Matthew Stafford: Should He Start or Sit?
by Goodbye Ladies on May 24, 2009 9:16 AM EDT
Email
PrintThe popular belief amongst fans and national media is that the Detroit Lions would be best served to sit shiny new quarterback Matthew Stafford on the bench for his rookie year. For instance, Sports Illustrated's Don Banks has advocated for giving Stafford a "redshirt" year, and Drew Sharp, ever the prophet, has vaguely counseled the Lions to exercise "patience" in rushing Stafford into the line-up.
I have always been skeptical of the idea that whether a quarterback starts as a rookie has much bearing as to whether he succeeds or not. The reasons why quarterbacks succeed or fail is sort of a "chicken or the egg" problem. Did Joey Harrington fail because he was on the Detroit Lions, or was he on the Detroit Lions because he was Joey Harrington (in other words, did Joey Harrington fail because he was in a "bad situation," or did the Detroit Lions draft Joey Harrington because they are poor judges of talent). I have been firmly in the camp that professional quarterbacks were "born and not made," but I understand why others think differently.
One of the key tenets of the "quarterbacks are made" theory is that teams are best served to sit quarterbacks for a year rather than bruise their fragile psyches by giving them too much playing time too early. Although I recall that the excellent blog over at Pro Football Reference did a similar study, I couldn't find it, so I was curious enough to try it on my own. I took all of the quarterbacks drafted in the first round since 1998 and measured the strength of the correlation between their pass attempts as a rookie and their ultimate success in the NFL. This, I think is the best measure, because the more a quarterback has to drop back to pass as a rookie the more chances, in theory, that that quarterback has to have his confidence shattered by an ill-tempered defensive lineman. "As a measure of "success" I used the same DYAR/pass metric that I used when I tried to evaluate Matthew Stafford's chances to become a quality NFL starter. Here is a list of all of the quarterbacks drafted in the first round since 1998 sorted by the number of pass attempts that they had as a rookie:
| QUARTERBACK Peyton Manning David Carr Matt Ryan Kerry Collins Joey Harrington Joe Flacco Byron Leftwich Tim Couch Matt Leinart Vince Young Ben Roethlisberger Ryan Leaf Cade McNown Patrick Ramsey Kyle Boller Donovan McNabb Eli Manning Alex Smith Akili Smith Jay Cutler Michael Vick Steve McNair Rex Grossman JaMarcus Russell Aaron Rodgers Philip Rivers J.P. Losman Chad Pennington Carson Palmer Jason Campbell Daunte Culpepper |
ROOKIE ATT. 575 444 434 433 429 428 418 399 377 357 295 245 235 227 224 216 197 165 153 137 113 80 72 66 16 8 5 5 0 0 0 |
Sure, David Carr is number two on the list, but Peyton Manning is number one and Matt Ryan is number three. Despite the heavy rookie workload, Peyton Manning has become one of the greatest quarterbacks of all time, and Matt Ryan, based on his rookie numbers, projects to be a perennial pro bowler. While Carson Palmer sat for his entire rookie season and became a very good NFL quarterback, except for five passes, J.P. Losman sat out his entire rookie year and has been one of the worst quarterbacks ever drafted in the first round. Alex Smith, Akili Smith, Rex Grossman, and JaMarcus Russell, all had comparatively little playing time as rookies and have not performed well afterwards.
The correlation between rookie pass attempts and success is exceedingly low with an R-squared of 0.04%, which means essentially that a 0.04% of why a quarterback succeeds can be explained by how much playing time they receive as a rookie. It's not statistically significant, and actually runs opposite to the conventional wisdom: quarterbacks who received lots of playing time as rookies actually performed slightly better over their careers than those who did not. Basically, the conclusion that we can draw is that whether a quarterback plays early or not has absolutely no bearing on whether they will ultimately succeed or fail.
So should Stafford start or sit this year? I would suggest that there is actually an advantage to getting Stafford in the line-up sooner rather than later. Although there is no significant correlation between rookie pass attempts and success there is a fairly strong correlation between rookie quarterback performance and ultimate NFL success. Although it is true that most rookie quarterbacks perform poorly, the quarterbacks who turn out to be truly awful perform much worse than those who become good to great quarterbacks. What follows is a list of all of the quarterbacks drafted in the first round since 1998 sorted by their rookie performance (minimum 50 pass attempts). Again, success is measured in DYAR/pass, DYAR being an advanced stat from Football Outsiders which (among other things) adjusts for the strength of competition:
| QUARTERBACK Ben Roethlisberger Steve McNair Peyton Manning Byron Leftwich Matt Leinart Vince Young Patrick Ramsey Cade McNown Jay Cutler Joey Harrington Kerry Collins Rex Grossman Eli Manning Kyle Boller Michael Vick Tim Couch JaMarcus Russell David Carr Akili Smith Ryan Leaf Donovan McNabb Alex Smith |
DYAR/Pass in Year 1 2.81 1.31 0.95 0.73 0.61 0.26 0.084 0.024 -0.10 -0.61 -0.64 -0.67 -0.93 -1.17 -1.42 -1.46 -1.92 -2.07 -2.46 -2.54 -2.82 -4.49 |
Admittedly, this is a small sample size, but the relationship is strong. Everyone at the bottom of the list turned out to be poor quarterbacks except for Donovan McNabb--and he had a huge bounceback season in his second year. If Stafford gets significant playing time this season, the Detroit Lions will have two full season of tape on Matthew Stafford, and if he is a bust, the Detroit Lions can move on.
This is intuitive to me. College football quarterbacks who perform at a high enough level to become high NFL draft pick are not delicate snowflakes: they are tough guys. If you're Peyton Manning, you're going to eventually be Peyton Manning whenever you start--irrespective of whether you throw 36 interceptions and are sacked 60 times. However, if you're J.P. Losman, it doesn't matter if you have two years to learn and sit on the bench--when you are inserted into the line-up, you are going to fail.
The Detroit Lions almost made the right play with how they treated the "development" of Joey Harrington. Although they correctly inserted Joey Harrington into the line-up early, they refused to move on when he failed. The Detroit Lions, determined to throw good money after bad, refused to acknowledge their mistake. This failure came with a huge opportunity cost. If they had recognized that they had failed with Harrington, they were in a great position to draft Ben Roethlisberger in the 2004 NFL Draft. Similarly, if Stafford plays poorly in his first two years, the Detroit Lions shouldn't hesitate to pull the trigger on a quarterback like Jevan Snead or Jimmy Clausen in the 2011 NFL Draft.
So put Stafford into the line-up, and if he succeeds, great. And if he doesn't, at least the Detroit Lions need not again pass up on a two-time Super Bowl quarterback in a future NFL Draft.
Cross-posted at the "Goodbye, Ladies" Draft Report
This is a FanPost and does not necessarily reflect the views of Pride of Detroit or its writers. FanPosts are valued expressions of opinion by passionate and knowledgeable fans.
89 comments
|
4 recs |
Do you like this story?
Comments
I'm lost
good post until you started to bring up algebra…j/k
If Stafford is the best QB during camp, then start him. If not, sit him. Pretty simple
I agree with Jazzy
If he is ready and Schwartz thinks he is, put him in. If not, let him sit for a while.
good homework,
I think your models could use a little work (I love statistics too, took three years worth in college- almost became an actuary) but the point illustrated does bear some consideration. I think that Eli is a different kind of QB than his brother, playing more from the heart than the head, and this model doesn’t reflect that. Nor does it consider Vick’s ability to win games, even though he will NEVER be considered a great QB. I DO believe, however, that your basic argument is very valid: a great QB is going to be great regardless of his “handlers” . The point may be moot anyway, if Culpepper can’t make it through a season. I, for one, am ready for our new era to begin; I simply am convinced that Dauntless Daunte is poised to have a CAREER SEASON and that Stafford will probably have to wait his turn.
P.S. Terry Bradshaw started as a rookie, too, and stank up the place for a couple of years b4 becoming the icon that he is. How do his stats fit into your models?
Hey MyPride,
Unfortunately (see my response to KDawg below) my “models” for projecting from rookie performance aren’t ready to be rolled out except to make the general observation that knowing how well a quarterback plays is useful information. Another problem is that Bradshaw pre-dates Football Outsiders’ DYAR statistic, so there would be nothing to plug into the equation. However, one thing I will say about Bradshaw’s rookie performance is that, although his numbers were generally abyssmal, he had a pretty good yards/attempt for a rookie, which I have found (anecdotally, not statisically), to be the best conventional statistic for evaluating quarterback performance.
by Goodbye Ladies on May 24, 2009 5:51 PM EDT up reply actions
+1
didn’t see Montana on there either, I don’t remember the 49er’s being good right off the bat neither. I do think you have a valid point though on if your going to be good your going to be good no matter what,but some of those guys that were on your list also played for some horrible teams.
