold, but it is still notable. Luck is only a hair behind Stafford, but has a whopping 14 more wins. While we often hear the Luck/Wilson debate framed as a stats vs. wins battle, it’s not as though Luck hasn’t been remarkable at winning games, too (and with less help than most young quarterbacks).

Tony Romo is to Luck in this graph as Brady is to Wilson: Romo is tied with Luck with 33 wins, but had a smaller percentage of his team’s yards. And, obviously, from a projections standpoint, that Luck accomplished what he’s done at a younger age than Romo makes it much more impressive. As for Brady, well, no one is surprised to see him at the top of the QB wins portion of this graph. He was Wilson before Wilson, reaching two Super Bowls in his first three years as a starter, and shouldered even a larger load of his team’s offense.

How about the quarterbacks from 1987 to 1999?

Five quarterbacks form the outer hull here: Jeff Blake, Mark Brunell, Brett Favre, Donovan McNabb, and Kurt Warner.

• Blake was a very interesting player that I’m sure most fans don’t give much thought to anymore. He threw one of the most gorgeous deep balls you’ll ever see, and was a shining light in a sea of 1990s Bengals passing misery. After spending time on the Jets bench, he moved to Cincinnati in ’94, where he started his first game. He led the NFL in yards per completion that year, and two years later led the NFL in game-winning drives with five. He was responsible for nearly 75% of Cincinnati’s offense during his first 48 starts, and while he had a losing record, he was great by Bengals standards.
• Brunell finished just a bit behind Blake in terms of percentage of team yards, but was much higher in the wins metric. He led some excellent Jaguars offenses, and his first 48 starts include all of his magnificent 1996 season. Brunell was a very good quarterback who snuck into Brad Oremland’s top 52.
• Favre had one more win than Brunell, and was about one percentage point behind him in percentage of team offense. McNabb was three wins ahead of Favre, and about two percentage points behind him. Warner was three wins ahead of McNabb, and about three percentage points behind him. Warner/McNabb/Favre/Brunell basically form a straight line in the graph above, and all were obviously very good quarterbacks who were successful very early in their careers (which, I’ll note, started at a variety of ages for this group).

Let’s move on to the next set of quarterbacks, whose first start came between ’76 and ’86:

Danny White leads the way here with 37 wins. His first notable playoff appearance came in relief duty in ’78, when he helped the Cowboys come back and defeat the Falcons in ’78. Then, in White’s first three seasons as a starter, he took Dallas to the NFC Championship Game each time. He had a lot of success playing on some very good teams.

On the other hand, White fell short of 60% of the Cowboys offense. If we move along the outer edge of the chart, our next Pareto Front point belongs to Dan Marino. Despite what you might have heard, he was quite a winner early in his career, beginning 33-8 and totaling 35 wins through 48 starts. He was also, unsurprisingly, a huge part of the Dolphins offense. But he still trails Randall Cunningham in this department, our third standout quarterback on this era’s outer hull. Given his rushing prowess, it’s probably not too surprising to see, but Cunningham was an enormous part of the Eagles offense (albeit one that lacked talent).

Finally, let’s look at the quarterbacks whose first start came between ’60 and ’75. Again, we have three quarterbacks forming a Pareto Front:

No one should be surprised to see Daryle Lamonica or Joe Namath there — Lamonica’s winning percentage is legendary, while Namath often was the Jets offense to an enormous degree for this era, as these numbers imply — but Jack Kemp standing out is a bit of a surprise, at least for me. When I think of Kemp, I think of the guy won a pair of titles as part of a very good Bills team, but not a star quarterback (particularly in ’64, where the Bills ranked last in pass attempts and first in rush attempts). But Kemp’s first 48 starts cover his 28 starts with the Chargers, where he was in a very different situation. He ranked 2nd in the AFL in passing yards in both ’60 and ’61, and wa very pass-happy in 1963.

Does your “percentage of team yards” include YAC? Or does it use something like YAC+ to give the QB some but not all credit for YAC?

• Nope, just percentage of raw yards. Putting aside the merit of the question, it’s not as though we have YAC data going back to 1960, anyway.

• It doesn’t look like you’re including sack yardage. I’m sort of on the fence as to whether you should. I mean, it *is* a part of team yardage, and it’s definitely a stat some QBs are better at avoiding than others. It would really knock down someone like Cunningham, to be sure.

Good point. And I’m sure it probably wouldn’t affect the scatter charts too much.

• Anthony Brown

Shouldn’t Hall of Famer Sonny Jurgensen have a dot on one of your charts?

• Jurgensen had 4 starts as a rookie in 57, before going back to the bench behind Van Brocklin in 58-59. So technically, he doesn’t belong on the chart of “QBs whose first start came between ’60 and ’75”.

• That’s exactly right.

• If you played just the sound of the Blake video and asked me to name the quarterback featured, I might never have gotten it, even after a couple hints.

• Richie

I think about Jeff Blake often. But that’s only because we had an owner in our fantasy football league last year who tried drafting Jeff Blake 3 different times. (He may or may not have had some adult beverages in his system.)

• Wolverine

Blake may not have been a great quarterback. but that Cincinnati passing game with Blake, Carl Pickens, and Darnay Scott was certainly entertaining. Makes me wonder if the offense could have been something special if Ki-Jana Carter hadn’t suffered his injury. At the very least, Jeff Blake prevented the Bengals from being completely unwatchable during the middle/late 90’s.

I see Russell Wilson as an average starter who lucked into a great situation. If he had to throw 40 passes per game and overcome a poor defense, I think his efficiency would suffer greatly. If he stays in Seattle long enough, the defense will regress and he’ll be facing this exact situation. Be curious to see how he handles it.

• sacramento gold miners

I think Wilson would still excel if the Seahawks had to throw the ball more, but I’m not sure Tony Romo or the likes of Andy Dalton would have lifted Seattle to back to back Super Bowls. Also, Wilson outplayed Manning in the New Jersey game, and how many people saw that coming. I can’t discount Wilson’s obvious ability just because his defense has been great.

