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. StaleTacos

    StaleTacos Well-Known Member

    1,663
    1,759
    113
    Jan 16, 2022
    It doesn't need to pass through a scientific journal. It just needs to make more sense in evaluations, and it does. So they use it more than passer rating.
     
    resnor likes this.
  2. cbrad

    cbrad .

    10,659
    12,657
    113
    Dec 21, 2014
    How do you know they "use it more than passer rating". Source?

    Or are you making that up? I doubt anyone ignores passer rating for any game. DVOA I'm pretty sure they don't look at as often, but maybe you have the data to show otherwise.
     
  3. Sceeto

    Sceeto Well-Known Member

    13,501
    6,245
    113
    Oct 13, 2008
    New York
    Omg
     
    resnor likes this.
  4. cbrad

    cbrad .

    10,659
    12,657
    113
    Dec 21, 2014
    Yeah that's my feeling too about the constant goalpost shifting by the anti-Tua crowd.
     
  5. StaleTacos

    StaleTacos Well-Known Member

    1,663
    1,759
    113
    Jan 16, 2022
    Passer rating makes little sense in a vacuum. It's far too simple. For example, it overvalues completion % without looking to yards per attempt. I've read IAY/A have a higher correlation to winning than passer rating, but you wouldn't bring that stat up here for the Buffalo game now would you?
     
  6. cbrad

    cbrad .

    10,659
    12,657
    113
    Dec 21, 2014
    I'll take that as an implicit admission you just made up the claim people use DVOA more than passer rating. Passer rating is always used.

    But now you're spewing some more nonsense. What do you mean that passer rating doesn't look at yards per attempt? Y/A is one of the 4 components of passer rating. You don't even know what passer rating is.

    Also, can you link to a site that claims IAY/A has a higher correlation to win% compared to passer rating over large sample size? I've never seen a claim like that before. Let's first see if you can produce any evidence for this latest claim of yours.
     
  7. TheHighExhaulted

    TheHighExhaulted Well-Known Member

    5,819
    4,665
    113
    Jan 15, 2008
    This is as far as it needs to go. It's pretty clear he is just making things up as he goes.
     
  8. StaleTacos

    StaleTacos Well-Known Member

    1,663
    1,759
    113
    Jan 16, 2022
    You got any links stating passer rating is always used in front office evaluations? No one is blinding using them to evaluate a player quite like you. They used advanced statistics like FO (DVOA, DYAR, ect...), and of course have their own advanced statistics departments. No one is just looking up passer rating to evaluate QBs.

    Yards per attempt is not the same as intended air yards per attempt (or CAY/PA). If a QB throws a 3 yard pass and it goes to the house. This matters.
     
    Last edited: Dec 20, 2022
    resnor likes this.
  9. resnor

    resnor Derp Sherpa

    16,327
    9,874
    113
    Nov 25, 2007
    New Hampshire
    Good thing football isn't science.
     
    Sceeto likes this.
  10. cbrad

    cbrad .

    10,659
    12,657
    113
    Dec 21, 2014
    Traditionally, passer rating is the baseline measure for comparing all other measures against, especially new machine learning techniques which btw are far more powerful than DVOA. Of course no one uses only passer rating. I never claimed they did. But that is the historical baseline against which newer metrics are compared. So yes it's "always" used but not solely, obviously.

    lol.. you are so confused. The stats that have a high correlation to win% are AY/A or ANY/A, not IAY/A. Tua is #1 in AY/A and ANY/A, which are both improved versions of yards per attempt. ANY/A's correlation to win% is on par with passer rating, yes.

    But IAY/A is NOT. Fairly low correlation. I'll let you look this stuff up so you can see your own ignorance more (if you care). And no IAY/A doesn't matter much if you go by the evidence.
     
  11. StaleTacos

    StaleTacos Well-Known Member

    1,663
    1,759
    113
    Jan 16, 2022
    Any links to show organizations use passer rating and compare it to actually advanced statistics? Like I said, you are the only person in the world that constantly brings it up. You are about as dated as a dinosaur in your statistical analysis.

