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

Possible Controversy Coming?

Discussion in 'Miami Dolphins Forum' started by Galant, Oct 6, 2019.

  1. AGuyNamedAlex

    AGuyNamedAlex Well-Known Member

    3,582
    2,579
    113
    Sep 12, 2015
    He said "That essentially means there are but six above-average quarterbacks in the league."

    That isnt what that means at all. It's a massive jump in logic. It means QB's who hit those thresholds in a year succeed. It in no way means only 6 QB are capable of hitting those thresholds in the right circumstances.
     
  2. cbrad

    cbrad .

    10,659
    12,657
    113
    Dec 21, 2014
    You’re not listening to what I’m saying. There are no such “thresholds” based on that graph.
     
  3. AGuyNamedAlex

    AGuyNamedAlex Well-Known Member

    3,582
    2,579
    113
    Sep 12, 2015
    The graph clearly shows win % in relation to QB rating correct?

    I'm not speaking of a threshold in the graph, I'm speaking of the threshold in rating (or whatever other method) to determine there are 6 above average QB in the NFL.

    Throwing aside the graph, because my issue isnt related to the data at all and you provided the graph afterwards, the implication of his words is that, there are 6 above average QB.

    So that implies if player X hit his threshold for being in that group of 6, he is no longer above average in general the next year if he doesnt? I'm more asking if I misunderstood his one sentence than anything to do with the graph.
     
  4. Irishman

    Irishman Well-Known Member

    573
    532
    93
    Oct 16, 2017
    High Point, NC
    He doesn't have to be good enough to beat us; he has to be bad enough for us to win.
     
  5. Puka-head

    Puka-head My2nd Fav team:___vs Jets Club Member

    8,605
    6,743
    113
    Nov 25, 2007
    Slightly left of center
    I think you have to worry about more than controversy. Is betting on sports and the NFL legal in FL? If so then a team that could be proven to have "thrown" a game could face legal repercussions. And even if it's not legal in Miami it is in other cities where we will be playing, the jesters home for one. If not legal trouble, someone could get shot!

    And if that were to happen the NFL would have to stomp it flat and make an example of any team that did it, there are BILLIONS of dollars at stake. I could see them suspending a team for a season, taking away all their draft picks etc.

    Better off just playing to win and taking what you get at the end of the year
     
  6. The Guy

    The Guy Well-Known Member

    6,598
    3,323
    113
    Oct 1, 2018
    If I recall correctly, you don’t hit 10 expected wins as a team, which is what I meant by an appreciable increase — that which could reasonably have a team expect to make the playoffs — based on the quarterback’s passer rating until you get to the top six quarterbacks in the league in that regard.
     
  7. cbrad

    cbrad .

    10,659
    12,657
    113
    Dec 21, 2014
    Yeah for 10 expected wins that's about right. You can use that equation in the graph I posted and you get a rating of 105 in 2018 values that's needed for about 10 expected wins (or 102 if you want 9.5 and round up). Since it might not be obvious how to look at years before 2018, note that 105 is 12 passer rating points above 2018 league average so just add 12 to league average in other years to see:
    https://www.pro-football-reference.com/years/NFL/passing.htm

    And you tend to get around 6 QB's.
     
    The Guy likes this.
  8. The Guy

    The Guy Well-Known Member

    6,598
    3,323
    113
    Oct 1, 2018
    So just to reiterate this because I think it’s important in terms of understanding the quarterback position in the present day NFL: at any one time, there are only six quarterbacks in the league who are functioning in such a way that they give their teams the probability of making the playoffs. The other 26 in quarterbacks in the league are functioning in such a way that their teams are unlikely to make the playoffs.
     
  9. Unlucky 13

    Unlucky 13 Team Raheem Club Member

    51,930
    63,007
    113
    Apr 24, 2012
    Troy, Virginia
    That does seem to be really, really hard to believe.
     
  10. KeyFin

    KeyFin Well-Known Member

    10,488
    12,821
    113
    Nov 1, 2009
    Especially when 12 teams make the playoffs every year!
     
    resnor and Irishman like this.
  11. Unlucky 13

    Unlucky 13 Team Raheem Club Member

    51,930
    63,007
    113
    Apr 24, 2012
    Troy, Virginia
    And there are always seasons where quarterbacks have very good seasons individually on teams that fail to make the playoffs, due to the defense, running game or luck!

