1. This site uses cookies. By continuing to use this site, you are agreeing to our use of cookies. Learn More.

Great Older Article on Fitzpatrick

Discussion in 'Miami Dolphins Forum' started by KeyFin, Jun 6, 2019.

  1. cbrad

    cbrad .

    10,659
    12,657
    113
    Dec 21, 2014
    I think that's oversimplifying things. I think the main reason Tannehill isn't a starter is that with 6 years starting he's made it clear he's not anyone's QB of the future. There are other QB's that aren't as good as Tannehill or are similarly good/bad but there's more uncertainty with them because they haven't played as long. This includes rookies and I'd say QB's with 4 or fewer years starting.

    Another reason he's not a starter is that there are a bunch of starters that are no better than Tannehill but it just makes no sense to switch. Eli of the last few years falls in that category as does Dalton. I'd even put Carr there though maybe some want to see another year of him?

    Finally, there's the case of Flacco which makes no sense at all from a performance standpoint. Does GM Elway think there's some of that old Flacco left? Not sure but Flacco has been a bottom tier QB performance wise for many years now, certainly much worse than Tannehill and much worse than most starters.

    So I think it's mostly that Tannehill has had 6 years to prove he isn't the solution as well as there being similar ability QB's where it just makes no sense to switch. And Mariota is on the verge on falling into this category too, but he's still at 4 years starting.
     
    resnor likes this.
  2. KeyFin

    KeyFin Well-Known Member

    10,488
    12,821
    113
    Nov 1, 2009
    Last season Tannehill was 20th overall with a 92.7 rating...Rodgers was 10th with a 97.6. The difference between 10th and 20th is a whopping 4.9 points. The difference between 20th and 16th is 2.2 points. Saying Tannehill is below average is a gross simplification of starting quarterbacks- a few more bombs to Stills, a few less strip sacks and he's a top 10 player.
     
    resnor likes this.
  3. Losferwords

    Losferwords Member

    77
    60
    18
    Sep 1, 2012
    Thill was an albatross... can’t believe his defenders are still out in force.

    I hated every waking second with that mouth breather at the helm... every waking second.
     
    djphinfan likes this.
  4. The Guy

    The Guy Well-Known Member

    6,598
    3,323
    113
    Oct 1, 2018
    The third sentence in your post is precisely why “top 10” has very little meaning.

    “Top 10” isn’t good enough and isn’t something to aspire to, unless it’s also top four. Tannehill could’ve been any of quarterbacks five through 10, and there still wouldn’t have been a significant enough difference from what he actually was.

    Such is the nature of quarterback play in the present-day NFL.
     
  5. djphinfan

    djphinfan Season Ticket Holder Club Member

    111,648
    67,540
    113
    Dec 20, 2007
    Gross my azz..nice comparing passer ratings as your system of grade..and he still winds up below average.

    Aaron Rogers makes plays with his legs that tannehill can only dream of..
     
  6. djphinfan

    djphinfan Season Ticket Holder Club Member

    111,648
    67,540
    113
    Dec 20, 2007
    I’m not over simplifying anything..facts are facts..

    I think if anyone is going to make this argument that this guy was an average qb they better draw the line as to what is average relative to qbing in the NFL..

    no one in the league would give up their guy for tannehill..fact.
     
  7. KeyFin

    KeyFin Well-Known Member

    10,488
    12,821
    113
    Nov 1, 2009
    Well, Brady wasn't top 4 any of the past 4 seasons and he's been to 4 Super Bowls in a row. That tells me top 10, top 4 and even top 1 doesn't mean enough if the rest of your team isn't up to snuff.
     
    resnor likes this.
  8. KeyFin

    KeyFin Well-Known Member

    10,488
    12,821
    113
    Nov 1, 2009
    Sorry- I guess someone being objective is offensive to you somehow. I guess I should have compared them by shoe size or fishing skills instead.
     
    resnor and Hooligan like this.
  9. cbrad

    cbrad .

    10,659
    12,657
    113
    Dec 21, 2014
    Umm.. dude, first of all please quote the correct numbers. Rodgers was 13th, not 10th. Cousins was 10th with a 99.7 rating, meaning that last year we're talking about a 7 point difference in passer rating between 20th and 10th.

