Luck, Griffin, & Wilson vs. Tannehill & Weeden: An Objective Analysis (Part I)

Discussion in 'Miami Dolphins Forum' started by shouright, Jan 3, 2013.

  1. Bumrush

    Bumrush Stable Genius Club Member

    30,573
    36,272
    113
    Nov 25, 2007
    Yeah, but watching Wilson maneuver in the pocket and extend plays is a far different experience that watching Tannehill.

    Tannehill may be more athletic than Wilson, but sometimes agility and pocket presence counts for more. Nobody will argue that Marino was a physical specimen with shot knees, but somehow he was always able to step away from pressure and make big plays. I would call it a combination of instinct and situational awareness.
     
    2socks and Bpk like this.
  2. Fin D

    Fin D Sigh

    72,252
    43,684
    113
    Nov 27, 2007
    Look, I'm not being cryptic. I don't need the definition of statically significant, I need you to explain why you think 7 more attempts is significant. I don't know why you refuse to answer that simple question, but I suspect there's something in your training in your chosen profession that would explain why a person refuses to answer a simple question.

    Again, 10 boys and 10 girls. You are saying that the average person of these 20 has 1 testicle. Then you're saying the boys all have 2 testicles, so that's 50% above the average which would be statistically significant, and yet all of that is not an accurate portrayal of reality.

    Stop.

    You said stat X could be eyeballed, I eyeballed stat X, then you said I can't do that. That's horse****. Why the hell can't you just admit that was wrong of you?
     
  3. shouright

    shouright Banned

    22,845
    8,861
    0
    Dec 13, 2007
    Hypothetically speaking, if the average deviation from the mean in the sample is let's say only 3 attempts, then a deviation from the mean of 7 attempts is so far from the norm (the mean) that it's deemed as due to something other than chance, and therefore deemed to be statistically significant. That degree of difference from the mean is explained theoretically by something other than random variation in the sample. It varies from the sample non-randomly, and therefore significantly. In other words, there is some inherent "meaning" associated with it.

    Now, on the other hand, if the average deviation from the mean were 7 attempts (or something a whole lot closer to 7 than 3), then of course 7 attempts wouldn't be a statistically significant difference from the mean. It would be only an average difference from the mean, and therefore thought to be due only to random variation in the sample. It wouldn't have any inherent meaning or significance.
     
    DevilFin13 likes this.
  4. maynard

    maynard Who, whom?

    18,425
    6,346
    113
    Dec 5, 2007
    clearwater, fl
    I think FinD is requesting a chart :shifty:
     
  5. maynard

    maynard Who, whom?

    18,425
    6,346
    113
    Dec 5, 2007
    clearwater, fl
    While I cannot assert this with 95% confidence, I believe Wilson to be more athletic
     
  6. shouright

    shouright Banned

    22,845
    8,861
    0
    Dec 13, 2007
    Actually you could probably take a look at their combine numbers and do just that, or at least explore it. There are people here who do that all the time. In fact, CK called Jason Allen the most athletic player in the draft one year, based on that kind of analysis.
     
  7. DevilFin13

    DevilFin13 Season Ticket Holder Club Member

    10,053
    7,111
    113
    Mar 22, 2008
    Many of you are correct that technically we don't know what the coaches were asking Tanny to do. But given the data we have (a significant enough number of pass attempts), I think after a full season we can look back and say fairly confidently that there was a concerted attempt to not throw downfield. I also think it's a fairly safe assumption to say the skills of our receivers had something to do with that, though not everything.
     
    2socks and shouright like this.
  8. shouright

    shouright Banned

    22,845
    8,861
    0
    Dec 13, 2007
    Well we do know there was an effort not to have Tannehill run a whole lot for the bulk of the season, so that he wouldn't get hurt. The coaches have said as much, and I don't believe there's a reason to doubt their truthfulness on that particular issue.
     
  9. Bumrush

    Bumrush Stable Genius Club Member

    30,573
    36,272
    113
    Nov 25, 2007
    There is a difference between running a lot and making plays when the pocket breaks down.. I probably watched 4 or 5 Seattle games and Wilson had this uncanny ability to keep plays alive and make throws. Tannehill showed a few flashed of that, but nowhere near Wilson.
     
  10. maynard

    maynard Who, whom?

    18,425
    6,346
    113
    Dec 5, 2007
    clearwater, fl
    Can I eyeball it?
     
  11. Dol-Fan Dupree

    Dol-Fan Dupree Tank? Who is Tank? I am Guy Incognito.