Statistical analysis
What did you use for margin of error? What correlation formula are you using? Are you attempting to determine that there is a linear correlation between rookie pass attempts and DYAR? How did you calculate the linear coefficient of correlation? Characteristics of the linear coefficient of correlation include the facts that it is not based on any specific division of measurement, and the result of calculating the linear coefficient of correlation produces a number between negative one and positive one. As the linear coefficient of correlation moves towards negative one, the negative linear association between two sets of data becomes more apparent and sturdy, whereas when the linear coefficient of correlation moves towards positive one, the more apparent and sturdy the positive linear association between two sets of data becomes. If the linear coefficient of correlation is zero, then the linear association between two sets of data is said to be nonlinear. As such, as the coefficient of correlation moves towards zero, the linear association between two sets of data becomes less sturdy and apparent.
On what basis are you forming the likelihood of probabilities that rookie QB’s will have similar outcomes if they have equal attempts? Did you use the relative frequency approximation of probability, or the classical approach to probability? The relative frequency approximation of probability can be used to determine the probability of any event, which is the compilation of effects or conclusions obtained by performing a specific course of action, where the amount of occurrences of any specific event within a given course of action is divided by the amount of times the given course of action was replicated. Using this method will achieve a value of probability that is approximate, rather than precise, and the more frequently the procedure is replicated, the more accurate the value of probability will become. The classical approach to probability, which necessitates uniformly possible results, is a much more accurate method. This method determines the probability of any given event, which has a uniformly possible result, by dividing the amount of manners by which the given event can happen by the amount of singular uncomplicated occurrences. These two methods differ in that the classical approach determines a precise result, and necessitates uniformly possible results, whereas the relative frequency approximation method determines an approximate result that does not necessitate uniformly possible results. Since it would be very difficult to assume that there would be uniformly possible results for NFL rookie QB’s, it pretty much necessitates that you use the relative frequency approximation method, which is not absolutely accurate.
I would like to see your excel spreadsheets with these sets of data and the subsequent scatterplot that shows the linear or non-linear correlation. I find your analysis to be interesting. You stated that the r-squared number was 0.04. To me, this means that there is not a strong correlation between rookie attempts and DYAR. I would also like to see a hypothesis test against the statement that quarterbacks who received lots of playing time as rookies actually performed slightly better over their careers than those who did not". If you based the decision of whether or not to play Stafford solely upon the statistical correlation that you have drawn here, then the obvious answer would be no. I find it interesting that you did not provide data to support the statement that “there is a fairly strong correlation between rookie quarterback performance and ultimate NFL success”. I could go on and on…..I must say that in this case statistics is not the be all and end all of making this type of decision. A definite element of intuition must come into play, and there are many other factors that will determine whether or not a rookie QB is successful in this league. Great post.
Whoah KDawg
You a statistics professor or something?
I had a harder time reading your post than his!
Spreadsheets
Hey KDawg,
Thanks for the good word. Most of my knowledge of statistics is self-taught, so I will admit that your post did send me to Wikipedia a couple of times. As far as methodology, I used a plain, old, ordinary least squares linear regression analysis, which is how I calculated the R squared. I tried (somewhat unsuccessfully) to pare down some of the statistical jargon in order that people without a significant background in statistics could understand it.
Here are links to two of my spreadsheets (although the second one doesn’t go through very well):
http://spreadsheets.google.com/pub?key=r7pk8CzRYbYGwfSPfByzuWg&output=htmlhttp://spreadsheets.google.com/pub?key=rQWOKvuWLrwEzUpFW-XzJPw&output=html
The scatterplot doesn’t show up in these, but it is basically a straight line, which was enough to convince that there wasn’t a “fit” that would suggest that rookie pass attempts are meaningful. As far as my statement that “quarterbacks who received lots of playing time as rookies actually performed slightly better over their careers than those who did not,” it is based solely on the observation that the R squared for rookie pass attempts is positive (albeit small). One bias that is actually built into the study is that the measure of success includes their rookie year. Because quarterbacks almost always play more poorly as rookies than they do later in their careers, the R squared for those who play early would probably be slightly higher if I measured success solely by post-rookie performance.
As far as the last statement about there being a correlation between rookie success and ultimate success, I probably should have phrased that differently. I ran a very “quick and dirty” regression on success versus rookie year and I got a high R squared, of nearly 50%. However, I didn’t feel confident in the R squared number because my success measure contains rookie performance as well, which I knew was inflating the R squared. I felt pretty confident though with an R-squared that strong that even if I factored out the rookie performance there was still an R-squared there, and to be honest I was simply too lazy to go back and re-calculate everybody’s success rate. I think it’s pretty obvious from just eyeballing the chart that with players like Roethlisberger, McNair, and Manning on the top and a bottom four with Akili Smith, Alex Smith, and Ryan Leaf is going to yield some sort of correlation. I will probably be more motivated to do the legwork after this season when we have rookie numbers from Stafford (and more DYAR from current QB’s) to see if we can make any predictions about how he turns out. Anyway, it was intended to be more of a side point that a player’s rookie performance is useful information, which I don’t think you necessarily need a regression to support.
I agree that this is not a be all end all of making this type of decision, but I would be inclined to put Stafford in sooner rather than later. Let’s not do the Drew Stanton thing again!
by Goodbye Ladies on May 24, 2009 5:36 PM EDT up reply actions
I'll take a look at these tomorrow
And Nate, I am not a stats prof, but I play one on TV and I stayed at a Holiday Inn last night…..lol….j/k
I got a 4.0 in college stats.
The spreadsheets are interesting
While I believe that you made a good attempt at utilizing some data that would allow you to draw a correlation that would significantly support some sort of prediction, one way or the other, I do not feel that the data that you used helps to support the idea of starting Stafford sooner than later…..in fact I think it better supports the idea that he should sit for a year or two. The reason I say this is because there is not a strong correlation between rookie attempts and QB success in the future. This is also not to mention that we have seen over and over what happens to rookie QB’s in Detroit when they are thrust into the mix prematurely. If Stafford were the #1 pick for the Falcons last year (like Matt Ryan), then I would agree that getting him in there sooner than later would be the best idea. I also feel that Atlanta has a better O-line than we do, which makes a big difference in itself in my opinion. Even if the data you used in this study showed a strong positive correlation, I am not too sure that the correlation would support starting Stafford in Detroit sooner than later. Unfortunately, there are too many other negative factors to consider with our current team to say that it would be a good idea to start him this year. I am not saying that he should not start if he is the best QB on our team beyond a reasonable doubt, because I feel that if he proves to be better than Culpepper that he should start. However, unless he proves that he will do a better job than Culpepper immediately, I believe that he should not be the starter and I think it would suit him much better to sit and learn behind the experienced professional. He needs to work on his accuracy, as he has been a bit erratic in practice. They are calling it rookie jitters, but I think he is simply overthrowing the ball and trying to throw it too hard. He needs to relax and work on his touch in my opinion. He has a cannon, there is no doubt about that. He can put balls into places that other QB’s could never dream of doing, but if he can’t make the touch throws too then he is not going to be any better than the young Brett Favre was, prior to becoming the starter in Green Bay. I feel that having a couple of seasons to develop his accuracy, learning to make all of the throws consistently, will suit him much better in the long run than tossing the iron into the fire right now. JaMarcus Russel, Matt Leinart, Aaron Rodgers, and Vince Young were all outstanding college QB’s…..and we have seen how well Russel, Leinart, and Young did when they were made the starter as rookies. Rodgers is the exception to the rule…..why? He sat behind one of the best to ever step on the field, and he learned to make the throws and how to read the defenses by watching #4 go to work. I do not feel that Matt Stafford is a Matt Ryan. I do not feel that he is a John Elway or a Dan Marino. I think he is an Aaron Rodgers, and I think that he fits that mold much better than the Matt Ryan or John Elway mold, both whom were ready to play right away.