    So DVOA rated Tua low AND PFF rated low against Buffalo. Passer ratings actual correlation to winning is .51, which I hate to break this news to you, but that's just a moderate correlation. There are better ways to evaluate QBs. Your tunnel vision of passer rating is exhausting. Tua has the same QB rating as Minshew in his first 20 games. Did you think Minshew was a franchise QB after the 1st 20 games too?
     
    Last edited: Dec 20, 2022
    Sceeto likes this.
  12. cbrad

    cbrad .

    10,659
    12,657
    113
    Dec 21, 2014
    lol.. asking that question just shows you don't have a clue what people do. And I've advanced statistical analysis myself, as in developed new methods. The precise opposite of dated. I just understand the serious limitations of these new methods people are selling as if they're somehow capturing individual ability in a way statistics can't yet. Better to use limited statistics properly than act like you have a stat that really captures the contribution of an individual player (e.g., DVOA and ESPN's QBR).

    Nope.. passer rating's correlation to win% across NFL history is 0.67. The sites you are getting that from did the analysis wrong. They combined multiple years into a single year not taking into account passer rating inflation. Have to do it year by year and then average, or you miss the effect of rule changes. Again you have lesser talent doing this kind of stuff. And a correlation of 0.67 translates to about 45% of variation in win% explained, which is huge for a purely offensive stat.

    Regarding Tua vs. Minshew, Tua was below average over his first 2 years. He's elite this year. I'm not defending his first 2 years.
     
  13. StaleTacos

    StaleTacos Well-Known Member

    1,663
    1,759
    113
    Jan 16, 2022
    Still missing the mark of a strong correlation. Wow. Using data from 30+ years ago for future evaluations. You know the game has changed, right? I'm surprised you don't work in a football front office. Have you applied?

    Oh OK. You are just cherry picking your data then?

    How about Alex Smith? He rocked a 105.4 rating in 2017 leading the NFL, and they STILL drafted Mahomes that offseason. I guess they didn't use passer rating on their evaluation, huh?
     
    Last edited: Dec 20, 2022
    Sceeto, hitman8 and resnor like this.
  14. cbrad

    cbrad .

    10,659
    12,657
    113
    Dec 21, 2014
    The correlation is slightly higher now than before, but it's actually been quite stable. And why would I work in a FO? Pay is less and the research I do now is more interesting (have two jobs, one as a researcher at a top university and another as director of clinical product development at a silicon valley firm).

    Not cherry picking. Tua's 3rd year is significantly different from his first 2. It actually makes less sense to put them together than it does to say something is different in year 3.

    Regarding Smith (rating was 104.7 btw) that was statistically an anomaly if you use statistical tests. So using statistics you would say there's good reason to find a better QB. KC hit the jackpot with that. Not a good attempt at a counter-example of the utility of stats.
     
  15. StaleTacos

    StaleTacos Well-Known Member

    1,663
    1,759
    113
    Jan 16, 2022
    So got turned away huh? Want to show your work? And again, still not a strong enough correlation to winning.

    Correct. You just described cherry picking. Thank you.

    Statistically an anomaly? You mean like Tua's 3rd year compared to the previous 2?
     
    Sceeto, KeyFin and resnor like this.
  16. cbrad

    cbrad .

    10,659
    12,657
    113
    Dec 21, 2014
    What do you mean got turned away? Wouldn't ever consider applying (and never did.. not an appealing job). And for offense alone 0.67 is a huge correlation. Half the game is defense.

    It's not cherry picking when you show something is significantly different. I don't think you know what cherry picking means. It means you are selecting some subset of the data where you cannot show independently that the conditions were different from the rest of the data. But if the passing game is "not the same" from a statistical perspective in year 3 for Tua you can analyze that separately.

    Same type of test yes, but of course with much larger sample size with Smith at that time.
     