Wilson is NOT better than Tony Romo.

• WR

What’s average about Wilson’s career so far? He’s won 75% of his games, including the postseason, played in 2 super bowls, and posted some of the best rate stats around. His 98.6 passer rating would be the 2nd best of all time if he had enough attempts to qualify. Luck is at 86.6. Now I’ll grant you that it’s only 3 seasons, and he has had a great defense and running game, but his career so far has been spectacular.

2014 was his best year yet. He posted career highs for completions, passing yards, and fewest interceptions. In addition to all of that, he ran for 849 yards, at 7.2 yards per attempt, and scored 6 rushing TDs without losing a fumble. He plays the position in a manner that is still considered unconventional, and doesn’t have the raw tools of a Luck or Manning. But to say that he’s just an average talent seems laughable, and is reminiscent of your Brady criticisms. There’s more than one way to play quarterback.

Now maybe I’m wrong, and Wilson will fall off a cliff in the next couple years. I don’t see that happening, unless he gets hurt. I don’t expect Wilson to outplay Luck for the rest of their careers, but so far, it’s clear to me that he’s been better than Luck by a significant amount.

I don’t care about win %. That’s a team stat, not a QB stat. It just so happens that Wilson has benefitted from the best scoring defense in the league from 2012-14, which may have aided him in winning games.

But the main reason I consider Wilson average is because he’s spent his entire career passing in optimal situations. The Hawks defense has allowed him to play with the lead, which is why he only has to throw 25 passes per game. Moreover, with the threat of Lynch, defenses typically play eight in the box, leaving single coverage on his outside receivers. That’s a huge luxury that most QB’s don’t enjoy. I want to see how Wilson performs when he has to throw 50 passes against dime coverage because the defense knows what’s coming.

QB’s who are currently better than Wilson:

Rodgers
P Manning
Luck
Brees
Rivers
Romo
Roethlisberger
Ryan
Newton
Flacco

QB’s in the same tier as Wilson:

Stafford
Kaepernick
E Manning
Bridgewater
Dalton
Palmer
Tannehill
Foles

• Wolverine

I would like to take the middle ground here:

1)I agree win% is a hugely overrated stat in evaluating quarterbacks, since the quality of your defense is at least as important as the quality of your quarterback (one example, Matt Ryan is still really good, but he no longer wins 10-13 games per year only because the Falcon’s defense sucks)

2)Adam, I think you’re not giving Wilson quite enough credit. I agree that Wilson is not as good as Rodgers,Luck,Brady,Manning,Rivers,Romo,Big Ben, and Ryan. I think he’s way better than Newton and Flacco. Yes, he’s only asked to throw much, and only in favorable situations, but he’s still incredibly efficient in those situations. I feel like people knock Troy Aikman’s career for the same thing. If he was asked to carry his team (i.e. when the defense eventually regresses), his efficiency will go down, as will the Seahawks’ win pct (slighlty, I think), but I still think he can put up pretty good numbers. He’s smart, accurate, has a nice deep ball, and can run.

3)Wilson in the same Tier as Stafford, Dalton, Kaepernick, etc? I’m a Lions fan and unabashed Stafford apologist, but even I think that’s ludicrous.

4)I think Wilson is somewhere between the 6th to 10th best QB in the league. Not quite as good as his efficiency numbers suggest, but still pretty damnn good.

• Tom

Good points Wolverine, completely agree – Wilson is probably not as great as some think, but he is, in opinion, in a tier above Dalton, Stafford, etc., even if its simply because of his running ability.

I would say Win % is massively overrated, and more to the point, it’s just misused. Brady is a “winner” in large part because of Malcom’s interception, Warner is a “winner” because of Jones’ tackle, etc. This is why Burke and others started calc’ing Win Probability…to find what portion of a team’s win can be attributed to who. Expected points is great at this as well – that stat is perfect for showing what portion of a team’s margin of victory is attributed to which unit, or player. What’s especially great about it is that it takes into account field position, which otherwise we’re basically guessing at – as in “Wilson always has great field position because of his defense”, etc.

In 2013, San Diego had MOV of +3: +10 for offense, -7 defense. Seattle had a MOV of +11.6: +5.2 offense, +6.4 defense (rough numbers from my own database). The numbers are there, we can see which unit ” carried the load”, dig a little deeper, see which player did the most, etc.

• sacramento gold miners

Not sure we can ever accurately assign specific percentages to different players, I’ll go with the most important position on the field having a significant outcome on the result. It’s true Butler saved the day for the Patriots last season, but Brady’s dissection of the Seattle defense in the fourth quarter was huge. Mike Jones made a nice tackle in that Rams SB win, but it was Warner, despite getting hit, launching the TD pass to Bruce.

At the end of the day, the better QBs usually win, and those players have made key plays long before end of games to make that happen. Russell Wilson is on a HOF path, barring injury, or off field issue. I like his future much more than Tony Romo, who needs more postseason success.

• Richie

Yes. When your main criteria for being good is winning, the better QB’s usually win.

Circular logic.

• Tom

Agreed, overall, QB’s deserve a lot of credit for winning games, and no, we can never precisely assign the proper credit to each player…and yes, over the long run, the better QB’s usually HELP win games. But my point was that we have some advanced stats now that get us closer to the truth. You’re saying you’ll go with the most important position on the field having a significant outcome on the result…yep, I agree, but I’m saying “How much”?

I get that using Win Probability is dicey, because we only have data since 1999, and there’s really only two places to get the data from, but I’m much more comfortable saying that Green Bay won 12 games last year, and Rodgers gets credit for 5 of them, as opposed to saying “Rodgers won 12 games last year”. Wilson? He gets credit for winning 3 out of the 12 wins last year for the ‘Hawks. This feels right to me.