    So what does a 7 point difference correspond to? For an average NFL team you have 34.5 passing attempts per game. You'll get the same rating as Tannehill (92.7) if you plug in the following:
    Completion = 22.2
    Attempts = 34.5
    Yards = 249.2
    TD's = 2.14
    INT's = 1.14

    So that's what the per game stats look like, on average, for a QB of Tannehill's passer rating in 2018. What does it take to turn that into 99.7? You could add 0.725 to TD's, which translates to 11.6 extra TD's per season. Or you could decrease INT's by 0.58 which translates to 9.3 fewer INT's per season. Or you could increase yards per game by 58.

    You think that's easy? It's not "a few extra bombs to Stills". Try 11.6 extra TD's thrown, or 9.3 fewer INT's thrown or 928 extra yards thrown in a season to go from 20th to 10th. There's a reason this isn't easy, and no Tannehill has NEVER been top 10 statistically.
     
    Last edited: Jun 14, 2019
    Hooligan, Surfs Up 99, KeyFin and 3 others like this.
  10. cbrad

    cbrad .

    10,659
    12,657
    113
    Dec 21, 2014
    You stated "If he was average he would have a job as a starter.."

    That is NOT a "fact".. it's called an "opinion". And my post explained why your opinion is probably wrong and oversimplifying things.

    You want a technical argument for justifying saying Tannehill is "average"? I'll give you one. Do a hypothesis test on his year by year z-scores using a 2-sample t-test and what you get is a 60.47% probability those z-scores come from an average QB (if the z-scores were all zero that would be 100% for reference, so even a "true" average QB that has random variability will not have 100%).

    Statisticians generally reject a hypothesis when it's less than 5% likely the hypothesis is true. So no statistician will reject the claim "Tannehill is average". Having said that, directly calculating his career z-score to be -0.1166 is much better. So it's definitely more accurate to claim Tannehill is "slightly below average" but you can't reject the claim he is "average". That's at least enough for me not to criticize someone making that claim.
     
    resnor, Pauly and The Guy like this.
  11. djphinfan

    djphinfan Season Ticket Holder Club Member

    111,648
    67,540
    113
    Dec 20, 2007
    I don’t agree with this, you need solid competent qb play, a qb who can make some plays when needed, good game management, and smart situational football player, you don’t have to have an elite qb imo, if that winds up numbers wise as in the top ten, that’s good enough to win a champ with.
     
  12. djphinfan

    djphinfan Season Ticket Holder Club Member

    111,648
    67,540
    113
    Dec 20, 2007
    You missed my fact, I said he is not a starter in this league.thats a fact..

    My opinion about tannehill has been more accurate then any stats can illustrate, that stats only support it.
     
    Last edited: Jun 14, 2019
  13. The Guy

    The Guy Well-Known Member

    6,598
    3,323
    113
    Oct 1, 2018
    Brady was 4th in 2015, 2nd in 2016, 3rd in 2017, and 12th in 2018.
     
  14. The Guy

    The Guy Well-Known Member

    6,598
    3,323
    113
    Oct 1, 2018
    What would be interesting in my opinion is if you were to list all of the current quarterbacks who in your opinion fit that description, and we could then analyze exactly what kind of quarterback play you’re talking about, statistically speaking.
     
    cbrad likes this.
  15. cbrad

    cbrad .

    10,659
    12,657
    113
    Dec 21, 2014
    This is actually a good suggestion because it's one way to show people who are skeptical of using a simple summary stat like passer rating that clearly doesn't incorporate everything you'd want, and that is technically a "team" stat, that it nevertheless tends to fit with most people's opinions of who the best/worst QB's are.

    In order to do this properly though you need to assign a rating, let's say from 0 to 10 where 0 is the worst and 10 is the best and you can use any number in-between (e.g. 4.3). With a rating one can calculate a correlation and quantify the similarity, without it you can't.

    Oh.. I should add that it's important that the ratings not just be for who you think are the best QB's. That won't help. You need to rate a representative sample of QB's from the league (as if you randomly chose them). Ideal is if you rate all 32 or come close to it.
     
    Last edited: Jun 14, 2019
    Surfs Up 99 and The Guy like this.
  16. KeyFin

    KeyFin Well-Known Member

    10,488
    12,821
    113
    Nov 1, 2009
    First off, I grabbed the stats from here- https://www.teamrankings.com/nfl/player-stat/qb-rating-nfl If that site is wrong, then it's wrong.

