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

Tua is not the Problem

Discussion in 'Miami Dolphins Forum' started by Galant, Nov 6, 2021.

Thread Status:
Not open for further replies.
  1. KeyFin

    KeyFin Well-Known Member

    10,488
    12,821
    113
    Nov 1, 2009
    In 2002, Brady was tied for 9th in QB rating with 85.7. Pennington was 1st with 104.2. However, the difference between 9th and 18th was 1.4 points total. 16th (which should be the league average) was 85.2....which is .5 behind Brady. So yes, he was technically "above average" by the slimmest of margins, but that's only because the majority of the league had an 84 or an 85 in passer rating.
     
  2. KeyFin

    KeyFin Well-Known Member

    10,488
    12,821
    113
    Nov 1, 2009
    Short answer? Yes.
     
  3. cbrad

    cbrad .

    10,659
    12,657
    113
    Dec 21, 2014
    Yes I agree, once again what you say is total rubbish. Matt Hasselbeck's career z-score is 0.0168. In other words he's about as prototypical an average QB as it gets from a production standpoint. You don't have a clue what you're talking about.

    z-scores are really good at showing who was above average vs. below average relative to their competition over time.
     
    Kud_II and resnor like this.
  4. cbrad

    cbrad .

    10,659
    12,657
    113
    Dec 21, 2014
    That's not how you calculate relative performance. Rankings obscure actual differences in ability because the difference between ranks N and N+1 isn't the same for different N. The way you do the calculation is to first get the distribution of team passer ratings (to equate sample size, which you don't have for individual QBs) for that year:
    https://www.pro-football-reference.com/years/2002/index.htm

    Calculate the mean. Listed mean for 2002 is 80.4 which is a tiny bit different than calculated mean of 80.22 because they're combining all teams to calculate mean (which is OK as long as that approach is kept constant throughout the calculation).

    Calculate the standard deviation, which is 9.3559.

    Now find out how many standard deviations away from the mean Brady was in 2002. Brady's rating in 2002 was 85.7, so

    z-score = (85.7-80.4)/9.3559 = 0.5665 which is exactly what I listed. Brady was solidly above average at 71st percentile in 2002.
     
    Kud_II and resnor like this.
  5. KeyFin

    KeyFin Well-Known Member

    10,488
    12,821
    113
    Nov 1, 2009
    Where does the calculated mean 80.22 come from? Because Aaron Brooks had an 80.1 rating at #19 overall. Vick was #18 w/ 81.6. I get you have a formula, but the formula feels broken if #18 is slightly above average and #9 is borderline elite....even when 9 thru 15 pretty much had the same score. Wouldn't that mean 15 of the top 32 QB's were elite? Common sense tells us otherwise.

    Pennington and Gannon were elite that season with a 104.2 and a 97.3 QB rating. I'm not sure how you can say Brady was 20 points lower than Pennington yet still somehow close to the same category. Your math is lying to you!
     
  6. cbrad

    cbrad .

    10,659
    12,657
    113
    Dec 21, 2014
    Like literally calculate the mean of the team passer ratings (so all QBs) in 2002. I gave you the link.

    Actually it's 80.2531.. was multitasking.

    No you're lying to yourself. I never said Brady was elite in 2002. I said he was solidly above average that year, which he was: 5.3 passer rating points above average when the standard deviation is 9.4.
     
    Last edited: Apr 5, 2022
    Kud_II and resnor like this.
  7. KeyFin

    KeyFin Well-Known Member

    10,488
    12,821
    113
    Nov 1, 2009
    How am I lying to myself? There's 32 starting QBs. #16 was Jay Fiedler with an 85.2 rating. That tells me 85.2 is average....because he's right in the middle. Half did better than that, half did worse. Brady was .5 better than that Fiedler, not 5.3 points better.

    Again, you don't need an algorithm to determine average and QB's shouldn't get "bonus points" because Harrington had a 59.9 rating at #32. Fiedler was average- no additional math is necessary.
     
    Phil Hutchings likes this.
  8. cbrad

    cbrad .