Hey KDawg,
I may be missing something, but reading your reply, I think you may be missing the point a bit. My point is not that Stafford will be BETTER if he starts early, but that IT DOESN’T MATTER if he starts. Accordingly, it makes sense to throw him in there sooner or later because his rooke year performance is “useful information” (although you can quibble with how useful it is) in evaluating his ultimate success or failure as a QB.
by Goodbye Ladies on May 29, 2009 6:46 PM EDT up reply actions
I got your point
I’m trying to support why I feel that he should not start right away, unless he proves to be the better man without a doubt. I feel that it does matter whether or not he starts, as your data clearly shows that there is no correlation between rookie attempts and future success, and also our line may not be ready to protect him as well as it should. Because of these facts alone, unless Matt Stafford proves beyond a reasonable doubt that he is a better QB than Daunte Culpepper, and that he has a better grasp on the playbook than DC, I think it is best for him (and the team as a whole) to sit and learn behind the experienced QB until we have another year to solidify his protection. I am not saying that I would be upset if Stafford started over Daunte, as long as he is the better man for the job and we feel like we can protect him. What sense would it make to invest in the future of the team, only to throw that future away by not protecting the investment? I agree that he could very well be successful as a rookie starter, but his individual future success has no bearing on whether or not he starts immediately or later. If he is not the better man for the job, then why start him right away? I do believe that you are misinterpreting what I am saying as well. Rookie year performance does not have to be evaluated in actual games. Stafford can be evaluated in practice, where he has a red jersey on and nobody can touch him when the O-line breaks down. If your point is that it does not matter if he starts or not, then it is contradictory to say that rookie year performance information (assuming you mean as a starter) is useful in evaluating the individual’s ultimate success or failure as a QB. Your data clearly shows that a rookie QB’s attempts (as a rookie) has no bearing on their potential future success in the league. Therefore it makes a big difference to me whether or not he starts this year, since I feel that he should sit unless he is the best QB we have, unless he understands the playbook better than DC, and unless we feel that he will be well protected (I am saying “we” but I mean the coaches…..I just say we because I feel like a part of the team even though I am only a fan).
thanks for the clarification
OK, now I understand where you’re coming from. Maybe I spoke a little too broadly initially. I think what looking at the data suggests is maybe not that “it doesn’t matter whether he starts” but that starting a rookie QB is not an intrisically good or bad decision vis-a-vis that individual QB’s development. If it really was an intrisically bad decision, you would expect poor performance to flow from rookie playing time, but that really doesn’t seem to be the case. Of course, you could look at the data, I think, as you have, and say well just because a a general proposition throwing in a QB is not bad, it may be bad in an individual case (and the inverse could be true: in some instances, SITTING a QB might be bad for his development). My hunch is that is putting too fine a point on it, but I can understand that perspective.
When I said “it doesn’t matter if he starts,” what I meant was that I don’t believe it matters as to his ultimate success as a QB. It, however, does matter as far as evaluating Stafford. Practice is great, but I would argue that there is no substitute for actual game experience in judging the quality of the player. It’s why scouts watch game film instead of just watching the combine and positional drills.
by Goodbye Ladies on May 29, 2009 11:50 PM EDT up reply actions
RUFF RUFF
RED RUM RED RUM . . . . . .SICK EM BOY! Butt could you use littler words next time, My mind had to look at that 4 times to figure it out. The funny thing is, as I started to read it before I got to he end I knew it was you.
lol
Sorry buddy…..the ol brain just took over. I’ll try to explain in layman’s terms. The term DYAR (Defense-adjusted Yards Above Replacement) is a complicated new statistic that NFL teams use to evaluate or measure a player’s total contribution over the course of every play of the season. It typically is used to show how valuable a player is versus the player that they replaced…..IE if Stafford replaced Culpepper, his stats could be used to show (in a cumulative, or in other words all added together, fashion) how many more, or less, yards that player adds to the team as a replacement. This stat is typically useless unless it is used over time, and it does not compare the player to any given average, but rather the player is compared to the level that he is expected to play at as a replacement player. The DYAR stat is also dependent upon (or affected by) the rest of the team around the replacement player. DYAR can not show how valuable a player is specifically. As such, Matt Stafford’s DYAR, if he were to start as a rookie, could prove to be 20% for example, but that would not mean that he added 20% more yards than DC did on an individual basis, but actually it would mean that Matt Stafford, playing in Detroit, in the Lion’s offensive scheme, behind the Lion’s O-line, had a DYAR of 20%, which would be phenomenal considering the fact that this stat is limited by the performance of all of the other 10 players that surround him in the offense. In the current situation, it would most likely be much less than 20%, with 0% being average, negative numbers being below average, and positive numbers being above average. However, DYAR is relatively insignificant when used for a first year player, since it is much better used over time. When trying to compare the number of attempts that a rookie QB makes in his first year (if he starts immediately) against his DYAR (which would measure his performance in terms of how many more, or less, yards he added to the team versus the player that he replaced) Goodbye Ladies was trying to see if there was a correlation, or in other words a connection, between the number of attempts that a rookie QB makes as a rookie starter and his statistical performance as measured in yards added to the team. His data showed that there is no connection, and the number of attempts that a rookie QB makes as a starter has no bearing on how well they will perform in the future. That tells me that whether he starts this year, next year, or the year after that will not have an effect on how well he will do at that time or in the future. How well he practices, the development of his accuracy and arsenal, his work ethic, and his understanding of the offense will have an effect on his performance. That is why I say that he has to prove that he is the better QB, that he knows the offense at least as good as Culpepper, and that the coaches feel that the O-line can adequately protect him before he becomes the starter for the Lions. I hope that this makes my point clear for everyone.
I think you're taking his analysis and outcome a "wee" bit too far, KDawg
This isn’t a study on the poster’s abilities in Statistics, but just a simple model that shows – using one method or control, in this case the DYAR – that “these” stats back up the claim that success is not necessarily contingent on whether or not a rookie plays in his first year and if he does, how well future success is translated based on a very usable and functional stat measurement (again, DYAR).
I’m a finance analyst and if I wanted to confuse people and attempt to show how smart I am, I could. We all have our strengths. I don’t think that’s what Goodbye Ladies was trying to do… but I can’t say that I feel the same for your rebuttals. Looks like belittling and self-inflation to me.
By the way… I aced Stats too (many years ago) and actually agree with some of your assessments (the pieces I remember). But still, all this jargon and rambling wasn’t really necessary.
Thanks for your opinion
But I was not trying to prove my intelligence. I don’t have to prove anything. I do not appreciate your insinuation that I am attempting to bolster my ego or to belittle anyone. Why don’t you go back and read again Drew…..your interpretation of what I said seems to be off kilter. I was stating my OPINIONS…..nothing more, nothing less. Sorry that I backed my opinion with intelligence. What is not necessary, if you ask me, is you jumping in and bashing me when the conversation took place 4 days ago. Not only was I trying to support my opinions with intelligent reasoning, but also I was looking to satisfy my own curiosity of how the test was conducted. At no point did I say that the original poster was absolutely wrong, nor did I say he was stupid or anything else derogatory. I asked questions. They were honest questions. I looked at the data. I looked at the spreadsheets. I formed my own opinions. I stated them. I backed them up with logical reasoning.
In the original post of this thread, the poster clearly stated
I took all of the quarterbacks drafted in the first round since 1998 and measured the strength of the correlation between their pass attempts as a rookie and their ultimate success in the NFL. This, I think is the best measure, because the more a quarterback has to drop back to pass as a rookie the more chances, in theory, that that quarterback has to have his confidence shattered by an ill-tempered defensive lineman. "As a measure of “success” I used the same DYAR/pass metric that I used when I tried to evaluate Matthew Stafford’s chances to become a quality NFL starter.
He later stated
The correlation between rookie pass attempts and success is exceedingly low with an R-squared of 0.04%, which means essentially that a 0.04% of why a quarterback succeeds can be explained by how much playing time they receive as a rookie. It’s not statistically significant, and actually runs opposite to the conventional wisdom: quarterbacks who received lots of playing time as rookies actually performed slightly better over their careers than those who did not. Basically, the conclusion that we can draw is that whether a quarterback plays early or not has absolutely no bearing on whether they will ultimately succeed or fail.
I never once disagreed with any of this. My argument was against the statement that Stafford should start sooner than later. Later in the post, this was said
I would suggest that there is actually an advantage to getting Stafford in the line-up sooner rather than later. Although there is no significant correlation between rookie pass attempts and success there is a fairly strong correlation between rookie quarterback performance and ultimate NFL success.