So, I’m not disagreeing on how important the QB position is…but using Wins, as a stat by itself and not part of a larger picture (and I’m not saying that you’re doing this), for judging a QB doesn’t seem right to me when there’s 21 other players out there fighting and scratching and making plays to get that win (forget Butler’s pick, what about all those extra yards Edelman fought for, and all the hits he took?).

Ok maybe I was being a bit harsh. IMO he’s kind of in a no-man’s land between the Ryan/Flacco tier and the Stafford/Kaepernick tier. Another part of the reason I downgrade Wilson is because he creates his own pressure in the pocket by holding the ball too long and running around. His offensive line has been sub par, but his style makes them look even worse.

• flyerhawk

One thing I would like to respond to regarding Wilson only throwing in ideal situations. I think it is extremely misleading to say that. The truth of the matter is that Wilson is MORE likely to throw in adverse conditions. When the team is down by multiple scores especially. Granted that doesn’t happen very often but when it does, Wilson usually puts the offense on his shoulders.

There is a reason why he leads the league in come from behind wins since he has entered the league.

Wilson also doesn’t get any garbage time stats. If the Hawks have a 2 score lead, they take the air out of the ball.

Your assertion is flat out incorrect. In his career, Wilson has dropped back 563 times with the lead and only 462 while trailing. His rating with the lead is 104.5 and it falls to 95.5 when trailing.

• flyerhawk

Correlation is not causation.

The Seahawks have won 36 of 48 games in those three season. Of course he has more pass attempts.

The Seahawks have run 861 plays while trailing. Wilson threw the ball 429 or 49% of the time.

They have run 1398 plays with the lead. Wilson threw the ball 506 times or 36% of the time.

But those percentages are irrelevant in answering the question of how optimal his passing environment has been. The vast majority of QB’s throw more passes while trailing than leading, and trailing is generally a sub-optimal circumstance for passing. But Wilson gets to throw more passes from the optimal state of being ahead. The fact that Wilson averages only 154 dropbacks per year while trailing is very telling in and of itself. The typical full time starter drops back 300+ times while trailing.

• flyerhawk

Your claim that the vast majority of QBs throw more passes while trailing doesn’t really make sense.

For every QB that throws while behind, there is a QB that throws while ahead. Kinda unavoidable.

Yes, Wilson gets to throw with a lead a lot. Which means he wins more but he also takes far fewer risks and has fewer opportunities.

Yes, the Seahawks are a very good team. That doesn’t make Wilson a poorer QB.

“For every QB that throws while behind, there is a QB that throws while ahead. Kinda unavoidable.”
How is that unavoidable? Let’s say a game is 28-0 at halftime. In the second half, the trailing QB will probably throw on 90% of his team’s plays, while the leading QB might only throw 20% of the time. It is a fact that more passes are thrown in the NFL while trailing than while leading.
I’m not saying Wilson is a poorer QB because he’s on a great team – I’m saying we don’t know how good he really is because he’s played in more optimal circumstances than just about any QB in the league. A few years down the road that will change and we’ll see what Wilson does when he has to shoulder the full offensive load.

• flyerhawk

The frequency of throwing may change but the situation is the same.

As I said already. I’m completely fine with saying that it is too early to tell where he really stands as a QB. Based on his play, we can say he is definitely one of the better QBs in the league. Barring injury, he is likely to have a pretty good to great career.

But we will need to see more from him to determine whether he will be a truly elite QB.

• WR

It’s clear that we have different methods of evaluation. I largely agree with your list, although I’m not convinced that guys like Flacco, Newton, and Luck should rate ahead of Wilson. I also think Wilson’s better than most of the guys on the lower tier list. Let’s compare Wilson to Luck, based on what they’ve achieved so far.

Luck throws the ball far more often, but is less efficient. Luck throws for more yards and TDs, and gets sacked a bit less. Wilson has a big advantage in rushing yards and TDs, and turns the ball over far less. Wilson is ahead of Luck in all the 4 categories included in passer rating, which is why his rating is so much higher. So do you want a QB who throws the ball more, and for more total yards, and produces more TDs while turning the ball over more often? Or do you want Wilson, who throws the ball less, but much more efficiently, creates fewer turnovers, and has better scrambling and rushing ability? It’s close, but I’ll take Wilson. Now, I think Wilson has a much higher chance of getting injured in the future, because running QBs often get hurt, so I wouldn’t trade Luck for Wilson right now. But if he stays healthy, I think Wilson has a great chance to enjoy a better career than Luck. We’ll see.

I know that you don’t like wins as a stat, but to me this is like saying that you can’t evaluate pitchers by looking at their career W-L record. The QB, in his career, absolutely has a huge impact on his team’s success, and I think it’s silly to suggest otherwise. Here are the top 11 QBs by win pct, including playoffs, minimum 70 starts

Otto Graham
Staubach
Montana
Manning
Roethlisberger
Rodgers
Stabler
Steve Young
Unitas

Do you see any names that don’t belong? Maybe Stabler, but that’s it. Career win pct is absolutely a valid part of evaluating QBs.

I’ll grant you that players like Manning, Marino, and Luck have better mechanics than guys like Brady, Wilson, and Montana. But the West Coast style of shorter pass attempts has been proven by Brady and Montana to be just as effective, perhaps more so, than the downfield passing of Manning and Marino. I don’t agree with your contention that players like Brady should be penalized for a shorter average completion length, and I think you’re applying that same mentality to the Luck-Wilson question. By any measure, Brady and Montana are two of the most successful passers in NFL history. How can that be, if completion length is of such paramount importance?

I agree that win % has at least some viability over a full career, particularly if the QB plays under several different coaches / systems. But Wilson has only played three years, all under optimal circumstances for winning, so I don’t put any stock in his win %. If he’s still maintaining 75% wins over 10+ years, then I’ll feel differently about it.