    10,659
    12,657
    113
    Dec 21, 2014
    Dude.. you can't redefine "average passer rating" to be the passer rating of the #16th ranked QB. There is literally a mathematical definition of "average". And that requires you to take all passer ratings and take the average using the formula. Average is what I said it was: 80.2531 if you calculate literally or 80.4 if you combine across teams.

    And the comment about lying to yourself referred to you suggesting I was saying Brady was elite in 2002 when I explicitly said otherwise.
     
    resnor likes this.
  9. KeyFin

    KeyFin Well-Known Member

    10,488
    12,821
    113
    Nov 1, 2009
    There's 32 starting quarterbacks- how can 18 of them be "above average"? Think about that for a moment. If they aren't rated against each other, then what do the ratings even mean? I am not redefining anything- it's common sense that average means you're right in the middle of expectations among your peers.

    Now, I see where you got the 80.2531 from- that's a league average divided out player by player. There were 11 QB's with a sub-80 rating and 4 of them with a sub-70 rating (plus Harrington at a 59). That's where the formula gives Tom "bonus points"...because the bottom few were simply that bad. On the other hand, there were only 2 really, really good QB's that year. But is that fair? Should we really pretend that #18 is slightly above average because Joey sucked? I think it's ridiculous.

    I love ya man, hope you're doing well, but I have to get back to work. =)
     
    Last edited: Apr 5, 2022
  10. cbrad

    cbrad .

    10,659
    12,657
    113
    Dec 21, 2014
    There were 107 QBs or players in 2002 that threw at least 1 pass and 45 QBs that threw at least 100 passes that year. You're going to just ignore everyone except 32 "starting" QBs. Some years you have more than 32 starting QBs. That's an artificial filter.

    And yes you're redefining the mathematical term "average". There is literally a formula for that. Use it. That's what everyone does, from pro-football-reference, ESPN, Amazon, name it. They don't use your suggestion of "average" = 16 because that makes no sense. You're first of all throwing away tons of data that do go into the average, then not caring about differences in passing attempts, and most importantly not even using the formula.

    Sorry you're wrong, not just about what the mathematical definition of average is, but about Brady being average to below average earlier in his career. Brady was well above average in his early years and elite later on.
     
    Kud_II and resnor like this.
  11. KeyFin

    KeyFin Well-Known Member

    10,488
    12,821
    113
    Nov 1, 2009
    Let's say we're in a 3-man 100 yard dash. You, me, and Galant. We race and you come in 1st, Galant is 2nd and I'm in 3rd. Do we need any additional math here to say who was average?

    You'll say yes. I'd say no, but we'll play this out. You ran the 100 in 10.00 seconds. Galant ran the 100 in 10.50. And I ran the 100 in 15.50 because I tripped and fell on my face at the starting line.

    You'll say the median time was 12 seconds, even though 1st and 2nd place destroyed that time. I'd say 10.5 seconds is average because that's what 2nd place actually ran...was he really faster because I fell down? Let's say I didn't fall and finished with a 9.9 instead. Now Galant is below average; how does that make any sense?

    Now, in this instance I can show you that the average median time for a top athlete running the 100 is 10.3 seconds, because we have tons of outside data to compare that to. Olympians are closer to the 9.5 mark. But for a quarterback, it's an isolated league- there's no outside comparison. Even CFL and arena league QBs don't compare because they play by different rules, etc.

    I get that you're grading on overall comparisons, but I don't feel like it gives you a true definition of "average".
     
  12. cbrad

    cbrad .

    10,659
    12,657
    113
    Dec 21, 2014
    Dude.. you literally have median and mean backwards. Both have mathematical definitions. The median in that case is 10.5 and the mean is 12. You just learned math wrong.

    Also, there is no such thing as an "average median time" when looking at one population.

    Anyway, this isn't a debate. You're just wrong. There are mathematical definitions here you just need to learn.
     
    Kud_II and resnor like this.
  13. resnor

    resnor Derp Sherpa

    16,327
    9,874
    113
    Nov 25, 2007
    New Hampshire
    I've got some ocean front property in Arizona that is an absolute steal. Interested?
     
    Kud_II, hitman8 and KeyFin like this.
  14. resnor

    resnor Derp Sherpa

    16,327
    9,874
    113
    Nov 25, 2007
    New Hampshire
    It seems you completely missed three point of my post.
     