I used statistical jargon, because it absolutely does matter what methods are used as to the results that are obtained. I felt that it was significant to the interpretation of the data, and so I asked questions. I also felt it was necessary to define the things that I was asking about, as not everyone has taken college statistics courses. I also made my point that statistics can not be 100% of the reasoning behind a decision, as other factors and mitigating circumstances must be considered. I argued against the statement that Stafford should start sooner than later, because I believe the opposite. Again, that is my opinion and I have the right to state it and back it up with logical reasoning. In fact, one of my posts began with
While I believe that you made a good attempt at utilizing some data that would allow you to draw a correlation that would significantly support some sort of prediction, one way or the other, I do not feel that the data that you used helps to support the idea of starting Stafford sooner than later…..in fact I think it better supports the idea that he should sit for a year or two.A small amount of confusion came from the posts that I made, but in the end, the original poster understood my angle…..something that I think you have failed to do (See the posts titled “I got your point” and “Thanks for the clarification”)
Your statement of “This isn’t a study on the poster’s abilities in Statistics, but just a simple model that shows – using one method or control, in this case the DYAR – that "these" stats back up the claim that success is not necessarily contingent on whether or not a rookie plays in his first year and if he does, how well future success is translated based on a very usable and functional stat measurement (again, DYAR)” is sematics. You reiterated what the original poster already said and what I already agreed with. Your statements of “I’m a finance analyst and if I wanted to confuse people and attempt to show how smart I am, I could. We all have our strengths. I don’t think that’s what Goodbye Ladies was trying to do… but I can’t say that I feel the same for your rebuttals. Looks like belittling and self-inflation to me” are hypocritical, in that here you are saying that I am bolstering my own ego as a braggart and belittling Goodbye Ladies, yet you utilize self-inflation and belittling against me at the same time…..how ironic.
Never before have I been so insulted in my entire life Drew. Congratulations. I don’t know if you had a bad day or if someone pissed in your Wheaties, but why don’t you try attacking my argument instead of me, thereby practicing what you preach. Goodbye Ladies, if at any point you felt belittled by what I was saying, I sincerely apologize. Never did I intend to “pull rank” or make you feel stupid. I was genuinely interested in what you were saying, and I was adding to it and drawing off of it. I am supremely humbled…..maybe I should post like a burger flipper from now on.
In retrospect
I want to make something clear to everyone here…..I think of all of you as equals (except RGB). Not just Drew, Sean, and people who are good at writing……ALL of you. I don’t care if I am smarter than you or if you are smarter than me. Unless you treat me as a lesser person or insist that you are right when it is plain as day that you are wrong, NEVER will I belittle you or force my opinion on you. I will state my opinions, and they may be worded strongly and adamantly, but please stop reading into them like I am attacking you personally or mounting an offensive against your beliefs. Generally, I am reading what you write and responding with my own opinions, backing those opinions when necessary. Drew….I still have respect for you. I still think of you as an equal. I do not want a war of words (even if I think I would win….lol…..I’m kidding) or an enemy. As I stated once before, war is never won by either side. I have a tendency to let my fingers fly when I am posting….I type what I am thinking. I apologize if my “jargon” is overwhelming. I like to use what I have learned. I was insulted, but I am over it. I am a Lions fan, true and blue. I would donate my left testi for them to win a SB….lol. I am amongst Lions fans…..and God bless you all. All of the stats, all of the money, the coaches, and the schemes are nothing without us, and none of those things make a damn bit of difference when it comes right down to winning or losing on the football field. The best 22 men on the team must go head to head with the best 22 men from the other team, and only half of them will prevail. Blood, sweat, pain, pride, and determination are what matters. Even skill, technique, and athleticism are secondary…..as a wise man once told me, “It all comes down to the little things, and who wants it more”. A fundamentally sound team, with heart and desire, will beat a team with more skill and athleticism 9 times out of 10 (opinion, but based on my own experiences too). My class DD HS football team beat a class A football team in my Senior season. They were bigger, faster, stronger, and had more skills. How did we win? We wanted it more, and we were fundamentally sound. They came into OUR HOUSE and we stopped them cold in their tracks. It’s never the size of the dog in the fight. It is the size of the fight in the dog. I do not disagree that Matt Stafford should and will start when he is the best QB on the roster and when he is ready. I do not see it happening this year. He might play some, but Daunte Culpepper is going to lead this team (opinion). What does Matt Stafford have to lose? What does Daunte Culpepper have to lose? Who is going to want it more? Those are the questions, to which the answers will determine which of the two will be starting (in my opinion). Again, the best man should start in every position…..not just at QB. And that is all I have to say about that.
I got down to the last line
and all I could think about was reading it like Forrest Gump. I agree with you. I think Culpepper will most likely be the starter for most of the season. Schwartz made it clear that Stafford will start when he’s ready. My guess is that that means when Schwartz thinks he’s ready. No set of statistics will ever predict the future. To be honest, I found the original post a little to excessive. I love to analyze things inside and out and I consider myself to be a good numbers guy, but when it comes to football, it’s something that I like to sit back and enjoy and have a good conversation. I love comparing stats but this post is a little too far beyond what I care to get into as far as details go. That said, I’m glad I can come to this site and share my opinion and comment on the opinions of others. I don’t always share the same viewpoint as others and I’m sure none of us ever do 100% of the time. As far as Stafford starting, my gut tells me maybe around Thanksgiving. My head tells me he should wait until next year. My heart tells me the Lions will throw him to the wolves about week 3. Whenever he starts, for the sake of the Lions, I hope he’ll be successful.
Emotional post their KDawg
I have never really thought that you’ve ever tried to belittle people with your opinion nor do I think you believe yourself to be better or smarter than anyone here. I’ve always found you, Drew, Sean and many of the other regular posters here very cordial in conversations. Sure, there have been times when things have gotten heated, but we are all passionate fans, and that’s to be expected. I think sometimes the fact that these are just words on a screen doesn’t help. It shows how real emotion and tone can create different reactions in face-to-face conversations. Without such capabilities on a computer screens, I fear there will always be misunderstandings.
PS. I also got the whole Forrest Gump thing with the last line.
2009 = The start of the Lions Golden Age (We hope).
by Hyperion Ecta on Jun 2, 2009 8:03 PM EDT up reply actions
Same here, KDawg
I’ll always say what I feel and hopefully you never take that as a personal attack (although I think we’ve had this conversation before). Sometimes I feel the need to make a strong statement, but I only do that because we are all adults and should be able to handle a little criticism in our opinions. Never are we all going to agree on everything and sometimes the issue gets a little heated…. that’s okay… it really is. Never in my mind am I attacking a person, only the thoughts, opinions and arguments that person is making. That person is accountable for those words and they are open for interpretation in my book. I may have been close to crossing a line above, but I was still not attacking you KDawg… my opinions were solely based on how I interpreted your words.
Again, you said it right…. this site is by Lions fans, for Lions fans and all about Lions fans, which we all are.
KDawg... I was very interested to see how you would respond...
to my comments. Look, I’m not interested in getting into pissing matches with you each time I opine on one of your comments or posts. Never is my intent to intentionally rile you up or to attack people. I was just calling it like I saw it, that’s all. When I read your comments…. and I read them thoroughly, I felt like your inquisitions were a bit too intrusive from a poster’s perspective. Every tiny element of his post was over-scrutinized and I felt it was, for lack of a better term “bully-ish”. I felt bad for someone who was posting his first fanpost and then got absolutely grilled. I felt like he was a new student proud of his work, then the teacher sat ripping him in front of the students. I was just being honest in how I read your comment and I felt like stating how I saw it.
I knew when I logged on this evening, I would see a huge defensive retort to my comments… and I knew that before I posted my comment. I was hoping that you would simply say “hey, I didn’t mean to come off like that” or “that’s your opinion, but I disagree”. Instead, you say you’ve never been so insulted in your whole life? I don’t want to attack you or your argument. I’d like to simply disagree or agree respectfully. It looks like I read this differently than your intentions, but that doesn’t discount how I felt when I read your post. You seem to have trouble with anyone that disagrees or calls you out on an opinion (not argument, but opinion). I’m sorry you feel that way.
Also, I wanted to say that I think you’re a smart guy. You’re witty and obviously educated and you have many great opinions – a lot of which I agree with wholeheartedly. And no, I would rather you don’t “dumb down” your comments. But the post above definitely came off a little “showboat-ish”, regardless of true intentions. If you can’t at least see that someone might take your comments like that, then this whole argument can probably be discarded, because your opinion is biased. You just have to realize not every reads your views the same way you write them. I’ll point out one more thing above that helped me feel as though your comments were a bit condescending, then I’ll let this argument go, because I believe that your intentions were more inquisitive than I originally thought. You responded to davis0169 a little later and basically inferred that the subject matter was “complicated” and took for granted that he didn’t know what the word “cumulative” meant, because you gave him the definition. Now, I know that he basically said he didn’t understand much in the post… that’s fine, but you in turn figured he didn’t know that word because of it’s larger complexity. I think he knows what that word means. It’s the statistical jargon that sails over everyone’s head and makes them feel small. I guess I shouldn’t try to speak for everyone, though. Maybe we all understand statistics at your level.