    Second, I wasn't being literal when I said a few extra TD's to Stills. But if we have to do the math (based on your calculations), an extra 4 TD's per season with three less picks and 300 total yards would get you to the 99.7. After all, you're not going to gain 4 more TD's for zero yards....so this isn't an either/or scenario. Divide all three stats by 16 and you're literally talking about 1 or 2 passes a game falling a different way.

    However, I was trying to get to the 97.6 in the first place (not your 99.7)....which would literally be a few more TD's to Stills, the yards that go with it and a few less picks. My statement was essentially correct...and I'll repeat that it was a guestimate in the first place.
     
  17. djphinfan

    djphinfan Season Ticket Holder Club Member

    111,648
    67,540
    113
    Dec 20, 2007
    Tannehill is fatally flawed imo, so while he may stay afloat in the slightly below average Dept. when it come to qb’ing, he’s 100 percent flawed..and I would never have him as someone who could function his job at the level I believe you need from that position.p to win a ring.

    Non elite Qbs who might have a little better numbers who are not fatally flawed imo would be someone like Matt Stafford or a Derek Carr, or a Dak Prescott, I think their skillsets would not keep me from winning a championship, their skillsets are sufficient at the position.
     
    The Guy likes this.
  18. cbrad

    cbrad .

    10,659
    12,657
    113
    Dec 21, 2014
    Ahh. OK, that site calculates combined passer rating for regular season + playoffs. For an "apples to apples" comparison you want to compare regular season only. It's best to use pro-football-reference if they have the stats. Easy to download too:
    https://www.pro-football-reference.com/years/2018/passing.htm

    And while you're right that even 1 pass each game falling a different way would get you to from #20 to #10, let's remember that these aren't randomly chosen passes we're talking about. These are TD's, INT's and 20+ yard passes we're talking about! How difficult do you think it is to have let's say one CRUCIAL pass falling differently for every game in the season? VERY difficult.

    In fact, that's not what how that difference occurs. It occurs because the QB (or any other component of the passing offense) on one team incrementally but systematically increases the probability of more TD's, fewer INT's and more yards over many many plays. How do we know? The variance would be HUGE if it were due to let's say 1-2 crucial passes per game. QB's would jump around in ranking like crazy because those 16-32 passes have the same effect as small sample size. That's not what we observe.

    In other words, it's really difficult to do what you're suggesting because you have to do it consistently.
     
    Surfs Up 99 likes this.
  19. The Guy

    The Guy Well-Known Member

    6,598
    3,323
    113
    Oct 1, 2018
    But why does just Tannehill get the one or two passes to fall a different way? If he gets it, then everyone else should get it, and then the statistics are just as discrepant as they are now.

    Again, you have to root for Tannehill to be the beneficiary of something, while rooting for other quarterbacks not to be. If you had a much better quarterback, you wouldn’t have to worry about what any other quarterbacks are doing.
     
  20. Sceeto

    Sceeto Well-Known Member

    13,500
    6,244
    113
    Oct 13, 2008
    New York
    :deadhorse::deadhorse::deadhorse: :headwall:
     
    The Guy, Hooligan and Surfs Up 99 like this.
  21. The Guy

    The Guy Well-Known Member

    6,598
    3,323
    113
    Oct 1, 2018
    Part of what I hear you saying here, which I think is accurate, is that those quarterbacks wouldn’t almost certainly produce a loss in the playoffs, as Tannehill would. Obviously a Super Bowl win is contingent on playoffs success.
     
    djphinfan likes this.
  22. Surfs Up 99

    Surfs Up 99 Team Flores & Team Tua

    1,950
    1,785
    113
    May 5, 2016
    Lol! Can’t argue with this. However, it’s the off-season. Since we really don’t know enough about current coaches and team we have to be able to talk about something. :-)
     
    The Guy likes this.
  23. Pauly

    Pauly Season Ticket Holder

    3,696
    3,743
    113
    Nov 29, 2007
    Yet you still feel the need to contribute to the corporal punishment of the deceased equine.
     
    Silverphin likes this.
  24. cbrad

    cbrad .

    10,659
    12,657
    113
    Dec 21, 2014
    Is the horse really dead though? I feel like it's moved even since last season but definitely since 2016 and 2017. I think 7 years of data DOES have an effect. Far more acceptance Tannehill isn't our solution at QB even if the arguments for why may differ. That was NOT the case before the 2018 season. Many posters predicted he'd be a top 10 QB in 2018 because he finally had the right coach and surrounding cast, even saying injuries wouldn't be an excuse (stated before the season). And when it didn't happen.. sure we heard the usual arguments about injuries/surrounding cast, etc.. but something changed.