  15. resnor

    resnor Derp Sherpa

    16,327
    9,874
    113
    Nov 25, 2007
    New Hampshire
    Wrong about what? By what metric?

    Brady played at SB winning levels from the beginning.

    Again, how much due to cheating? I don't have the answer to that.
     
  16. resnor

    resnor Derp Sherpa

    16,327
    9,874
    113
    Nov 25, 2007
    New Hampshire
    What a quality, fact filled post.

    I'd argue that your post is rubbish.
     
    mlb1399 and Sceeto like this.
  17. resnor

    resnor Derp Sherpa

    16,327
    9,874
    113
    Nov 25, 2007
    New Hampshire
    Bledsoe never won a Super Bowl. Tua has not demonstrated a Brady level of play. So what makes you think either of those two guys would have won? Who was Brady throwing to that year? Branch? And who else? Troy Brown?
     
  18. resnor

    resnor Derp Sherpa

    16,327
    9,874
    113
    Nov 25, 2007
    New Hampshire
    No, Key, knowing Fielder's rating does Not tell you he's average. If everyone else fell below 75, let's say, then Jay would have been way above average. It tells you nothing without the context of where that fell in regards to other QBs.
     
    Last edited: Apr 5, 2022
  19. KeyFin

    KeyFin Well-Known Member

    10,488
    12,821
    113
    Nov 1, 2009
    In 2002 with Buffalo, Bledsoe's QB Rating was better than Brady's. He also went to the pro-bowl that year. So my short answer is that I think NE could have won with either QB- their defense was just that elite.

    For Tua, I was comparing him from a talent perspective to Brady in 2002....not career Brady. This whole conversation came up on what Brady was like early on, so that's the only thing I was comparing. If Tua was today's age in 2002 though and on that NE team, I think they still would have had a shot at winning. Again though, my answer goes back to that defense more than anything.

    But for Tua specifically, he does not make many mistakes and he delivers the ball on target. With a team around him and his job primarily being, "Don't mess up, just get the ball to the speedy guy quick", he can do that as well as anyone. I just don't think a lot of people can envision what that will look like in a balanced offense behind a competent line.
     
  20. KeyFin

    KeyFin Well-Known Member

    10,488
    12,821
    113
    Nov 1, 2009
    LOL, I did the math right, didn't I? I mean, we got the same answers. I just suck at math definitions apparently.
     
  21. Ohio Fanatic

    Ohio Fanatic Twuaddle or bust Club Member

    32,070
    22,827
    113
    Nov 26, 2007
    Concord, MA
    It was well acknowledged by everyone (except you) that Brady wasn't an elite QB for first few SBs. I don't agree with many that he was just a game manager, but he was clearly a tier 2 QB that won superbowls because of the outstanding team and coaching. Same for Big Ben in his first SB.

    Its funny how people don't see it but then argue against the idea of Tua being able to win games because the supporting cast is way better. Tua has won games - with a good defense, but a historically bad Oline, bad coaching. Why would he not be able to win a lot more games when you dramatically improve his supporting cast? Don't get me wrong, Tua IMO will never be elite, but he is a system QB that needs a good system, coaching and team to win a SB. It's really no different than Drew Brees. Greg Cosell, who has more knowledge than everyone on this board put together, was dead on before Tua was drafted. He stated Tua needed a good Oline and a coach with the right system and he can be a consistent winning QB. Tua has not had that at all to start his career.
     
    Phil Hutchings likes this.
  22. resnor

    resnor Derp Sherpa

    16,327
    9,874
    113
    Nov 25, 2007
    New Hampshire
    C'mon, man. That Bills team was stacked, he had Thurman Thomas go for 1500 that year. And their defense was wicked. That's not even remotely the same situation that Brady had in NE.
     
  23. resnor

    resnor Derp Sherpa

    16,327
    9,874
    113
    Nov 25, 2007
    New Hampshire
    I'm not saying Brady was elite his first couple seasons. But, as cbrad has shown, Brady played at levels equal to the Super Bowl winning QBs. So HIS PLAY was what you needed to win the Super Bowl. That CANNOT be said of Tua. I really don't want to talk about Brady and the ****ing Patriots. They're cheating lying bastards, which is why we'll never actually know how good Brady and the Patriots were.