I’ll finish by saying, I’m sorry that I gave you your life’s worst insult. That, I did not mean to do. I just felt that the fanposter was being assulted a bit, although he did hold his own. I really just felt that the post was taken far too literally and Goodbye Ladies was simply using known stats to make the point that rookie starts don’t always equal ruined players. Again, KDawg… no hard feelings on my side.
love is the air..
Flowers are blooming. Birds are chirping. I feel like the Captain of the love boat. Now who wants to be Ike? And who wants to be Gopher?
At least we agree on some things
I am not interested in bantering back and forth with you every time I post either. You think I was showboating, and that is fine. You think I was condescending, and that is fine too. I do realize that my comments can and will be interpreted by the people who read them, and that some people may place inflections on my words that are simply not there. Unfortunately, as I and HE stated, words written on a computer screen have no tone or emotion and can easily be misinterpreted. I have no right, nor the power, to force you to feel any differently than what comes naturally to you when you read my posts or anyone’s. However, you do have the power to curb your responses. While you say that you do not wish to attack me personally, the fact of the matter is that I took your statement that I was “self-inflating” as a personal attack. The comment had nothing to do with the argument. It was directed towards me, the person who posted prior to your comment. To me, the words mean that you think that I am egotistical and that I am “showboating” to inflate my own self-esteem. I do not see this as disagreeing with my argument in any sort of respectful manner. You had no foundation or support to counter the argument, and you did not offer a counter argument. You re-stated exactly what was already said, semantically, and you took a couple of shots at me based on your own personal view of what my intent was. Do I have trouble with anyone who disagrees with me or calls me out on an opinion? Absolutely not. Will I counter and use logical reasoning to support my viewpoint? Of course I will. I understand very clearly that there will be people who disagree with what I say and what I believe. That is the nature of opinion. I have no qualm with it. Do I have trouble with people who have no counter opinion, but rather they just attack me on a personal level? You’re damn right I do. Will I sit back and take it? If you think so, you have another thing coming. Was I being defensive? Yes sir, that is the general and natural response to an offensive attack. Do I harbor ill will or hard feelings? No. Why? There are no hard feelings from my end because you do not know me from Adam, because I realize that words on a computer screen can be taken out of context, and because we are both human. If you clearly understood what I was saying, you would realize that at no point was I scrutinizing, grilling, or being critical of what Goodbye Ladies had posted. I was inquiring based on my own curiosity, and my questions were directed to satisfying it. As far as my comment to Davis is concerned, yes I did take some things for granted. However, I was attempting to cover all of the bases so that there was no further confusion. If Davis had taken what I wrote to him as condescending, then I can almost guarantee that he would have said so. I accept your apology, and I hope this makes it clear where I was coming from. There are two sides to every opinion.
Okay man, I give on this one...
I was taking some liberties based on what appeared to me to be your intent, which as you’ve pointed out, wasn’t. I responded in a reactionary way and probably shouldn’t have. It was hypocritical for me to attack your comments for what I percieved as attacking his. But know that I was attacking what I felt was your intent versus attacking you as a person. That’s probably hard to tell from cold comments alone. As for me countering your opinion… I didn’t need to, because I agreed with much of your assessment on the topic. My comment to you was a new topic based on the tone of your questionnaire and to see how you would respond to my point of view.
Look, I’m a direct person. I think you’ve figured that out about me by now. I don’t make a point to intentionally rile feathers, but I also don’t shy away from giving an honest opinion, which can sometimes hurt feelings a little. I just don’t think twice about a little heated debate now and then. After all… this is football. In my work environment, conflict is way of life and typically a means to an eventual compromise. What you write here is open for discussion, argument and interpretation. You just have to realize in an environment like a blog, people are going to disagree and there is going to be conflict. I’d just say, don’t take it so seriously. To be honest, this is a really tame blog site. Sean does a great job keeping it that way and I personally like it that way. I’m not into vulgarity or nonsensical obscenity. That’s why I chose this as my home blog site for the Lions. But that doesn’t mean we shouldn’t be able to speak our minds a little and not be able to “call it like we see it”. If you can’t hack this site, maybe blogging isn’t your thing (not talking directly at you KDawg with that last comment – you can obviously handle yourself). Okay, so I guess in a nutshell, I’m just saying don’t take the comments so much to heart. We don’t know each other personally, so why would the attack be personal? Yes, that is hypocritical of me to say… and yes, I’ll be taking my own advice.
But at the end of the day, I don’t want to come off like a jerk or overly insensitive or a troublemaker. If you knew me personally, I’d like to think you would see the direct opposite of that. I especially don’t want to be attacking anyone personally and I may have done that here (but do realize it wasn’t intentional). I apologize again for that.
One word…… TRUCE?
One more small comment, KDawg...
I didn’t think you were either showboating or condescending. I thought the tone of your questions were showy and condescending. Not sure if you agree, but to me… that is a very big difference.
Truce
Drew……I am a man. As a man, I do not cry when someone “hurts my feelings”. Unfortunately (I guess), I typically have the general male response to hurt feelings, which is anger. In this case, I was more offended than hurt…..since the source of the offense came from an acquaintance and not someone that I know on a personal level. I honestly think that you and I would be friends in person. I also feel that way about just about anybody here. I am a joker….a fun loving guy. We are different, in that I do enjoy vulgarity and obscenity on occasion….lol. But that does not mean I do not know when it is appropriate and when it is not. I come here to post because I think that this is the best Lions forum I have ever seen. I have been to many others, and this is the only one where I usually respect what is being said by a majority of the posters. That is not to say there are not some posts and posters that I do not enjoy. I am a kind and generous person, but my kindness is often mistaken for weakness. People try to take advantage of me a lot, and that is a mistake. I understand where you are coming from, in feeling that I was over doing it with the technical terms of statistics. I guess I just wanted to show that I knew what I was talking about. I did not intend to be showy or condescending, and I was angry that someone felt that I was. Regardless, I will work on not taking comments too personally. Once again, we’re cool.
Great Post - rec'd
Some of the statistics went a little over my head, but I get your point, and agree: I think Stafford should get every minute of playing time his contract requires (The 35% number I’ve seen in several publications). Maybe more, depending on how the season goes. If Daunte tanks early (which I hope and really don’t think he will) then we throw him in. I don’t get the notion that somehow our very intelligent #1 overall QB is going to have some sort of epiphany or revelation during the offseason, vs. starting at some point this year.
I want to give Daunte his time, but by drafting a QB #1 overall, you have given the team to him. There is no reason to hold out on that, unless Mr. Culpepper has a Pro Bowl type of year. We have nothing to lose!
the problem i have with this
everyone of those Quarterbacks. Were not in the same situation. Couch had injury problems. Harrington was picked by the GM and not the coach. His style of play did not translate to the system being used. Leaf was a head case. Akili Smith should have never been drafted in the first round. He was a product of over-hype by his agent. on and on and on. What about teammates? And the line" If you’re Peyton Manning, you’re going to eventually be Peyton Manning whenever you start—irrespective of whether you throw 36 interceptions and are sacked 60 times. However, if you’re J.P. Losman, it doesn’t matter if you have two years to learn and sit on the bench—when you are inserted into the line-up, you are going to fail. " That does not make any sense. What about Steve Young? Failed in the beginning. Than became a Hall of Famer.
What I am saying is that. There is no one way to develope a QB. Or every team would do it. The human factor is and will play the biggest role of a QB or any NFL hopeful. Does the system fit the skill set of that player? Is it a stable coaching situation.
Orlovsky sat for 3 years? He came in and played….decent. But every game he seemed to be getting more and more comfortable.
I don’t care who you are, you don’t get better until you play.
If Stafford is equally talented and has just as much of a grasp on the playbook as Culpepper, he has to be the guy.
I can't agree with that 100%
Guys do get better playing…..in practice. You think Aaron Rodgers was as good as he was last year as a rookie? Give me a break. The guy was developed behind Favre, plain and simple. Yeah, he may continue to get even better now that he is playing in actual games, but saying that you don’t get better until you play is not true. I agree with you that if Stafford proves to be as good as Daunte, or better, and he understands that playbook on an equal, or better, level than Daunte, he has to be the starter. Until he proves those things however, he has to continue to get better in practice and develop into that role.
What is your criteria for 'ultimate success in the NFL'?
When I look at your list of NFL QBs picked in the first round, if you look at those who threw over three hundred passes in their rookie year, I see one success; Peyton Manning. Matt Ryan and Joe Flacco are unproven quantities and the rest of the guys could be classed as underachievers.
Even the next grouping, from less than 300 to over 100 attempts, is full of question marks.