There’s no denying that Wilson has better efficiency numbers than Luck. But it’s more than just the difference in attempts. Luck has played behind poor offensive lines (even worse than Wilson’s), and has dealt with mostly below average defenses and special teams. He’s never had a stud RB to ease the workload. Luck has been THE guy for the Colts since week 1 of his rookie year. So while I agree with your hypothetical about volume vs. efficiency, the circumstances for Luck and Wilson have been so divergent than I don’t think it’s fair to compare their efficiency directly. Let me ask you: How many games do the Colts win the last three years if Wilson is their QB?

You’re right I don’t like short throwing QB’s, but Wilson is not a short thrower. In 2014 he did rely on YAC more than in previous years, but overall I’d consider him a deep thrower, which he has proven to be good at. For me to put him in the upper tier, I need to see him consistently make throws of all types and distances, and do so against defenses that are focused on the pass. At some point the Hawks defense will fade and we’ll have an answer.

The list I made before was off the top of my head and very general. I do think Wilson is better than the QB’s I placed in his tier, but not so much so that I’d move him up to the next tier. If I had to rank them Wilson would be around 12th.

• flyerhawk

How do you measure the quality of the Colts line versus the Seahawks line? Most people grading the Seahawks line would rate 2 guys, who have missed significant time, as above replacement level players. The rest of their line is graded at below replacement, particularly on passing plays.

You also ignore the fact that Luck has played in the WEAKEST division in the NFL his entire career. He has played the Jaguars and Titans 6 times each. They once again have the easiest rated schedule in the league this year.

If you want to say that Wilson should get an incomplete because we don’t know enough about him, that’s fine. But you are talking out of both sides of your mouth. On the one hand you are saying we don’t know even to grade him and on the other you are grading him. Which is it?

I’d prefer to give him an incomplete grade. Honestly that’s why I’m calling him “average.” We really don’t know how he’ll perform under a more diverse and difficult set of circumstances. He could end up being a fantastic QB who carries his team, and in that case I’ll freely admit that I was wrong.

While I do think Luck is and will continue to be better than Wilson, I’m not sold on Luck automatically becoming the best QB in the league at some point, which is the way many people talk about him. You’re right about him facing weak schedules year after year, which has undoubtedy made him look slightly better than he really is.

• Tom

If I’m understanding Adam correctly, I think he is saying specifically that Wilson doesn’t have enough seasons yet under his belt to use Wins as a way of evaluating him.

• flyerhawk

Which is completely fair. But we also shouldn’t judge him negatively because he is on a very successful team.

• Tom

Yes, agreed. I think it’s more about putting his career in context, rather than judging him negatively…at least that’s how I’m approaching the whole thing. I think he’s damn good, he made some freaking great third down completions last year in the SB, among his other feats.

• Tom

I understand your main points, but I have to throw out that using Wins to evaluate a pitcher is just not the same as using wins for a QB. Not saying they’re completely different, but not nearly as equal as you might think. A pitcher has, theoretically, the possibility of winning an entire game himself by striking out every opposing batter, and then, when he comes up to bat, hits 1 home run (yes, I’m being extreme to make a point). Football is the ultimate team sport, and since guys aren’t playing both sides of the ball anymore, there’s just no such thing as a single player “winning the game” (in the overall sense, not the end-of-game Malcolm Butler sense). We can say that Roger Clemens just about won a game by himself in 1996 when he struck out 20 batters, but no QB in NFL history, except for maybe in the old days when a dude could be the field goal kicker, the running back, the linebacker, as well as the QB, has come close to winning a game in that fashion. I’m alright with Wins used (in some fashion) for pitchers, not so much for QB’s, if at all.

• WR

I just want to make clear, I’m not saying win pct for QBs is important because I think QBs should get all the credit when their team wins. I’m saying that a quarterback’s career W-L record has to be part of the evaluation of his career. Single season W-L records for pitchers are almost useless, but as a career figure, they’re very useful. The same is true for quarterbacks.

There is a very high correlation between efficiency at the QB position and winning, and Wilson is another example of this. The winningest QB of the 1950s was Otto Graham, who had the best career yards per attempt figure in NFL history. The winningest QB of the 1960s was Bart Starr, who had the highest passer rating from 1960-69. In the 70s Staubach had the best win pct, and the highest passer rating of anyone with at least 1000 attempts in the decade. In the 80s, Joe Montana had the best passer rating and best win pct. In the 90s, Steve Young had the best win pct, and the best passer rating. In the 2000s, Peyton Manning had the best passer rating, and was right behind Brady for win pct. In the 2010s, Rodgers has the best rating, and has a great W-L record.

It’s a consistent theme throughout NFL history. The QBs who win the most games are almost always at the top of the list for efficiency. Wilson has a fantastic rating of 98.6, and a .750 win pct. I don’t think that’s a coincidence. Even without a great defense and Beast Mode, if he continues to post rate stats at that level, he’ll continue to win a lot of games.

• sacramento gold miners

Bob Griese had detractors who felt it was mostly the supporting cast, and the Dolphins would be in big trouble if Griese had to air it out more. It was just a false assertion, as Griese was highly effective even when the Dolphins were more run-oriented.

In terms of trying to find a formula to address winning, it kind of feels like going down a rabbit hole. The W-L record will always be important when evaluating the career, and that does include the postseason.

• eag97a

Agree with your points with regards to short passing qbs amd schemes. There are also arguments that short passing is harder and should be given more credit. I personally think short passing and long passing is a wash requiring different skill sets. Short passing requires accuracy, anticipation and tighter windows but is a little bit easier nowadays because of rules against the defense and in my mind this raises the stock of older short passing qbs like montana and ken anderson. Long passing requires arm strength and with todays game recognition and anticipation that defenders will commit PI. For me the best passers are the ones who excel at both with Peyton being the best example. For me Russell Wilson has had a very very great start to his career but he needs more work with all of his skill sets as with any other young QB. I’m very excited to watch RW career progression.