    I mean, I've resisted calling Brady the GOAT because of that, but I mean, you can't argue with the production with or without superstars on multiple teams, for like a quarter of a century. And I fell like his okay has improved and he's gotten stronger as he's gotten older. I can't deny the dude is an elite QB.

    Also, regarding Tua, I think you and are in agreement on him, from what you just said. I've said repeatedly that I believe Tua will have a decent career. The problem is where we picked him, we could have had a better QB. I think Tua is Ted Ginn 2.0.
     
  24. plc001

    plc001 Season Ticket Holder Club Member

    1,085
    1,758
    113
    Jun 22, 2008
    To be fair, once we start adjusting for era, I start to see the stats as more of a novelty than anything else.

    It’s in the same realm as trying to predict successful nfl QBs using college stats.
     
    Phil Hutchings likes this.
  25. resnor

    resnor Derp Sherpa

    16,327
    9,874
    113
    Nov 25, 2007
    New Hampshire
    I'll say this, the beauty of those Patriots offenses was that it was almost plug and play. They basically only went after players that fit their system, so it was like a well oiled machine. If McDaniels can put together an offense like that, then yes, Tua could be Brady. But I just don't think it's as feasible to do that as people hope.
     
  26. cbrad

    cbrad .

    10,659
    12,657
    113
    Dec 21, 2014
    Well regarding Tua.. first of all I'm glad we have an offensive minded coach and got Tyreek Hill. Tua's in his 3rd year and statistically it's years 3-4 where QB performances tend to start plateauing (4 vowels consecutively in that word!) so this is it. Either Tua's z-scores are in that 0.7+ range (depending on league average rating that's high 90's or low 100's) or I think you look for a new QB.

    Dolphins did something smart: we kept our 1st round draft picks in 2023 which looks like a decent year to draft a QB, so that should be the plan. Tua is either well above average in 2022 or you go QB hunting again. You also don't have to get rid of Tua in 2023 while bringing on another QB since he'll still be on a rookie contract, unless of course he can be traded for a high pick (reasonable actually).
     
    resnor likes this.
  27. Pauly

    Pauly Season Ticket Holder

    3,696
    3,743
    113
    Nov 29, 2007
    There is a long term stable correlation between a team’s NFL passer rating within a season and win% (around 0.65 iirc). Cbrad has run the stats back to the 1970s and I’ve run them independently for the last 20 years.

    What has happened is that as rules have changed and teams have changed their strategies and schemes overall passer rating in the league has been improving. Teams are becoming better at passing efficiently. You could just use rankings, but the problem with that is that the steps between each rank are unequal, so by using z-scores you are measuring how much better or worse than average a team is.

    If you are skeptical I highly recommend running the math for yourself.
     
  28. TheHighExhaulted

    TheHighExhaulted Well-Known Member

    5,819
    4,665
    113
    Jan 15, 2008
    That's incredible considering Thurman Thomas retired in 2000.
     
    Phil Hutchings likes this.
  29. plc001

    plc001 Season Ticket Holder Club Member

    1,085
    1,758
    113
    Jun 22, 2008
    CBrad is smarter than I, and he’s good with his numbers. This isn’t and shouldnt be construed as a slight to him. But it seems to me the logic that could tell me how good a QB from history would perform in todays game could be used to accurately predict college QB success in the NFL.

    The NFL has been throwing millions at that problem. I think some of the same problems would exist in predicting how good Marino would be in todays game. There are too many variables to quantify. Too many assumptions. The type of things that make hasselback (as another poster stated) look like a pro bowler in the data.
     
  30. cbrad

    cbrad .

    10,659
    12,657
    113
    Dec 21, 2014
    Two corrections:
    1) z-scores don't tell you how well a QB in the past would have performed today, or vice versa. They tell you how well they performed relative to their own competition, but in a unit of measurement (i.e., z-scores) that allows you to compare across eras. So you're comparing how impressive QBs were relative to their own competition in a standardized unit of measurement. That's why this is nothing similar to predicting NFL success from college. It's purely a post-hoc analysis.