Is Eli Manning a success? Sure he won a Super Bowl, but if his career ended now I would consider him to be this decade’s Jeff Hostetler. What about Cutler and Vick? Successes? Too early to tell?
The last group, less than 100 attempts, has arguably four successes [Pennington, McNair, Palmer, and Daunte] and two probables [Rodgers and Rivers] with two more unprovens [Russell and Campbell].
To me, your data is proving the exact opposite of your conclusion.
Really though, you should only be looking at players who have played at least five years in the NFL. Any shorter time period is too early to tell if a QB is a success or not.
But I applaud your effort and thought to come up with this theory. I just don’t agree with it.
I like where you are going
A couple things:
In your first analysis, you mention that the R2 is .04 and that the number is not significant. You also mention your results are contrary to popular belief. It is contrary to common sense, but only in that popular belief is that starting early hurts a QB and you found that there is no relation between when a QB starts and if they are a success. Don’t over extend your findings by claiming that starting early helps.
Also, as mentioned above (NorthLeft12), you need to tell us how you decided a QB is a success or not. At least, give us who you thought was a success (I could look at the data you posted and I will later).
I am wary of mystical formulas like your DYAR. All DYAR is is what the guys at that other site think are predictive of the QBs success. You have a lot of good collected data already, I would rather you just added more variables to your model and ran a larger regression to find out which ones help predict success. Maybe some variables like draft position (higher picks go on teams that have fewer other good players), wins in the season (maybe all the losses in their first year get in their head and make them fail), college wins (if you think a good QB is born, not made). Throw all these in your regression model and you will be able to see not only if any of these factors have predictive power, but also if their are any interactions that may be interesting.
Your argument is essentially a nature versus nurture for QBs. IS the QB a success because of the training in the NFL or is the success inherent in the QB? You put up Manning as an example of a guy who was going to succeed. I will put Tom Brady up on the other side. Does anyone think that if he was picked up by the Raiders instead of the Pats he would have 3 rings and multiple trips to Hawaii?
I do like what you did and praise you for it.
ATL Lion and NorthLeft
NorthLeft and ATL Lion,
You guys both bring up some good points, so I think I will address them both here. The metric is DYAR/Pass. Football Outsiders has a detailed explanation of DYAR on their web site, so I won’t repeat that here. I decided to use DYAR/pass because players who are mediocre players, but who have a lot of DYAR because they pass a lot. Football Outsiders is a pretty reputable site and I actually prefer their statistics over conventional ones, so that is why I used it. It is also why I didn’t go back into the past too far because the statistics only go back to 1995 or so, so anybody drafted before that I unfortunately could not include (which is also why I can’t go back five years like NorthLeft suggested).
Of course, the QB’s all come in at separate points in their careers. I had nine years worth of DYAR/pass for only about ten quarterbacks, so I modeled their careers to be able to “age” data for younger QB’s like Flacco, Ryan, Russell, etc. It’s a small sample size, I know, but the results were linear and pretty intuitive to me, so I went with it. Utlimately, it didn’t end up making that much of a difference. I didn’t bother “aging” guys who were out of the league.
Anyway, a big X-factor in this is Matt Ryan, which I think will be the most controversial, but that’s the way the numbers came out. He had what can only be termed an unbelievable season for a rookie. By modeling him after the ten quarterbacks, he only got a very mild bump because basically the model thought that he really couldn’t get much better.
Anyway, the spreadsheets that I posted in response to KDawg have all of the numbers, but they aren’t sorted, so here’s what they look like. They are pretty intuitive to me, although I know folks will disagree with where Ryan is:
Peyton Manning 2.744892
Matt Ryan 2.63676
Philip Rivers 2.17185
Carson Palmer 2.06987
Jay Cutler 2.04699
Ben Roethlisberger 1.88942
Chad Pennington 1.782236
Aaron Rodgers 1.61086
Steve McNair 1.46273
Daunte Culpepper 1.3582
Jason Campbell 1.33138
Joe Flacco 1.30781
Eli Manning 0.996776
Byron Leftwich 0.96917
Donovan McNabb 0.856406
Matt Leinart 0.769617
Kerry Collins 0.407623
Vince Young 0.301921
JaMarcus Russell 0.267778
Patrick Ramsey -0.096764
Kyle Boller -0.1109
Rex Grossman -0.13236
Joey Harrington -0.205373
Michael Vick -0.244951
David Carr -0.320752
J.P. Losman -0.48435
Cade McNown -0.536972
Tim Couch -0.670668
Alex Smith -1.52016
Ryan Leaf -2.108401
Akili Smith -2.404624
Those were the numbers that I used. I agree that there may be ways for those who have more formal statistics background to improve the methodology, but I stand by the ultimate result. Again, if you just use the number from Ryan’s rookie year, he comes in at approximately 2.56 DYAR/Pass, so the aging doesn’t make much of a difference.
I can understand NorthLeft’s position that there are more “successes” on the bottom of the top. However, the reason that it all works out the way I posted was because of the relative success. Manning had the most attempts and has been head and shoulders above almost everyone else. Also, some of the really bad players (the Smiths, Grossman, Losman) are on the lower end of the rookie attempts scale. Of course, you’re free to disagree with it if you would like and do research as a counterpoint, but that’s why it comes out the way it does. I would only suggest that whatever it is be “comprehensive,” and not just plucking Carson Palmer out of the bottom of the list and looking at him.
As far as the statement about rookie quarterbacks performing better, it’s just a statement of fact. Based on my numbers, they performed slightly better. I stated in my post that it’s not statistically significant, and I will say that I don’t believe that playing early generally helps or hurts a quarterback. I just think it has no bearing at all.
If you check my blog, I actually have a post where I more fully analyze what makes rookie quarterbacks successful by breaking down the Lewin Career Forecast, which is pretty much the gold standard in my opinion for predicting NFL success at the position. I will say that I have been tinkering with that a little but haven’t posted anything about it, although I still stand by my guess that Stafford/Freeman will end up like Kerry Collins and Sanchez will end up like Joey Harrington. I will say that I did throw in draft position (within the first round) and there was no correlation.
And no, Tom Brady may not have three rings if he was on the Raiders, but I think that he would have multiple trips to Hawaii. Tom Brady took very little “developing”: he was on the bench for only a year and performed very well as early as his second season. I refuse to believe that his emergence is due mostly to one brilliant year of coaching. If that was the case, then why aren’t Kliff Kingsbury and Rohan Davey burning up the league? Anyway, thank you both for reading and the good word. I can go on and on and on talking about this stuff!
by Goodbye Ladies on May 24, 2009 9:07 PM EDT up reply actions
Maybe I am not understanding
Are these numbers the numbers you used as a measure of “success at QB”? And these are the DYAR/pass attempts?
OK, maybe Brady only being on the bench for a year was a bad example. How about Hasselbeck. He was on the bench for many years before getting his starting shot. The point is, like the nature/nurture debate, there is no good way to tell if it is coaching for a year or inherent ability that makes a good QB. Maybe if we get identical twin QBs.
Hyperion makes some good points later on about the difficulty of trying to base a measure of NFL success on an individuals statistics. The NFL is not an individual sport. QBs can have bad years in the middle of their careers because of lack of talent around them. A bad line, no receivers, etc. Look at Peyton Manning in 2001. The defense was terrible and he was constantly playing from behind, forcing the ball. The next year Dungy came in and revamped the defense. There is so much more going on in football than one individual.
That said, I am still a fan of looking at factors that can help a QB. Is a good receiver better than a good RB? Does a good O line help better than a good Defense? Actually, Schwartz looks at stats in a similar way. There are some articles about how he ranks player positions by how important they are to a win.
Good post
There’s also some great comments here, very involved and very, very long. I’m a statistics man myself by I don’t tend to subsribe to predicting someone’s career on the career’s of others. I believe a QB has his own talents and will meet his own specific environments in his career. There is a correlation between certain things, like you have brought up, but we will only know how good Staff is when he hits the field. Good post anyway.
I also believe Schwartz will pick the best QB to play on the day. I suspect Staff to get a fair bit of playing time this year.
2009 = The start of the Lions Golden Age (We hope).
agreed Hyperion
I really do not buy into these " proposed methodology" and " approximate value systems." To say you could put a number on a person. And predict his career before he even steps on the field. It does not make sense to me. And I am a stat guy. I understand the method that is being used. Football is played by people. And not programed machines. Where do you account for the Humanity in the game? I applaud the effort. I really do. But when you put numbered values on everything. Where is the fun of sports talk? Who had the better defense? Well, the 2006 Ravens had a grade of 46.7. While the 1978 Steelers had a grade of 41.3. What does that really mean? Statistical analysis is fine and interesting. But I would rather discuss football the old fashion way. In a bar, with complete stangers, over some beers. The way God intended.