    The fervor was vastly diminished. I mean this place had a veritable civil war (not very civil sometimes though lol) over Tannehill for many years. Today you see disagreements but that fervor is gone. So I do think 7 years of data and predictions proven wrong has had an effect.

    Maybe we should invent the phrase "beating a dead camel" because unlike a horse a camel might simply refuse to budge despite repeated beatings until it has an utterly compelling reason to move.
     
    Pauly, Bumrush and djphinfan like this.
  25. djphinfan

    djphinfan Season Ticket Holder Club Member

    111,648
    67,540
    113
    Dec 20, 2007
    That is correct and why I use the term fatally flawed, once you see it clearly and you understand how the flaw impacts the game you should move on..theres no working around fatal flaws..
     
    Last edited: Jun 15, 2019
  26. Sceeto

    Sceeto Well-Known Member

    13,500
    6,244
    113
    Oct 13, 2008
    New York
    "When life gives you dead horses, make some horseburgers".
     
  27. KeyFin

    KeyFin Well-Known Member

    10,488
    12,821
    113
    Nov 1, 2009
    I didn't say to give anything to Tannehill- my post wasn't necessarily about him at all.
    Sure, but remember that game Tannehill had 4 TD passes for 300+ yards, but he only threw the ball a total of about 30 feet on the day? I'm thinking it was Arizona where the little slant passes kept turning into 60-80 yard TD's. Those scores weren't RT at all, they were 100% Landry, Hartline, etc....and several here kept going back to that game that season saying it unfairly inflated RT's year-long stats.

    Was it 2015 maybe? I'm sick as a dog today so I'm not going to look it up. But anyway, that was one game that severely impacted his season's stats. Then this past year, we saw a similar game work the opposite way...where he was strip sacked in our own red zone and then threw a pick 6 while getting hit a few minutes later.

    Before that game, I remember you actually questioning if you could be wrong and RT could end up "elite"....and of course we know what followed in the weeks ahead with the December meltdown. The truth of the matter is that if we took back about 8 TERRIBLE plays from that final month and just handed off the ball instead, his stats would be very different on the year.

    And again, this isn't a "RT is amazing" topic since it could apply to any QB. I'm simply saying this because there's less of a difference between 5th place, 10th place and 30th place than people perceive. It's not like we can say the #8 QB is really good but #18 sucks....there's actually very little separating the two.
     
  28. cbrad

    cbrad .

    10,659
    12,657
    113
    Dec 21, 2014
    First of all, let's see if you even believe what you're saying. Andy Dalton's average rank by passer rating in 8 years starting is 16.6. Russell Wilson's average rank by passer rating in 7 years starting is 7. In other words those two QB's are on average ranked 10 apart.

    You contend there's "actually very little separating the two".

    Personally I think that's absurd.

    Second of all, I don't think you understood the post you just quoted because that should put the nail in the coffin of this idea that the primary difference between #7 and #17 is just a few key plays. Yes, theoretically you can make up the difference with "just a few key plays", but the expected variance in QB rankings IF your hypothesis were correct is so large that you'd expect every QB's rankings to span the spectrum over their career, but they don't. Wilson was ranked top 10 in all years except 1. That's not what you'd expect if your hypothesis were correct.

    No, as I pointed out, the difference is likely due to a fundamental difference in average ability of the QB's (or other key component of the passing offense that's remained relatively constant), as evidenced by the Wilson vs. Dalton comparison. Remember as usual that sample size issues come into play here. Obviously in any given year you could argue random variation resulted in what seems like a misleading result, but over longer periods of time your theory doesn't hold water.
     
    Pauly, The Guy and djphinfan like this.
  29. Surfs Up 99

    Surfs Up 99 Team Flores & Team Tua

    1,950
    1,785
    113
    May 5, 2016
    I am not sure if I ever predicted that Tannehill would be a top 10 QB, but I sure hoped he would be. It’s the same with all of our guys. I sure hope that Parker will be a legit #1 WR and Harris a dominant edge rusher. The list goes on. I am probably naive and in some ways I stay ignorant on purpose. The hope is what keeps me going. :-)
     
    djphinfan likes this.
  30. The Guy

    The Guy Well-Known Member

    6,598
    3,323
    113
    Oct 1, 2018
    Precisely. At some point with that line of reasoning you get to the conclusion that there are simply no differences in individual ability among quarterbacks, and variation in their performance is caused exclusively by their surroundings or some collection of random events.