    2) As I stated, the career z-score for Hasselbeck is 0.0168, so basically average. There's nothing special in his production.
     
    plc001 likes this.
  31. TheHighExhaulted

    TheHighExhaulted Well-Known Member

    5,819
    4,665
    113
    Jan 15, 2008
    Even in you confused Thurman Thomas for Travis Henry, he ran for 1438 yards. Where do you get 1500 yards from?

    Bledsoe wasn't exactly throwing to Hall of Famers either, Eric Moulds and Peerless Price weren't great players.

    How were they that stacked and went 8-8?
     
  32. TheHighExhaulted

    TheHighExhaulted Well-Known Member

    5,819
    4,665
    113
    Jan 15, 2008
    Hasselbeck's "z score" was amongst the highest in the decade of the 2000's.
     
  33. cbrad

    cbrad .

    10,659
    12,657
    113
    Dec 21, 2014
    Do you just make up stuff? I doubt you even bothered to do the calculations.

    Here are his z-scores from 2001-2009 (he had only 19 attempts in 2000 so we can ignore that):

    2001: -0.7680
    2002: 0.7909
    2003: 0.9283
    2004: 0.0229
    2005: 1.6637
    2006: -0.4312
    2007: 0.7597
    2008: -2.5063
    2009: -0.5301

    His 2000's overall z-score weighted by passing attempts is 0.2111 so barely above average. Again, there is nothing special about Hasselbeck even when you try to cherry pick a decade, which one shouldn't do when talking about careers.
     
  34. TheHighExhaulted

    TheHighExhaulted Well-Known Member

    5,819
    4,665
    113
    Jan 15, 2008
    You don't get to choose what's relevant to fit your calculations. I would argue that a decade of someone's prime is more than enough data to judge him.

    Was Hasselbeck not amongst the top of the list in "z score" for the decade? I see top 10 according to my calculator.
     
  35. cbrad

    cbrad .

    10,659
    12,657
    113
    Dec 21, 2014
    "Choosing what's relevant".. yeah you're still cherry picking. And the entire purpose of z-scores is to get AWAY from rankings!!! You don't rank order z-scores. You look at the actual z-score lol.

    Hasselbeck was average overall and only slightly above average in your cherry picked decade.
     
    hitman8 likes this.
  36. OwesOwn614

    OwesOwn614 Well-Known Member

    3,757
    3,787
    113
    Jul 8, 2020
     
  37. OwesOwn614

    OwesOwn614 Well-Known Member

    3,757
    3,787
    113
    Jul 8, 2020
     
  38. OwesOwn614

    OwesOwn614 Well-Known Member

    3,757
    3,787
    113
    Jul 8, 2020

    [​IMG]
     
    Phil Hutchings likes this.
  39. Galant

    Galant Love - Unity - Sacrifice - Eternity

    19,127
    11,058
    113
    Apr 22, 2014
    "This week, ESPN host Mike Greenberg asked the latest variation of the question he routinely poses about the NFL’s 2020 quarterback draft class, a group including the Dolphins’ Tua Tagovailoa, Cincinnati’s Joe Burrow, the Chargers’ Justin Herbert and the Eagles’ Jaylen Hurts.

    But this time, the answers carried more weight because an active seven-time Pro Bowl player was involved in the discussion. Cameron Jordan, the Saints defensive end and member of the NFL’s 2010 All-Decade team, was asked which of these third-year quarterbacks is ready to take the biggest jump.

    Jordan said it’s Tagovailoa, because of the impression he left in Miami’s Monday Night victory against the Saints in Week 16 last season. “After playing Tua, [it’s Tua] because he has more room to grow,” Jordan said. “He’s poised. He doesn’t get rattled. After playing Justin [Herbert], you can hit Justin. He gets a little rattled after you start sticking him early. Tua stays calm. He has that demeanor about him. He’s been hurt more, so he has so much more room to grow.”

    Read more at: https://www.miamiherald.com/sports/...y-jackson/article259823070.html#storylink=cpy"
     
    Phil Hutchings, OwesOwn614 and KeyFin like this.
  40. Fireland

    Fireland Well-Known Member

    1,461
    1,234
    113
    Dec 29, 2013
    [​IMG]
     
Thread Status:
Not open for further replies.

Share This Page