    Two Tacos likes this.
  17. StaleTacos

    StaleTacos Well-Known Member

    1,663
    1,759
    113
    Jan 16, 2022
    As in you applied and failed to get a job. And of course, you could have made more money in football if you climbed up in higher levels of the front office. Perhaps social interactions with people held you back.

    Again, please show your data. 0.67 isn't gospel. A strong correlation is .70. And no, half the game isn't defense. There is also special teams which smart organizations weigh heavily.

    Na. That's pretty much exactly what cherry picking is. Players, schemes, and coaches change all the time. The player you are analyzing is the constant. When you discard the other data in the set because "reasons," that's cherry picking.
     
    resnor likes this.
  18. cbrad

    cbrad .

    10,659
    12,657
    113
    Dec 21, 2014
    lol.. no I make many times more than sports statisticians do.

    Data is all from pro-football-reference. I have a program that scrapes their data and analyzes it. And anyone with half a brain in statistics knows that you don't have fixed thresholds for whether something is a "strong" correlation or a "weak" correlation. There are contexts where 0.98 correlation is "weak" (e.g., if you have a model that should theoretically recover the parameters used in a simulation you better be getting 0.999+ correlation) and other contexts where a 0.3 correlation is strong (in many areas of medicine nothing correlates well to patient reported severity of a disease; dry eye is an example).

    Here 0.67 is a very strong correlation because it's only an offensive stat when half the game is defense.

    Nope you simply don't understand. If you can show statistical significance, then that means the effect of all those other changes that normally occur "can't explain" the difference.

    Anyway, I think we're done. You've proven a bit too many times you're willing to make up facts and act like you know stuff you have no clue about. And you're part of the "keep moving the goalposts on Tua" crowd. So we're done here.
     
  19. StaleTacos

    StaleTacos Well-Known Member

    1,663
    1,759
    113
    Jan 16, 2022
    I'm very happy for you. However, I'm guessing you don't make as much as a general manager where interpretation of data sets is needed. You know where data translates to the field and players. Seems to be where your robotic approach to football is most lacking and apparent.

    OK. So...
    1) You can't show your work.
    2) You still don't understand there is special teams.
     
  20. StaleTacos

    StaleTacos Well-Known Member

    1,663
    1,759
    113
    Jan 16, 2022
    [​IMG]
     
    resnor likes this.
  21. resnor

    resnor Derp Sherpa

    16,327
    9,874
    113
    Nov 25, 2007
    New Hampshire
    I say this about people ALL THE TIME. HAHAHAHA
     
    StaleTacos likes this.
  22. DonStrockRules

    DonStrockRules Member

    59
    87
    18
    Dec 19, 2022
    I dont remember Tua, himself, making any bad plays when Miami was up 8 early in the 4th Qtr. The playcalling was absolute **** and the OLine was Useless when the Dolphins tried to run it late.
    McDaniel has looked like a total POS late in all of our losses. His gameplans and total lack of adjustments made it impossible for Tua to find open receivers.
    Against Chargers: Dolphins down 6 pts and get the ball at midfield. We call 3 plays in a row that were doomed pre-snap. That game was SO Winnable!!
    We have All This Speed! Why dont we ever take advantage of it in a 'ball control' way late in these games? And if crossers arent open, that's a tell that tha defense is giving you something else but youre not taking it!
    Last year, Tua was actually really good at using check-downs.