    This is perhaps what happens when you draft a quarterback in the first round and have high expectations for him, and he doesn’t pan out. Your philosophy can shift to attributing causes for quarterbacks’ performances to surrounding variables, and away from factors internal to quarterbacks.
     
    djphinfan and cbrad like this.
  31. djphinfan

    djphinfan Season Ticket Holder Club Member

    111,648
    67,540
    113
    Dec 20, 2007
    I really enjoy how you analyze things and get to the bottom of something you are curious about..
     
    Hooligan and The Guy like this.
  32. KeyFin

    KeyFin Well-Known Member

    10,488
    12,821
    113
    Nov 1, 2009
    It's interesting you picked Andy Dalton since in 2015, he was 2nd in passer rating only to Wilson. So could Dalton play on Wilson's level? 2015 proves he can (ESPN's total QBR had Dalton 5th and Wilson 8th, for what it's worth). Am I arguing that Dalton is on the same level as Wilson? Nope. But on a great day (or season) for Dalton, there's not a ton separating the two....even though they are 10 rankings apart over the course of their career.

    It's also worth pointing out in that 2015 season, the Super Bowl QB's were #7 ranked Cam Newton and #34 ranked Peyton Manning (29th in ESPN's total qbr). In the end, the 2nd worst ranked QB on the season got the ring...not Wilson, Dalton or any of the top 10 QB's.

    To the 2nd part- I don't believe sample sizes come into play when you're talking about potential. Should Eli have statistically won two Super Bowls? No- yet he did. So did Hostetler, Rypien, Warner, Dilfer, Johnson, Flacco, Foles, etc....so did Peyton in his worst season ever. My only point here is that folks saying you have to have a top 10 or a top 5 QB to win are clearly wrong.

    Fitzpatrick could statistically be the top QB in the league next year and lead us to a Super Bowl. While it may be highly improbable, it's POSSIBLE. I don't think it's this massive, insurmountable gap between Fitz and Brady (or whatever name you want to throw in there). I think the difference in QBR and results come just as much from environment (supporting players, coaching, scheme, etc.) as it does from the QB's that are seen as "can't lose" types of players.

    So yes, I genuinely believe in what I stated earlier- the #17 quarterback overall is not light years behind #7. The offensive line plays a factor in QBR as does the run game, your receivers, your play calls, your formations, etc....yet none of that stuff makes it into statistics since it's taken as a constant. To me, all that other stuff is a larger variable than the guy actually throwing the football, and that's why I've always taken QBR with a grain of salt.

    That's how 34th ranked Payton Manning wins a super bowl....with a dominant defense and run game to carry his worn out arm. And hey, he's by far one of my biggest favorites of all time- I'm just making a point here that a quarterback rating doesn't tell the entire story of who a player is and the level he can compete on.
     
    Hooligan likes this.
  33. cbrad

    cbrad .

    10,659
    12,657
    113
    Dec 21, 2014
    Why are you focusing on anomalies?

    Out of 53 SB winners ONLY ONCE was the QB less than one standard deviation below the mean: Peyton Manning in 2015. His z-score was an astoundingly low -1.5863. Average z-score of SB winners is just above +1.

    Dalton's rankings are: 20th, 13th, 15th, 25th, 2nd, 15th, 17th and 26th
    Wilson's rankings are: 4th, 7th, 10th, 1st, 14th, 10th and 3rd

    Obvious even without fancy statistical analysis that Dalton's one great year was a total anomaly. So why focus on anomalies?

    To answer questions like how big the difference is between #7 and #17 you can't focus on anomalies. You have to focus on the "rule" which is best gleaned from the entire distribution of stats. And when you look at the entire distribution it's pretty clear the QB is the main difference: Wilson >> Dalton (the "rule", not the rare "exception").

    You say you think the surrounding cast is responsible. Don't know how you can reasonably argue that. Let me show you how difficult it is by starting with removing the effect of the defense. And we have to use z-scores here, not ranks because ranks don't actually measure anything. Why? Because the difference between rank 1 and 2 isn't a priori the same as the difference between rank 10 and 11, or between any other two neighboring ranks, and as you'll notice that's not a property of any measuring device.