• WR

I agree Eagle, and I find it very striking that the two most successful QBs of the last 35 years, Montana and Brady, both employ a similar style. I think that their style requires a strong ability to read defenses and avoid turnovers. That’s why other QBs can’t necessarily do it. To me, there’s a difference between being a successful passer, and showing off strong mechanics, and being a successful QB. I think both Brady and Montana are/were unusually strong in the mental aspect of the game, which helped them to overcome their limited physical tools. Wilson is still young, and is still getting better. If he stays healthy, his potential is almost unlimited.

• Patrick

Adam you think Wilson is average because he plays amazingly well, but only because he has a great running back and defense making him look great. He also has an average receiving corps, and until Graham came around his TEs weren’t all that impressive. The OL in Seattle is so shitty that Wilson has to run for his life on most plays. He is the QB who was pressured on the most plays last season. Sometimes he bails the pocket too early, but who can blame him when he is used to the pocket collapsing.
And what about Romo? His supporting cast have been even better than Wilson’s, at least on offense. His receivers are much better, TEs too (until Graham came around), he has the best OL in the league, and he had the leagues leading rusher on his team last year. Not to forget that the defense did a lot better than expected.
You say Wilson isn’t good because he’s in a great situation. Romo is in an even better situation offensively, but you think he is better than Wilson? That doesn’t seem like a fair argument. I think Wilson is great, but i also think Romo is great. Actually I think Romo is better, but there is no way Wilson is average. He’s a top 8 QB.

There are at least 10 QB’s who could’ve taken the Seahawks to the Super Bowl the last two years. See my list above.

As a 49ers fan, I’d kill for Baldwin; runs great routes, seems to make a ton of acrobatic catches

• Clint

While I think he’s a better passer than he gets credit for, he has the perfect situation in Seattle. I can’t see a world where Russell Wilson throws 40 times a game and still wins.

• roundtable

Common storyline with zero data to back it up. When you control for team strength, offensive performance is pretty constant when a team is either up 7 or down 7. Stats for the seven most average teams in 2014: Yards per play when down by 7: 5.36. When up by 7: 5.20. Turnover percentage goes up by .08% to balance out.
http://www.fieldgulls.com/seahawks-analysis/2015/7/26/9039799/seahawks-russell-wilson-earl-thomas-pete-carroll

Additionally, when you combine rushing yards and passing yards Wilson accounts for 66% of total yards on offense. More than Brees, Rivers, Eli Manning, and many others. It’s a bit weird that in an era when most people consider RBs to be a low value, fungible resource, many of those same people think Lynch is the *only* guy that makes one of the most efficient offenses in the game work. Wilson rushed for over 900 yards with an absolutely sick 8.4 yards per carry. Even when Lynch gets the carry, Wilson is usually freezing at least one defender. I love Lynch, but Wilson is massively more important to making the offense work.

• atyler2011

It is what it is, roundtable. I’d assume you are one of the 12s as I am. The same chorus keeps on “singing” and we are keeping on winning, hopefully it will end w/ a “better” note this time. Regarding to this RW’s meme of “not a factor” because he does not have the volume argument. First, who cares about volume, it is efficiency and effectiveness of the output, which in this case, win/loss record. For example, Matthew Stafford accounted the majority of his “contribution” to the team because of his volume and the team’s win/loss record, that is certainly true, but how effective is it? The guy still has a losing record as a starter, for one. The guy has a 3-31 against winning teams and 0-17 versus winning teams on the road. I guess his “contribution” hasn’t really “impact” when playing against quality opponents. Second, regarding to the strong defense and running game argument. People tend to forget prior to RW’s arrival, from 2010-2011 seasons, their structure and personnel were “similar” to the current roster w/ one exception at the qb position.

Let’s dissect into the defensive argument, in 2010 Seattle finished 25th ranked team in scoring defense and 7th in 2011, a big improvement; however, on the offensive side they didn’t improve as they ranked 23rd in scoring offense for both years. Both years 7-9 record. Now, let’s assume that team has the current team defense, what their record will be w/ the “same” defense? Let’s used the points differential as a measuring stick. From 2010-2011, Seattle had an average of -2.85 point differential, which probably meant the offensive unit was not up to par (this unit had Beast Mode and basically the same offensive personnel before RW’s
arrival) I’d argue they probably had an overall better talent on the offensive side, at least at the skill positions. This compares to 10.4 point differential from 2012-2014 w/ basically the same roster with the only exception at the qb position. Seattle had a 7-9 record for both 2010-2011 season. We can argue 2010 was PC’s first season so there was an adjustment, which is fair. But we did not see any improvement from the offensive side of the ball, both years ranked 23rd in terms of points scored per game, when the defensive unit made great stride, ranked 7th in scoring defense from 25th in 2010. Let’s assume the 2010-2011 also had the “same” defense as today unit, so the average scored against would be 15.2 for those 2 seasons resulted in a +4.55
point differential. The +4.55 translated into 4 more wins so if the 2010-2011 had the “same” defense, the overall record would be 16-16, a .500 record. The defensive point differential for the 2011 vs. today
unit is -4.5 (19.7 vs. 15.2) very close to the overall point differential when using the 15.2 average for the 2010-2011 team. On the side note, since RW’s arrival, the offense had added 6 more points to its scoring average, which is incredible, when considering most of the game outcomes determined by the spread between 3 and 7 points.

Regarding to the running game, when we talk about Seattle’s running game, it is about Beast Mode. In this context, I am not saying Beast Mode is not Beast Mode, but how Beast Mode has benefitted from having RW as their qb. They complement each other so well. For example, Beast Mode had a career average (w/o RW as the qb) of 3.9 yards per carry. He averaged 4.2 during the time in Seattle prior to RW’s arrival. After RW’s arrival, his average went up 4.8. You are talking an increase of .6 of a yard to almost a full yard, which is quite a lot for a running back w/ basically the same OL. He scored more touchdowns (due to the scheme), more receptions, more receiving tds (RW looked for his checkdowns) etc..