I don't agree with applying sabremetrics to Football in general
This seems to me to be an application of Bill James’ statistical work on baseball to football.
There are a few problems with that.
1. The nature of the two sports. Baseball is much more of an individual match up as opposed to football. ie. pitcher versus batter. There really is no individual match up in football. The closest would be WR on CB, but that still has to include QB performance, pass blocking/rush, safety help, field position, down, etc.
2. Individual performance in baseball is highly impacted by the environment. ie. stadium dimensions, dome, wind direction and speed. Football is played on a standard field and is impacted by extreme weather [snow, rain, high wind, cold] only.
3. The sample size. Baseball has ten times as many games a year, and careers last longer, which should mostly eliminate the impact of most anomalies and also supply a large enough data base to account for environmental impacts [ie. home vs away data].
I agree with HE and JC above, predicting future performance based on what similar QBs have done is very problematic. Just looking at the list of QBs in your comparison should show you that the NFL in general is struggling to figure out who will be successful and who will bust, and, as you tried to address, how best to develop them to achieve that success.
Makes for a great discussion thread though!
Agree with a lot of what you said but...
I that you are exactly right that there are problems with applying statistical analysis to football, for all of the reasons that you’ve mentioned. I would add though that there are those kind of problems with any type of football analysis.
Compare the NBA Draft to the NFL Draft. High picks in the NBA Drafts become busts much more rarely than high NFL picks. Why? Because basketball is much less complex and relies much less on teamwork than the NFL. A good point guard is going to be a good point guard, irrespective of his teammates. However, a good running back is not going to be productive without an offensive line and a QB who can take the safety out of the box. NFL players (college or pro) are tremendously difficult to evaluate (and not just through statisics) because their success relies so much on their teammates.
In other words, yeah, advanced statistics are not going to be as useful in football as baseball, but that doesn’t mean they’re useless.
by Goodbye Ladies on May 30, 2009 12:07 AM EDT up reply actions
North, I think you are right in a sense...
that football doesn’t lend itself as neatly to statistical models as say, baseball does. But I think the crux of the post was simply to state… in a quite complex way… that quarterbacks are individuals and you can’t judge the success of a rookie by either a) the team or situation he was drafted into, b) if he gets on the field as a rookie or even c) how well his rookie production translates to career standings.
I felt that Goodbye Ladies simply showed that individuality is an extremely large and complex variable in the quarterback success formula.
I do agree with you about moving on.
However, this is where you need to have people in charge who are good evaluators of football talent and secure enough in their positions to make the tough decisions like dumping a first round draft choice.
As the links you referenced showed, Harrington just did not make any progress in his first two years, despite loads of playing time and pretty good protection from his O line. The other factor not discussed is the physical capabilities of the QBs. The ultimate successes at QB mostly seem to have strong arms and need to learn how to control that ability. Weaker armed QBs seem to get by in college on smarts and the smaller physical advantage they have over their competition. In the NFL, the faster pace of the game is too much for them. QBs like Harrington, Carr, A. Smith seem to fit in that mold. Though I have not seen him play, I suspect Sanchez may fit into that mold too. Stafford seems to belong in the Aikman, Doug Williams, Cutler group.
It would be interesting to know how much of this kind of thought is going on within the Lions organization. I suspect there will be a lot more of it when camp is underway.
I'd let him watch Culpepper for a year
Remember Harrington
"There's no substitute for guts."
-- Paul "Bear" Bryant
Who?
There are periods of my life I black out, usually due to substance abuse in response to this guy that used to throw the football for us… his number was 3 I think….
#3 NOOOOOOOOOOOOOOOOoooooooooooooo
My blood presure can’t handle that number still to this day, Where are my heart pills???? and the mighty HEMP cousin.!
GBL
Let me see if I understand…
1. Starting or not starting has not been shown to affect an NFL QB’s future success.
2. There is a fairly strong correlation between rookie quarterback performance and ultimate NFL success.
3. Stafford should start so that Lions can more quickly evaluate him. If he’s a keeper, then great. If not, we can cut bait on him quicker and try it again in 2011.
If I’m wrong about your point, please let me know. If I am right, then let me say that I think you make a very good point about cutting bait and moving on if we need to. However, the fact that there is no correlation shows nothing at all. It does NOT mean that whether a QB sits or starts will NOT affect his future success. It does not mean anything at all. Simply not significant.
I don’t think that creating the perfect Stafford report card should be the driving force in the decision to start him or not. I believe that decision should be based on:
1. Will the rookie survive out there? Can our line provide enough protection to keep him from getting injured? If he gets sacked half the time he goes back to throw, it’s going to be darn hard to evaluate that post-op patient!
2. Is Stafford as good, or better than Culpepper? If the answer to either of these first two questions is no, then sit the Rook.
3. How is this individual motivated? Can he develop more by focussing his brain during game time? Remember, faster and better decision making is sometimes just as important as a rocket arm. Or, does he need to get in the game?
The Myth of the Lions Poor Pass Blocking
Last year, the Lions gave up 52 sacks to their opponents. Kitna and Daunte were sacked once every eight [8!] times they attempted to pass. That is horrible. Orlovsky was sacked once every eighteen 18 times he attempted to pass. That is around the average for the league. Some other QBs;
Cutler 1 per 56; Cassell 1 per 11; Rodgers 1 per 16; Roethilisberger 1 per 10; E. Manning 1 per 18; Rivers 1 per 19; Matt Ryan 1 per 26; Flacco 1 per 13; Favre 1 per 17.
I think the Lions pass blocking was a lot better than the public [uninformed] belief was. The high sack totals are almost totally traceable to poor QB play. Hopefully, Daunte is better prepared this year and will release the ball quicker to avoid a lot of sacks. If not, we will see Stafford sooner rather than later.
Good point North
I never thought the O-line was as bad as the media made it out to be. And when the Lions got a running game going. The sacks came way down. Kevin Smith’s best four games. All in the second half of the season. Lions only allowed 6 sacks. I have said this many times. Better play design. Better coaching. Better play calling. The Lions O-line could be average to good.
I agree with your agreement.
I felt the Lions were too predictable, and did their best when they mixed up their play calling. I still see our offensive line as weak in run blocking [we have been below average in rushing for a long time] and I can’t blame it entirely on our mediocre running backs.
I am eager to see how the offence performs in 2009.
I agree
I thought I’d look at some basic stats to see if how the Lions have gone in recent seasons. So, in looking at total sacks during the regular season from 2000:
2000 – 53 sacks (26th) – QB: Batch
2001 – 66 sacks (31st) – QB: Batch
2002 – 20 sacks (1st) – QB: Harrington
2003 – 11 sacks (1st) – QB: Harrington
2004 – 37 sacks (16th) – QB: Harrington
2005 – 31 sacks (10th) – QB: Harrington
2006 – 63 sacks (31st) – QB: Kitna
2007 – 54 sacks (30th) – QB: Kitna
2008 – 52 sacks (31st) – QB: Take your pick
So when you look at those numbers, you’ll see a distinct period of 4 years by which we were pretty good in pass protection. 2 1sts is actually very good. Those years were with Harrington. Beforehand and afterward we shot right up again. I think this line, especially with some of the upgrades we have made, can protect a QB…as long as he doesn’t hold the ball for a long time. Culpepper does have a tendency to hold the rock for a while so I don’t expect a huge change in numbers but maybe a return to the high 30’s or low 40’s would be nice.
2009 = The start of the Lions Golden Age (We hope).
by Hyperion Ecta on Jun 1, 2009 8:21 AM EDT up reply actions
Quite an eye opener isn't it?
I still find it hard to believe that the Lions were as low as eleven [11!] in 2003. I guess Harrington did have some redeeming qualities. Too bad he could not do much on the positive side [ie. completion %, TDs, yards per attempt].
When you break down 2008 between Orlovsky and Daunte/Kitna you will see the same kind of performance. Especially if you weight it by the pass attempts.
Exactly Right
I think that shows how the quarterback has a much greater effect on the number of sacks given up than he is typically given credit for. Although David Carr is often cited as a quarterback who was “ruined” by his offensive line, when Carr was replaced by Schaub, the total sacks the Texans gave up was reduced by half—and that’s with pretty much the same offensive line and the same offensive system.