    Dalton has a career z-score of 0.1311 while Wilson's is 1.1217. Remove the effect of the defense (using the best fitting line between points allowed and passer rating) and Dalton's z-score becomes -0.0213 while Wilson's becomes 0.702.

    Both were helped by good defenses though Wilson was helped a lot more. But look at the magnitude of the difference between raw z-score and defense adjusted z-score. Even with Wilson's often elite defense that was 0.41 standard deviations removed. That's a lot but it's WAY smaller than the total you'd have to remove to get the two similar to each other, which is about 0.7. Can't do that by talking about surrounding cast, especially when it's arguably the case Dalton had a better surrounding cast on offense than Wilson.

    So aside from what you WISH were true, what actual evidence do you have that the difference is relatively small? So far you've made your argument based on anomalies. The "rule" however is that the difference is big, and the hypothesis that fits best is that it's due to the one constant in these stats: the QB.
     
    The Guy likes this.
  34. The Guy

    The Guy Well-Known Member

    6,598
    3,323
    113
    Oct 1, 2018
    There is a ton separating the two. When you do an analysis of variance (ANOVA) on their season by season passer ratings, the difference between them is significant at the 0.011 level.

    What that means is that even when both QBs' season to season variation is taken into account (e.g., Dalton's great season you mentioned), there is nonetheless a 98.9% probability that Dalton doesn't belong in the same category as Wilson.

    For QBR the difference is significant at the 0.0079 level.

    If I've interpreted this incorrectly, I trust that cbrad will provide the correct interpretation. The p-level numbers are accurate, however.
     
    Last edited: Jun 15, 2019
    cbrad likes this.
  35. cbrad

    cbrad .

    10,659
    12,657
    113
    Dec 21, 2014
    Yup! Verified that too. Interpretation is that the means of the populations those samples come from are different.

    I'll just add that when you adjust all ratings to a common year, Wilson's 112 ratings and Dalton's 120 ratings are significant at 0.0073, so 0.73% probability they come from the same QB.
     
    The Guy likes this.
  36. Bumrush

    Bumrush Stable Genius Club Member

    29,473
    34,332
    113
    Nov 25, 2007
    The Tannehill discussion should move to the Other NFL forum.

    He’s no longer relevant in Miami and will never be relevant in Miami again.
     
  37. KeyFin

    KeyFin Well-Known Member

    10,488
    12,821
    113
    Nov 1, 2009
    I didn't pick 2015- you did when you wanted to compare Wilson to Dalton....that was the closest comparison year.

    I could have presented that argument a number of ways though. For example, your #1 QB on the year won 10 games. Yet the Bengals, Pats, Broncos, Chiefs, Cardinals, Vikings, and Panthers won more. The Packers, Jets and Steelers tied 10 games- that's ten teams who did as well or better than the team with the top rated quarterback.....that's 31.25% of the league.

    How is that possible that just a hair under a third of the league won more than Seattle with their top D and top QB? The answer is ridiculously simple- QBR is not the only thing that matters in football. There is A LOT MORE to the equation.

    Do I have to explain that more successful teams generally pass less because they're generally playing with the lead (and taking less chances)? Or that strong running teams tend to be high scoring offenses...because of the play action pass? These are not anomalies as you say but basic football 101- the less you ask your QB to do, the more he'll generally have success. Of course, you picked the biggest anomaly in football to begin with in Russell Wilson...but it is what it is.

    Getting back to the original argument though, you asserted that career average #17 overall Dalton could not be as effective as #7 overall Wilson...the difference between them was too great. History proved that false when the Bengals won 12 games and Seattle won 10- that's the only point I was trying to make here.

    Sorry for any typos or anything I messed up- I have a 102 fever. Going to lay back down but I'll argue with you more later!
     
    Last edited: Jun 15, 2019
    resnor likes this.
  38. cbrad

    cbrad .

    10,659
    12,657
    113
    Dec 21, 2014
    Hey.. hope you get better soon! I have nasal polyp surgery on Monday and I react really bad to this prednisone I have to take until then so who knows if my mind is working properly!! Having said that let me respond to your key points.

    Precision matters.. in wording that is.