Personally, I don’t think RW can consistently throwing the ball 40 or 50 times per game because of his physical limitations, but that does not mean he cannot “manage” and continues to be success at what he is doing w/ his unique skill sets. This is how Seattle builts its team, strong running game, defense, limit of turnovers, and a NECESSITY for a qb who can make big plays when needed. If you look at his play from that perspective, then you would value his “contribution” even more because he does not have a lot of opportunities to do it.

One last note, people say defense wins championship, but the saying does not reflect the reality of it. For example, in the last 14 SB winners, only two number one scoring defenses had made and won it- the 2000 Ravens and 2014 Seahawks.

“The same chorus keeps on singing and WE keeping on winning”

When you refer to a team as “we” during an analytical discussion, it immediately makes you look biased and likely pushing an agenda. I don’t understand how the `Hawks defensive deficiencies in 2010 and 2011 are the least bit relevant to the evaluation of Russell Wilson, considering he was STILL IN COLLEGE. The defense made a huge jump in 2012, and would have done so whether Wilson or Flynn or Tarvaris were playing QB.

Nobody is saying that Matt Stafford is good because he throws a lot and puts up big volume stats. But it’s disingenuous not to acknowledge that Stafford’s job is harder and more demanding than Wilson’s. While Stafford has Megatron which obviously helps, he has to throw against nickel and dime coverage pretty much all the time, while Wilson enjoys single coverage and even cover 0 on occassion!

I agree with you that Lynch and Wilson make each other better, so it’s not fair to give Lynch all the credit for defenses stacking the box against the Hawks. That being said, I still believe that Lynch is more responsible for Wilson’s success than the other way around.

• atyler2011

First, did I say that I was not a Seahawks fan? So I’d think that gives away a little bit about my persective. You can question my “allegiance”, but please do not use that to say I am “biased” in my analysis. If you think that I am, then you are more than welcome to refute it by providing some of your own. Second, “Hawks defensive deficiencies in 2010 and 2011 are the least bit relevant
to the evaluation of Russell Wilson, considering he was STILL IN
COLLEGE”. You just provide more “credence” to my argument. The fact of the matter when RW was in college, the team made up, from 2010-2011, was basically the same, w/o RW at qb. It advanced further my point if the argument of having the “same” defense w/o RW, the team is still having the same success. That was the gist of the argument. By having the “same” defense as of today’s unit, they are not as good as having RW as their qb. Third, nobody has said whose job is much harder or easier, MS or RW. It is each individual qb’s performance and how each player has “contributed” to each team’s success. An organizational philosophy on how to build its team is very much dictacted how they will play on the field. Detroit org. believes in MS’s arm, not much his mind, so they build everything around him so he can “carry” the team. Based on his career so far, I don’t think they have much success, don’t you think? Regarding to Beast Mode’s argument, yes you can say that Beast Mode “contributed” more to RW rather than vice-versa. But you haven’t provided any data to make me believe in such hypothesis. Anybody can say that he or she believes in fairies, but that does not mean they exist. Or do they? One last note, MS’s career ANY/A (most correlated metric to a qb’s performance) is 5.95 compared to 6.93 for RW. That is almost a yard more per attempt.

You should really look at WPA as a percent of team wins.

• atyler2011

O.K. just a quick look and haven’t gotten a chance to analyze, but here is a thought about its analysis based on this metric for 2014 season only. Here is the list of qbs who were ahead of Andrew Luck- Matthew Stafford, Ryan Tannehill, Andy Dalton, Mark Sanchez, Drew Stanton, Colin Kaepernick, Teddy Bridgewater, Alex Smith, and Brian Hoyer (this is not include the future HOFs like Peyton Manning, Tom Brady, Drew Brees, etc..) Oh I forgot Cam Newton. So should I take this metric seriously or not? Based on this metric, do you really think these qbs are ahead of Andrew Luck? Seriously?

wilson is only responsible for 21% of his team’s wins over his career – it’s fairly low. (WPA/WINS). Rodgers is an astounding 43%. Flacco is just 17%. Seems like most QBs are in the 25% to 33% range (Kap 25, manning 31, Brady 25, etc)

Fitzpatrick is at just 3%.

Alex Smith is 7.5% since 2011, a shocking -26% in 2005 through 2011.

Seems like a good metric; I want to do a regression when I get a chance and see central tendencies.

• atyler2011

What is Andrew Luck’s career number? That would be an interesting thing to know. Are these just regular seasons only or playoffs as well? I don’t know the metric so it is not that I am knocking it, but to see all those qbs are “better” than Andrew Luck does not seem to be “credible” in its analysis.

Combined, 27.7%. He’s really made up for a lot of boneheaded first half plays with 4th quarter brilliance that sorta evens out.

His reputation is overblown because of it.

Ps I also don’t think it’s a linear relationship; good teams have good quarterbacks usually, and bad teams have bad quarterbacks usually. So if you’re on a high win team, your QB is probably fantastic. But in Wilsons case he’s below expected WPA.

• atyler2011

So what is the top-10 qbs, in terms of this metric, during the last 3 years? The NFL median? “So if you’re on a high win team, your QB is probably fantastic”. Does it apply to Drew Brees, Phillip Rivers, or Joe Flacco over the last 3 years?

• Tom

I put a list up above…I ordered it in terms of what percentage of a teams wins can we attribute to the QB. Wilson is pretty low, not because he stinks, it’s because his team is insanely good. The Legion of Boom can win some games on their own, and they have.

• atyler2011

Thanks, Tom. However, this is the reason that I am not really “buying” into this analysis if it does not factor in SOS. Here are the total SOS rating according to Pro-Football Reference (negative means “weak”, positive means the opposite) for 2012-2014.