Usually when a team does not give up a lot of sacks, it is a testament to the QB’s ability to quickly progress through his reads and to get the ball out of his hands fast. However, with Harrington, he just made his first read, and if that man wasn’t open he went immediately to the checkdown, hence no sacks.
by Goodbye Ladies on Jun 1, 2009 9:13 PM EDT up reply actions
So the answer is
Daunte is the problem and he will most likely be our starting QB next year.
Expect the massive amount of sacks to continue
Just shows the quick 3 step drop passes in the west coat offense
It’s hard to get sacked on a two yard pass
by Runnin' Rebel on Jun 1, 2009 1:17 PM EDT up reply actions
I completely agree with you, North
The Stafford “will get killed behind our line” theory just doesn’t hold much weight for me. He’s much more mobile than Kitna and even Culpepper for that matter. This shouldn’t be a point of concern.
Its not his mobility that will reduce his sack total.
It will be his quick release and recognition of what is happening downfield. I think that a QB’s mobility is hugely overrated in keeping his sack numbers down. M. Vick was sacked 33, 46, 33, and 45 times in the four full seasons he played. He missed most of 2003 [he played five games] and the Atlanta QBs were sacked 35 times that year.
Roethilisberger is another example. He is big, strong, and elusive but he still gets sacked almost as much as anyone in the league. Then there is Matt Cassel and Tom Brady. Brady, 21 sacks in 2007. Cassel, 47 sacks in 2008. I don’t think there is any question who is more mobile.
I’ll concede that mobility can help a bit, but I always wondered if some of the scramblers put their linemen in a poor position when they moved out of the pocket too early.
Scrambling QB's
tend to rely on their feet to much, which obviously leads to them running at the wrong times. So, I agree that scramblers aren’t as unsackable as they would seem. Scrambling is more about picking the right time to run then running it because you can.
2009 = The start of the Lions Golden Age (We hope).
by Hyperion Ecta on Jun 2, 2009 8:06 PM EDT up reply actions
Maybe I should have used the word "pocket presence"
I think that more adequately describes what Stafford has versus saying he’s mobile or a scrambler. He seems to have the Montana or Brady-esque ability to glide around in the pocket and avoid pressure to reduce sacks. But he is mobile enough to bootleg as well.
Okay… so my point is that he just shouldn’t get sacked as much as Kitna… how ’bout that :)
But I agree with you North, that his quick release may become one of his best attributes. It’s the Marino rule… quick release is better than quick feet!
Nobody should invite the sack as much as Kitna
Drew, I know I am picky about some wording but watching, listening, and reading about football we are continually bombarded with myths and stereotypes. One of the oldest and most easily debunked is the relationship between the “mobility” of a QB and his sack rate.
I think you worded it perfectly.
Culpepper should start
Stafford should become the #1 anywhere in between Halloween and Thanksgiving.
Sounds good to me
2009 = The start of the Lions Golden Age (We hope).
by Hyperion Ecta on Jun 1, 2009 7:41 PM EDT up reply actions
Fantastic stuff, Goodbye Ladies!
I know I’m late to the discussion here, but I just wanted to say that I thoroughly enjoyed this post. You came to a conclusion that I’ve touted for many years, but was simply too lazy to put working statistical models to good use and back up my claims.
My conclusion has always been… whether you are talking about drafting a college quarterback or evaluating future successes… that quarterbacks are individuals, through and through. You cannot simply lump them into a category based on school, scheme, style or any other historical correlation one can make about said quarterback. Sure some are a little easier to categorize than others… i.e. so-called “system” QBs, but even within those stereotypes are individuals with unique skillsets.
So back to the main objective… whether or not to start Stafford sooner rather than later… I would simply say this: Start him when he is the best QB on your roster. I hope (strongly) that it is Week 1 against New Orleans. Not a big fan of the “new, post ’04” Daunte Culpepper and the sooner we can find out if we have our future secured with Stafford, the better. This silly notion that he must sit a year simply because the Lions might “ruin him” or that he’ll get “killed” behind our line is borderline absurd (he’s a little more mobile and athletic than Culpepper or Kitna). If he’s better than Daunte by the end of camp… hands down, he should be the guy at New Orleans.
I’ve maintained throughout the process of the Lions interrogating and drafting Stafford, that if Stafford was the true-blue franchise QB for Detroit, he should start as soon as the system is in place and learned. That should be by Week 1. My reasoning has been that even on a poor team, great quarterbacks will show flashes of brilliance. He still needs the franchise to put the other pieces in place, but for his part… the talent and ability is there. Like you said Goodbye Ladies (by the way… love the handle!), Peyton Manning was always destined to be Peyton Manning. Never was he going to be Ryan Leaf. His team success would always be measured by the team that the Colts put around him, but he wouldn’t have been less of a quarterback, ability and intelligence-wise. His stats may have been a bit different, but he was always Peyton Manning.
Likewise, Stafford is Stafford. The Lions can help his progress through good coaching, successful scheming and of course, by giving him protection and targets. That said, Matt Stafford will not work any less, be any less intelligent, have any less arm strength or have less of any of the intangibles that God so graciously blessed him and now Detroit with. I can’t wait to see him perform. I’ve been on the Stafford wagon for a while and I think that I am about to see the reasons that I liked him so much come to fruition.
Anyway… great post, Goodbye Ladies. One of the best I’ve read in a while. Keep ’em coming!
I am a recent convert to the play Stafford soon group.
I had always agreed with conventional wisdom that a rookie QB should sit and watch a year or so before playing, but I agree that it should be totally the judgement of the coaches whether or not he gives the Lions the best chance to win, then he plays.
I have to admit I am getting pretty excited about the 2009 Lions. I am not expecting miracles, but I am expecting a much improved team on both sides of the ball. A competitive team will be a nice change after a year and a half of incredible incompetence.
I can't wait for the season
It doesn’t help that my Wests Tigers aren’t going brilliantly at the moment. It feels like forever since I’ve won any football games as a fan.
2009 = The start of the Lions Golden Age (We hope).
by Hyperion Ecta on Jun 2, 2009 8:07 PM EDT up reply actions
I'm guessing rugby
I think that’s huge in Australia. But that’s only based on what Russell Crowe said when he came to a Michigan game in 2007.
Pride of Detroit, SB Nation's Lions Blog
In a way
Australia is immensely diverse in it’s sports. We have 4 major football codes. There’s AFL or Aussie Rules which is probably the one you think of when you think Australian football. Ben Graham & Sav Rocca are former AFL players turned punters. Then there’s Rugby which is split into two different games – Rugby League (which I follow) and Rugby Union. League isn’t hugely well known internationally, only having larger followings in Australia, New Zealand and England. Union however is fairly well known and when most people say rugby, they picture union. I believe it’s fairly big in New England actually. Then there’s soccer which has a decent following here.
Anyway, the Wests Tigers, (which is the logo on top of the Lions logo in my display pic) are my Rugby League team and are the only team I care for more than the Lions. Luckily, League season runs when the NFL season is off so there’s no clashing.
Also, Russell Crowe owns the South Sydney Rabbitohs, a Rugby League team, which is very strange for us because most teams are run by leagues clubs so they have no real owner. He actually does a pretty good job as a owner, let’s just say he’s no Al Davis or WCF.
2009 = The start of the Lions Golden Age (We hope).
by Hyperion Ecta on Jun 3, 2009 3:03 AM EDT up reply actions
Hey, thanks for the information
We have quite a bit of rugby here in the states as well – just that not much of the professional stuff is widely known or displayed. I actually go and watch my neice play in her city league here. It’s a rough sport. I have respect for those who play it.
Hooroo!
Yeah, his rugby team is how he and Lloyd Carr became friends. Carr went to a game and spoke to the team and Crowe returned the favor after the Appalachian State and Oregon debacles.
Pride of Detroit, SB Nation's Lions Blog
Stafford
I’ve been saying I’d prefer Stafford to sit the majority of the year, but there are some great posts here that are making me alter my thinking. My new opinion is If he clearly outplays Daunte then he should start day one. If it’s a very close race then Daunte should start and let’s put Stafford in after the bye week and get him 9-10 games this year. My main conern with starting him day 1 is the fact that we seem to have a tough opening strech of games and the pressure to get a win after no wins last year is a lot to put on an underclassmen rookie QB.
Also, when I was college I spent a semester in australia (by far the best decision I ever made), and Aussie Rules Football is great. My way of describing it would be a mix of football and soccer…and as a big fan of playing both of those, I really enjoyed watching AFL games.
Comments For This Post Are Closed
Secondary Sidebar
User Tools
FanPosts
Community blog posts and discussion.
Certain photos copyright © 2012 by Associated Press or Getty Images. Any commercial use or distribution without the express written consent of Associated Press and Getty Images is strictly prohibited.