    40,544
    33,044
    113
    Dec 11, 2007
    While I believe Wilson is much better at escapability, I do not believe that Wilson is more athletic.
     
    gandalfin and Eop05 like this.
  12. Fin D

    Fin D Sigh

    72,252
    43,684
    113
    Nov 27, 2007
    Yet, you're taking the mean of the 5 best QBs and the mean of 5 rookies and telling me 7 attempts a game difference means Luck is a pro and Tannehill is less so. There's a middle ground you're neglecting along with a correlation with reality that is missing as well. 7 attempts a game is not telling us anything significant. You're basically saying its only important because 7 is more than 3.
     
  13. Stringer Bell

    Stringer Bell Post Hard, Post Often Club Member

    44,356
    22,480
    113
    Mar 22, 2008
    I think what would be telling would be the efficiency of those downfield passes the team did make. It certainly is clear that they limited the attempts downfield. However, I'm not sure that limiting those attempts was actually a well-founded decision or not.
     
  14. Stringer Bell

    Stringer Bell Post Hard, Post Often Club Member

    44,356
    22,480
    113
    Mar 22, 2008
    It certainly is greater than one standard deviation.
     
    shouright likes this.
  15. Fin D

    Fin D Sigh

    72,252
    43,684
    113
    Nov 27, 2007
    And that translates to what info? What does it tell us?
     
  16. Sumlit

    Sumlit Well-Known Member

    4,796
    2,760
    113
    Feb 27, 2012
    Miami
    As I said before, this kind of statistic is meaningless. What does it matter than Tannehill had 7 less passes per game than the average top passing QBs? What does it mean that Rodgers had almost the same as Tannehill?

    Those stats give you absolutely no insight or knowledge that pertains to how Tannehill played this season or will play from now on.

    More smoke and mirrors. More fuel for endless futile arguments.
     
    FinNasty likes this.
  17. jim1

    jim1 New Member

    5,902
    3,054
    0
    Jul 1, 2008
    If that's your goal then be honest about it and do it right. Your earlier attempts, including a rather sorry manipulation of Aqua's data for your own purposes, were suspect.
     
  18. Stringer Bell

    Stringer Bell Post Hard, Post Often Club Member

    44,356
    22,480
    113
    Mar 22, 2008
    You can just run a t-test.

    If you run a t-test, you get the following results:

    The two-tailed P value equals 0.0095
    t = 2.7690
    df = 30
    standard error of difference = 3.225

    These numbers, especially the P-value means that the two numbers are statistically significant. Usually anything less than 0.05 would warrant rejection of a null hypothesis.

    *Also worth mentioning that I get 30.25 for Tannehill's average attempts, not 32. Not sure if I missed something, or if Shou did.
     
    shouright likes this.
  19. Fin D

    Fin D Sigh

    72,252
    43,684
    113
    Nov 27, 2007
    I wasn't clear, what does it tell us in the real world? How does this info tell us anything practical about tannehill, the way he was used, his talent level, etc.?
     
  20. shouright

    shouright Banned

    22,845
    8,861
    0
    Dec 13, 2007
    Actually what I'm telling you that, in terms of passing attempts, Luck is non-sigificantly different from the average of the best veterans in the league (i.e., he is not unlike them), while being significantly different from the average of all the other rookies (i.e., he is unlike them) who started for their teams this year. In other words, in terms of passing attempts per game, he played a whole lot more like the best veterans in the league than like his fellow rookies.

    On top of that, Luck had significantly more attempts downfield per game than even the average of the best veterans!

    In other words, he shouldered quite the load for a rookie quarterback, and I think the data show that clearly.
     
  21. shouright

    shouright Banned

    22,845
    8,861
    0
    Dec 13, 2007
    Depends. :)
     
  22. shouright

    shouright Banned

    22,845
    8,861
    0
    Dec 13, 2007
    I disagree.
     
  23. shouright

    shouright Banned

    22,845
    8,861
    0
    Dec 13, 2007
    I cut out the Jets game away. He barely played.
     
    Stringer Bell likes this.
  24. VManis

    VManis Season Ticket Holder Club Member

    6,374
    11,425
    113
    Nov 10, 2010
    I think the sample size here is simply too small to draw any meaningful conclusions. All sorts of game varibles will influence these statistics including how often were they playing from behind, how many overtime games are included in the sampling, how good of a running game the teams have, how good/bad of a pass defense were they facing, what were the weather conditions etc.....
     
  25. shouright

    shouright Banned

    22,845
    8,861
    0
    Dec 13, 2007
    Well it may not tell you, personally, anything.

    What it tells me, personally, is that Tannehill was used like a typical rookie, and not like one of the best veterans in the league. Andrew Luck on the other hand, was not.
     
  26. Fin D

    Fin D Sigh

    72,252
    43,684
    113
    Nov 27, 2007
    I'm with you to here ^....