AFC East- 4.1
AFC North- (12.8)
AFC South- (6.4)
AFC West- (.7)
NFC East- (1.7)
NFC North- (3.5)
NFC South- 6.6 (very surprise)
NFC West- 24.9

As I suspect, we can see most of the top-10 belong to the “weaker” divisions with an exception of Drew Brees, Matt Ryan, Cam Newton, and Tom Brady who have had played somewhat a “tougher” schedule, which means, but more than often, they are beating all the little sisters of the poor. The NFC West is almost 4X “tougher” than the closest division in the NFC South.

Regarding to the LOB, they are good don’t get me wrong, but this team is dictated by their DL. I don’t know if you’ve watched this team like I do (all the games), so I have an idea what this team is all about. This defense is great but they can’t win w/o RW due to how Seattle built its team. The fact that since his arrival, he has added 6 more points to the team scoring average in his first 3 years is astounding when the roster of this team was basically the same, on the offensive side of the ball, prior to his arrival when they were “pedestrian” at best. Put it this way, if you ask the Panthers or the Niners, whose team is built similar to Seattle, to trade CN or CK straight up for RW, what do you think they will do? Let’s not even compare their own individual stats, not even close.

• Tom

atyler2011 – Agreed, SOS is definitely a factor, and I imagine that his ranking (if we can even call it that) would improve on that list. Also, I admittedly have not watched every Seahawks game; probably a few during the regular season and then of course, all the playoff games and Super Bowls, so you know the team better than I. Also, there is no doubt from what you’ve posted, and others, that Wilson has improved his team immensely.

My main point is that using Win% as a stat for a QB, and Wilson in particular, doesn’t work very well, and that’s why I threw out the Win Probability numbers.

Wilson has a WPA of 9 games over the past three years (and to be fair to Wilson, he’s most likely getting dinged an entire game simply because of that one tragic pick to Butler) – about the same as M. Ryan, and better than Stafford, Big Ben, Cap, Newton, Rivers, Flacco and a bunch of other guys; he’s also averaged a QBR of around 68. So he’s really, really good, there’s no need to continue that discussion. But his TEAM is also very, very good, they won 42 games in three years, so his contribution is not as much, relative to his TEAM wins, as say Rodgers, who can account for almost half their wins. To say that “Wilson has a 0.750 Win%” almost doesn’t mean anything. Romo has a Win% of 0.604, so basically, he’s just not as much of a “winner” as Wilson? He just couldn’t get it done when it mattered, right? That’s ridiculous. And the best way we can show that is Win Probability – how much win percentage has this guy added. Romo has played well over the past three years, but it’s a TEAM game…for whatever reason, they can’t win games, but we don’t attribute that to one guy. It’s really hard to win games when your defense is giving up an average of 24.7 points per game (compared to Seattle’s 15.2 PA/G). I don’t really care who’s better, maybe Wilson is, but there is no doubt in my mind that Romo is far more important to Dallas than Wilson is to Seattle, and the numbers bear this out.

• atyler2011

Now regarding to TR’s argument, I am for one who believes TR is a very good qb (he belongs in my top-10 list). That being said, the guy hasn’t performed that well in “crunch” times with an exception of last year. It happens to the best of them (e.g., Peyton Manning). This is the whole gist of my argument, each team builds its team differently based on its philosophy. Dallas has always believed in TR’s arm and they should because he is very good. That being said, they were not that successful by going that route, so they refocus and rebuild their team to be more physical in having a strong OL and running game, the other part (defensive) has not come to fruition yet, but they are trying. It was no accident that TR had his best season because of Demarco Murray’s running. I don’t believe he will have the same level of success this year even if he has a sure future HOF TE and possibly the wide receiver in Dez and one or two OLs if they keep this up. When you build your team on the belief of a superstar, instead of a team, then of course, that “superstar” is expected to carry the team and be more “involved” into the scheme on how they play. In doing that, of course, their WPA will be higher compared to the ones that do not have to be so “involved”, but that does not mean the other player is less effective. It is just how each team is structured.

I am a person that is not into the QBR stuff. My argument with that metric is its inconsistent analysis of a player. For example, prior to last year, AL always scored high on the QBR metric, but he was just “middle of the road” when using other advanced metrics. How can that be? Is he just an outlier? When it comes to scientific research, the two most important things are reliability and validity. How can you do well in one but not the others if using the same or similar variables? I don’t know because they keep their methodologies in a “black box’ that nobody knows how they arrived at the score. I think they are making some changes this year regarding to its procedure in the weighted-average scoring for different situations. That is why I am a big believer in the ANY/A metric, you can argue its
weaknesses, but it is the highest correlated metric (besides the QBR
stuff but the sample used in the QBR analysis was very limited (don’t
remember the exact number but I believe it was between 20 and 30 qbs) w/
“faulty” assumption like only measuring the winning qbs and their QBR
scores, instead of all qbs, regardless if his team win or lose) when
using to measure an individual qb’s performance. It works because of
its reliability and validity. You can always run different input, but
the output will always be consistent. The funny thing is when comparing qbs with different advanced metrics, with an exception of ESPN’s QBR, those players are very similar in their position in his respective ranking between metrics.

Lastly, I have enjoyed our discussion and thank you for sharing your knowledge about the WPA stuff. I was aware of the metric, but never had a chance to look deeper. Thanks.

• Tom

atyler2011 – I’ve enjoyed the discussion as well…this is what a site like this is about: tossing out stats, taking different points of view, bringing up new stuff.

I know you weren’t arguing the Wins=Great argument, I was just making the comment that that’s the reason for my WP post (meaning, I didn’t bring that WP stuff in to argue that Wilson wasn’t good etc.).

One thing I have to comment on – it’s funny you compare Bradshaw to Wilson, because there’s a lot of similarities between the style of the two teams, the 2010’s Hawks and the 1970’s Steelers. Great defense, great running, and the long ball. Wilson struggled a bit in the last Super Bowl, but then he’d just throw this beautiful bomb to Matthews and all of a sudden, they’re in the game. Bradsahw same thing in ’78 and ’79…just when the other team thought they had them, TB would heave one out to Swann or Stalworth, and boom, the game is changed.