    Once concern I have is that maybe Tua just Doesnt know how to make the right calls pre-snap. Maybe he is there with audible choices or calling out protections and he's just terrible at it. It's like being terrible at RPO reads; handing off when he should keep and keeping when he should hand off. Stuff like that.
    I just have a lot of doubt about Tua's mental acuity.
     
    resnor and StaleTacos like this.
  23. TheHighExhaulted

    TheHighExhaulted Well-Known Member

    5,819
    4,665
    113
    Jan 15, 2008
    Hitman, you're not fooling anyone. Lol
     
    Springveldt and RGF like this.
  24. TheHighExhaulted

    TheHighExhaulted Well-Known Member

    5,819
    4,665
    113
    Jan 15, 2008
     
    The_Dark_Knight likes this.
  25. RGF

    RGF THE FINSTER Club Member

    6,066
    3,436
    113
    Nov 24, 2007
    NY
    I picked up on that right away. Especially after he commented on *his* own post almost immediately. He put me on ignore after I told him he has been exposed.
     
    djphinfan likes this.
  26. The_Dark_Knight

    The_Dark_Knight Defender of the Truth

    11,817
    10,319
    113
    Nov 24, 2007
    Rockledge, FL
    I'm just curious...prior to this season with Tua breaking out, I wonder who was the last quarterback that led the league in all of these stats and what the exact figures were?
     
  27. Springveldt

    Springveldt Season Ticket Holder

    My first thought as well when the first post of a newly created account was in the exact style of an existing member who then comments straight after with "good post". That account got put on ignore straight away.

    After reading the latest posts in this thread I've had to add another to the ignore list as I'm now convinced they aren't discussing things in good faith and they are actually a troll.
     
    RGF likes this.
  28. cbrad

    cbrad .

    10,659
    12,657
    113
    Dec 21, 2014
    It's a bit of a trick question because Pass TD-INT isn't really a stat. That is, there are many different ways to calculate who is "first" in that category (are you taking a ratio, which isn't even mathematically defined if INT = 0, if taking a ratio do you care about total TD or INT, etc.). It's actually cleaner to list INT% instead of some ambiguous Pass TD-INT concept, but they didn't do that because Tua is #4 not #1 lol.

    Anyway, regardless of these asterisks the answer is probably Nick Foles in 2013:

    Rating 119.2
    Pass TD-INT 27-2
    TD/Att 8.5%
    Yards/Att 9.1
    Yards/Comp 14.2

    Note that over the last 10 years there's been about a 5 point increase in average passer rating, so in 2022 that 119.2 is maybe more like 123 or 124. And Foles was #2 in INT%. So Foles is quite a bit more impressive than Tua. However, he only had 317 pass attempts, which is a bit on the low end for a season.
     
    Last edited: Dec 21, 2022
  29. Finatik

    Finatik Season Ticket Holder Staff Member Club Member

    4,323
    4,012
    113
    May 2, 2014
    SO Cal
    I just have to say DAMN. I'm in the med device sector and utilize statistics all the time and have a hard time understanding some of the concept he puts forth. To say something like this is so way over the top. I didn't think you could lose more credibility then you had up to this point. Just stop. Please. You're making yourself look really, really foolish.
     
    Last edited: Dec 21, 2022
  30. hitman8

    hitman8 Well-Known Member

    3,040
    2,485
    113
    Nov 11, 2016
    You guys are ridiculous.I have never in my life created a dummy account.
     
    Last edited: Dec 21, 2022
  31. hitman8

    hitman8 Well-Known Member

    3,040
    2,485
    113
    Nov 11, 2016
    Oh please, never have I created a dummy account. You just can't stand I agree with the guys opinion. Hate to break it to you but there are lot of dolphins fans who are not Tua nut huggers and hate Grier.

    I actually have nothing against Tua except I think he is not an elite talent and has some limitations. I do hate Grier though with a passion. The guy is incompetent and I can't wait to see him go.
     
    Sceeto and resnor like this.
  32. RGF

    RGF THE FINSTER Club Member

    6,066
    3,436
    113
    Nov 24, 2007
    NY
    He has been exposed and he can't handle it. It would do him a world of good to just admit it because it would at least explain the reason for his unknowledgeable, off the wall, idiotic, absurd, attention seeking posts.
    We'll alll have a good laugh and move on.
    Well, were having a good laugh anyway but right now its AT him and not WITH him.
     