    As you can see by going back to post #268 where I first introduced the Wilson vs. Dalton argument, at NO point did I choose 2015 nor imply one should look at their closest comparison year. You did that. What I was talking about was average ranking over their careers, so 2012-2018 for Wilson and 2011-2018 for Dalton.

    Similarly, I NEVER implied that "Dalton could not be as effective as Wilson" in a single season like 2015. In fact, I warned against precisely such a conclusion with the very last sentence in that same post #268: "Obviously in any given year you could argue random variation resulted in what seems like a misleading result, but over longer periods of time your theory doesn't hold water."

    As far as the other arguments, of course there's no argument about the contribution of other factors on win%. The question is.. can one provide evidence that those surrounding factors are influential enough to support your argument. But as I just showed you, it's really hard to remove the effect of the defense and then try to argue surrounding cast on offense could make up the difference. The best hypothesis is that the primary difference in those z-score ratings is from the QB.

    And there's nothing really special about Wilson vs. Dalton per se. It's just the first example I found where the difference in rank was about 10 and in approximately the same place you stated (#8 vs. #18) AND where the QB's started their careers at approximately the same time. The same argument would apply to any other pair, at least if we ignore issues with using ranks.

    Actually, let me expound on the problem with using ranks here. Because of how most stats are distributed, the difference between ranks 1-10 will usually be far greater (though not always of course) than between ranks 11-20 because the tails spread out. Then you have peculiarities at the bottom of the rankings where the spread is often even larger because you have backups. Anyway.. not that important but it's worth keeping in mind we need to use the actual ratings instead of ranks, and ideally z-scores.
     
    Last edited: Jun 15, 2019
  39. KeyFin

    KeyFin Well-Known Member

    10,488
    12,821
    113
    Nov 1, 2009
    Let me ask you this very simple, very basic question. What's the point of fans watching a football game?

    If your answer is anything other than "to see my team win", then you're lying. That's all any of us care about. I personally have no idea what ANOVA is (although it made me laugh since I went to Nova High in Davie right across from the Fins practice facility, hence I am A NOVA TITAN!). But what that formula can't tell you is when a pretty good QB like Dalton will outshine a great QB like Wilson.

    The truth of the matter is that you can't use QB stats to predict who will win the most games next season, who will win the SB or even what teams will make it to the playoffs. If that was all black and white then we wouldn't even watch or care...because the Miracle of Miami or the blowout Wildcat game or any of those great memories wouldn't have been possible. Is Ronnie Brown a better QB than Brady? The history books say he is on that particular day. The one true stat that mattered to us was that big fat W.

    My one and only point through this entire multi-page conversation is that you do not have to have a top 4 QB to win it all. It helps a ton, I'm not debating that for one second, but it's a team sport and a team effort all the way across the board. I mean, if Miami did win it all with Fitzpatrick this season but his rating said he was 21st overall, would you really give a crap?

    I sure know I wouldn't...because that's not the most important statistic. And I've given plenty of examples already of seemingly average QB's doing just that.
     
    The Guy and resnor like this.
  40. KeyFin

    KeyFin Well-Known Member

    10,488
    12,821
    113
    Nov 1, 2009
    Prednisone is my mortal enemy. Well, let me re-phrase that. My five acres that were riddled with poison ivy ten years ago was my mortal enemy. I've cut it, dug up roots, sprayed it with all sorts of sure-proof chemicals, cussed at it and everything else I can think of...and we still have some problem areas. So like clockwork, I end up at the doctor 2 or 3 times per year and told I have to take prednisone...which skyrockets my blood pressure and makes it impossible to sleep more than an hour or two a night. I only take it as a very last resort, it's evil stuff!

    Anyway, back to football. I guess the real question here is, "How good is good enough for a QB to take your team deep in the playoffs?"

    We know the #1 overall QB doesn't usually win the super bowl (since Brady is rarely #1 and the bastard keeps winning them). It seems like the opposing QB is generally top 5 but I'm too lazy to back that up right now...I really am about to go lay down until this fever breaks. For the conference championships- what's the average ranking of the 4 QB's still standing? I think looking at the final 4 teams and their QB's overall ranking for the season would give us the biggest preview of how important being in that top 4 really is in the grand scheme of things.

    I'm not sleeping much with this cold so I might do that overnight for the past decade. Good luck with the surgery Monday!
     
    resnor likes this.

Share This Page