Also, yeah, WP is a black box, what can you do? Burke’s algorithm (or formula or whatever) is most likely his life’s work, and I can understand how he wouldn’t want to just release that…same with Schatz’ DVOA, etc. So I agree, I love ANY/A – it’s simple, it correlates well and we can all goof around with it.

• atyler2011

Also, from my previous post, I posited a hypothesis about this defense and how it “applied” to Seattle’s 2010-2011 season based on its points differential. Basically, it showed if the 2010-2011 team had the “same” defense as the today unit, they would have had won 4 more games that resulted in a 16-16 record, instead of 14-18; a .500 record. The 2010-2011 offensive unit was basically the same as the 2012-2014 team, but I could also argue the 2010-2011 had better talents at the skill positions, mainly the OL and receiver corps.

• atyler2011

Also, does the metric factor in the SOS? Let’s use Matthew Stafford as an example, he has a 35-42 record for his career? When you factor in he is 3-31 against winning team, which means he has done a lot of “damage” against all the little sisters of the poor clubs. So does it have that much meaningful in the context of when playing against better competition, he’s just simply “tepid” at best? We can say a lot of things about Joe Flucco, but I’d take that guy, in the playoffs or big games, over a lot of qbs in this league.

• Tom

atyler2011 – I could be wrong, but I’m almost certain the WP numbers of Brian Burke’s site don’t take into account SOS.

• atyler2011

Thanks

• atyler2011

Also, does the metric factor in the SOS? Let’s use Matthew Stafford as an example, he has a 35-42 record for his career? When you factor in he is 3-31 against winning team, which means he has done a lot of “damage” against all the little sisters of the poor clubs. So does it have that much meaningful in the context of when playing against better competition, he’s just simply “tepid” at best? We can say a lot of things about Joe Flucco, but I’d take that guy, in the playoffs or big games, over a lot of qbs in this league.

• atyler2011

Thanks. So his career is 27.7%, what is Matthew Stafford number career wise? Also, are these numbers take in the consideration of having high volume in their opportunities or per play basis? I am more interested in per play because that gives you more an indication the efficiency factor not just volume wise. Who cares if you produce a lot of output when you put in the same amount of input.

Regarding to Andrew Luck, last year was a very good year but I believe he was even “better” in his rookie season in the context of the situation like Peyton Manning transition and the team turnover (over 1/2 of the roster), and the pressure of replacing PM and be the franchise savior. On the field, the team had a NEGATIVE point differential and they were be able to win all those close games and made the playoffs. I thought that was an amazing job on his part.

• Tom

Totally agree…we remember is heroics, but those heroics wouldn’t be needed if he didn’t throw 3 picks in the first half, etc.

• Tom

Nick – I checked out the same thing, and got similar numbers.

Percentage of WPA per Team Wins, 2012-2014, QB’s with over 40 games, including playoffs:

A. Rodgers: 15.2 WPA/31 wins = 49%
T. Romo: 11.8/29 = 41%
D. Brees: 10.1/26 = 39%
M. Ryan: 9.3/24 = 39%
P. Manning: 14.2/40 = 35%
M. Stafford: 7.6/22 = 34%
B. Roethlisberger: 8.5/26 = 33%
C. Newton: 8.2/26 = 31%
P. Rivers: 7.3/26 = 28%
A. Luck: 10/36 = 28%
C. Kaepernick: 7.8/29 = 27%
J. Flacco: 8.5/33 = 26%
R. Tannehill: 5.3/23 = 23%
R. Wilson: 9/42 = 22%
A. Smith: 3.4/25 = 14%
A. Dalton: 2.9/31 = 9%
J. Cutler: 1.8/20 = 9%

Out of 18 QB’s, Wilson is 15th on the list. No, this is not saying that he’s bad or that Tannehill is better, etc. (for the record, Wilson has a better WPA/G stat than Tannehill, 0.16 to 0.11). What it says is that Wilson is a very good QB on an incredibly, historically great TEAM. To say things like “Wilson has a .750 win pct” as WR does below, basically attributes a TEAM stat to a single guy. I realize no one is saying that the QB is responsible for the all the wins, but when you attach the number to him, that’s basically what you’re doing. And it’s especially not fitting for Wilson. He does not have a 0.750 win%…as in Bradshaw did not have a 0.714 win% in 1976, and McMahon did not go 18-1 with Bears in ’85. These guys aren’t boxers…

• Bruce

The problem with the WPA analysis of Russell Wilson is that he is very instrumental in the run option and once he hands the ball off, he’s no longer a part of that stat, even though the threat of Wilson running, unique to only a handful of QBs, really has helped propel Seattle’s running game to another level in the past, and with Thomas Rawls coming back, and Wilson healthy again, we’re starting to see it again now. WPA doesn’t account for that (if I’m not mistaken.) I wonder how that would affect Wilson’s “9”…he’s often the 2nd leading rusher on the team, and that is ALWAYS playing havoc with defenses. How can you measure that? How can WPA measure that? To some degree the new generation of smart running, scrambling Wilson type QBs will challenge our classic thinking – and they should. That said, Wilson has proven himself a really good pocket passer (when he gets protection certainly), and a good to great leader of the team.

• Tom

QB Years Games WPA
12-A.Rodgers 3 46 15
18-P.Manning 3 53 14
9-T.Romo 3 48 12
9-D.Brees 3 50 10
12-A.Luck 3 54 10
2-M.Ryan 3 50 9
3-R.Wilson 3 55 9
5-J.Flacco 3 54 9
7-B.Roethlisberger 3 46 9
1-C.Newton 3 49 8
7-C.Kaepernick 3 51 8
9-M.Stafford 3 49 8
17-P.Rivers 3 50 7
17-R.Tannehill 3 48 5
11-A.Smith 3 41 3
14-A.Dalton 3 50 3
6-J.Cutler 3 41 2