  33. Finatik

    Finatik Season Ticket Holder Staff Member Club Member

    4,323
    4,012
    113
    May 2, 2014
    SO Cal
    Can we all just stop attacking each other. It's getting tiresome.
     
    The G Man likes this.
  34. Dol-Fan Dupree

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

    40,533
    33,035
    113
    Dec 11, 2007
    That Foles year was amazing.
     
  35. Two Tacos

    Two Tacos Season Ticket Holder Club Member

    11,121
    5,828
    113
    Nov 24, 2007
    For some reason I've never really thought about how strong that makes that correlation.
     
  36. cbrad

    cbrad .

    10,659
    12,657
    113
    Dec 21, 2014
    To put that 0.67 in perspective, the correlation between points scored and win% is 0.7515. Now.. it's really not the correlation you want to look at it's the square of the correlation because that translates to percentage of variance explained (if there's a linear relationship between variables, which there is here.).

    So 0.67^2 = 44.9% of variation in win% is explained by passer rating (a purely offensive stat), and 0.75^2 = 56% of variation in win% is explained by points scored. Furthermore, 44.9/56 = 80.2%. So about 4/5 of points scored is due to whatever is captured in passer rating. Shows you how little the run game matters in comparison.

    Other tidbits. The correlation to win% for points allowed is -0.7337 which is a bit less in magnitude than for offense, showing that offense is slightly more important in the NFL game than defense. Note that percent of variation in win% explained for both offense and defense are over 50% (0.7337^2 = 53.8%). That's only possible if offense influences defense (same team) and vice versa, and they do. Offense and defense are correlated with each other at -0.3291 (i.e., better the offense the better the defense), and if you square that you see that 10.8% of variation in offense is explained by defense and vice versa.
     
    Last edited: Dec 21, 2022
    Two Tacos likes this.
  37. The_Dark_Knight

    The_Dark_Knight Defender of the Truth

    11,817
    10,319
    113
    Nov 24, 2007
    Rockledge, FL
    Wasn’t my intent for it to be a trick question. I was genuinely curious the last time a quarterback led the league in all of those categories and what that team actually did during the season.
     
    resnor likes this.
  38. resnor

    resnor Derp Sherpa

    16,327
    9,874
    113
    Nov 25, 2007
    New Hampshire
    So Tua could be Nick Foles...did anyone ever think that he was a franchise QB? Lol
     
    StaleTacos likes this.
  39. cbrad

    cbrad .

    10,659
    12,657
    113
    Dec 21, 2014
    Yes, Tua could be Nick Foles. Can't tell what will happen in future seasons, and maybe Tua craps it up later.

    However.. there's no reason to deny Foles' accomplishments in 2013 saying "oh the defenses he faced were overrated" etc like so many are doing with Tua. It's the constant shifting of goalposts that's the problem. Just acknowledge Tua is playing like an elite QB this year and anti-Tua posters wouldn't get so much criticism for their stance.
     
    resnor likes this.
  40. The_Dark_Knight

    The_Dark_Knight Defender of the Truth

    11,817
    10,319
    113
    Nov 24, 2007
    Rockledge, FL
    You’re right, Tua could very well do a complete 180 next season and be a complete bomb…just like Wentz with the Eagles, Goff with the Rams, Stafford with the Rams, Murray with the Cardinals, etc, etc, etc…

    But to reiterate your point, no one can deny Tagovailoa’s accomplishments this season. He leads the league in virtually every quarterback statistic and has shown he can play quarterback at an elite level. For anyone to deny this, they’re either delusional or have an agenda.

    At the beginning of the season, what was everyone saying about Tua? This is a make or break season for him, right? Year 3, new head coach, all of these offensive weapons…if Tua can’t win with all of this, then it was time to move on from him. Everyone remembers that, right? Well he’s done it. He’s done what everyone said he had to do and as expected, it still isn’t good enough.

    Sounds a lot like my first wife…no matter what I did, it was NEVER good enough
     
    Fin-O likes this.
Thread Status:
Not open for further replies.

Share This Page