    ...but here ^ is where the disconnect is.

    Your data doesn't state that. 7 more attempts a game with 1 to 2 deep shots is not a big deal or indicative of anything tangible. I don't think it shows he was asked to do significantly more, nor does it show why he was asked do the little bit more he did.

    I guess what I'm saying his this tells us nothing more than looking at their total attempts which doesn't tell us much of anything.
     
  27. shouright

    shouright Banned

    22,845
    8,861
    0
    Dec 13, 2007
    Actually the sample size is a further strength of the exploration because, despite the small sample size, the differences found were still statistically significant! :)

    Trying to determine statistical significance with a sample that small is an uphill battle, but it was done nonetheless. That further illustrates the strength of the differences involved.
     
  28. Dol-Fan Dupree

    Dol-Fan Dupree Tank? Who is Tank? I am Guy Incognito.

    40,544
    33,044
    113
    Dec 11, 2007
    How?
     
    Fin D likes this.
  29. shouright

    shouright Banned

    22,845
    8,861
    0
    Dec 13, 2007
    Well, believe what you'd like, but IMO if he passed the ball overall significantly more than the average rookie, and he passed the ball aggressively (i.e., downfield) not only significantly more than the average rookie, but also significantly more than the best QBs in the league (!), he shouldered quite the load for a rookie! :)
     
  30. Dol-Fan Dupree

    Dol-Fan Dupree Tank? Who is Tank? I am Guy Incognito.

    40,544
    33,044
    113
    Dec 11, 2007
    What does that say about Matt Stafford? He passed more than anyone.
     
  31. Fin D

    Fin D Sigh

    72,252
    43,684
    113
    Nov 27, 2007
    Again, just because 7 attempts more a game is "statistically significant" in a sample size that has only extremes, it doesn't mean it is in the real world. 2 more shots deep a game isn't significant either. When I'm talking about significant, I'm talking about real world info, like why & how?
     
  32. mroz

    mroz Season Ticket Holder Club Member

    27,931
    28,421
    113
    Oct 26, 2008
    SF Bay Area
    For now I would say talent… Hard to be successful in the NFL (as a QB) with the receivers he had to throw to. Dont even want to get into the lack of running game and so on..
     
  33. shouright

    shouright Banned

    22,845
    8,861
    0
    Dec 13, 2007
    Well, like I said, believe what you'd like. I'm afraid this is probably the point at which we agree to disagree on this particular issue. :)
     
  34. mroz

    mroz Season Ticket Holder Club Member

    27,931
    28,421
    113
    Oct 26, 2008
    SF Bay Area
    Do you know what THill's turn over ratio was compared to other QB's? I am particularly interested in his INT per attempts ratio
     
  35. padre31

    padre31 Premium Member Luxury Box

    99,377
    37,301
    0
    Nov 22, 2007
    inching to 100k posts
    Odd:

    http://stats.washingtonpost.com/fb/tmleaders.asp?type=Receiving&range=NFL&rank=232

    Here we are ranked 22nd overall.
     
  36. Dol-Fan Dupree

    Dol-Fan Dupree Tank? Who is Tank? I am Guy Incognito.

    40,544
    33,044
    113
    Dec 11, 2007
    That would be important.

    From what shouright is trying to say with pass attempts, it seems to me that a more accurate statistic would be, "pre-snap reads", audibles and "post-snap reads"

    Pass attempts does not says much.
     
  37. padre31

    padre31 Premium Member Luxury Box

    99,377
    37,301
    0
    Nov 22, 2007
    inching to 100k posts
    I don't think so, as I watched the games he simply never went down the field until later in the yr, for much of the yr it was all underneath stuff and outs, later in the yr he had more success with the same players, going down the field.

    He also rarely rarely threw a ball into the EZ.

    Shou's stats don't surprise as I've made that point, based on observation, for most of the past year

    IMO, they protected Tannehill for much of the year.
     
  38. RoninFin4

    RoninFin4 Season Ticket Holder Club Member

    24,388
    49,019
    113
    Dec 11, 2007
    Cincinnati, Ohio
    How much of that is due to having 7 comeback wins in the 4th quarter though? I'd bet it's more of a situational thing than something actually game-planned for.
     
    shouright likes this.
  39. Section126

    Section126 We are better than you. Luxury Box

    47,525
    72,482
    113
    Dec 20, 2007
    Miami, Florida
    WE need a football version of the shot / spray chart.
     
    Dol-Fan Dupree and Stringer Bell like this.
  40. shouright

    shouright Banned

    22,845
    8,861
    0
    Dec 13, 2007
    You're getting ahead of me here. That's "Part II." ;)
     
    RoninFin4 likes this.

Share This Page