    Plus stats are highly useful, but never tell the whole story. Especially passer rating, which can get huge boosts from plays where a small easy pass becomes a long gain for a TD. IMO, you always need to look at it subjectively.
     
    resnor and Irishman like this.
  12. KeyFin

    KeyFin Well-Known Member

    10,488
    12,821
    113
    Nov 1, 2009
    As of today, the top 6 quarterbacks are-

    1. Russell Wilson, 126.3
    2. DeShaun Wason, 115.9
    3. Patrick Mahomes, 114.7
    4. Dak Prescott, 106.0
    5. Gardner Minshew, 105.6
    6. Marcus Mariota, 103.0

    So you heard it here first folks, the undefeated Patriots and 49ers won't make the playoffs- neither will division leading Green Bay or New Orleans!
     
    mbsinmisc, resnor and Irishman like this.
  13. The Guy

    The Guy Well-Known Member

    6,598
    3,323
    113
    Oct 1, 2018
    Well, obviously there are other parts of teams, as well, that can contribute to their success or failure.

    The point of the post was that only six quarterbacks in the league at any one time are functioning in such a way that they give their teams the probability of making the playoffs.

    In other words, a team doesn't become likely to make the playoffs due to the functioning of the quarterback position alone unless the quarterback is among the top six in the league in passer rating.

    The significance of this is that a quarterback's being a "top-10 QB" in the league, which is something bandied about quite often, isn't special at all, unless he's also top-six.
     
    Last edited: Oct 9, 2019
  14. cbrad

    cbrad .

    10,659
    12,657
    113
    Dec 21, 2014
    Yeah let's be clear about what that 105 rating => 10 expected wins means. "Expected" in statistics just means "average". In other words, if your QB year after year gets a 105 passer rating in 2018 numbers then statistically speaking you can expect 10+ wins about half the time. That's looking at a vertical line on that graph at 105.

    To see the passer ratings of teams that win 10 games, look at a horizontal line at 62.5% win% on that graph. You see there are a lot of QB's with less than 105 rating that win 10 games. In fact the distribution is skewed towards below 105 because of many other factors that contribute to a team having 10 wins.

    So none of this is counter-intuitive. On one hand, if you want to make it 50% or more likely that you will win 10 games you need a QB with a 105 rating in 2018 numbers. On the other hand, more than half of teams that win 10 games don't have such a QB. Both statements are true and shouldn't be confused with each other. And the reason both statements are true is because most teams don't have QB's with 105+ rating so even if the probability of 10+ wins is smaller for those QB's, the total probability of 10 wins is larger due to the larger number of such QB's (smaller probability + lots more teams = larger overall probability).
     
    Pauly, AGuyNamedAlex and The Guy like this.
  15. The Guy

    The Guy Well-Known Member

    6,598
    3,323
    113
    Oct 1, 2018
    Like I’ve said frequently, to win a Super Bowl you need either an elite quarterback and at least an average pass defense, or an elite pass defense and at least an average quarterback.

    Obviously if you have the elite quarterback and you consequently need only an average pass defense, you have quite a leg up on all the teams that have just an average quarterback and need virtually an entire other side of the ball (the pass defense) to function in an elite manner.

    This is of course why the quarterback position is prioritized so strongly in the present day NFL.
     
  16. AGuyNamedAlex

    AGuyNamedAlex Well-Known Member

    3,582
    2,579
    113
    Sep 12, 2015
    My only point was that the top 6 may change year to year, so going into a season how do you know who the top 6 truly are?

    Most likely if you are a top 10 QB you will finish in that top 6 sometimes and not others.

    So my point is that I understand you need your QB to play like a top 6 but he may not do it yearly. In that case is he good enough?
     
    The Guy likes this.
  17. The Guy

    The Guy Well-Known Member

    6,598
    3,323
    113
    Oct 1, 2018
    This is a good point, and of course individual quarterbacks’ passer ratings vary from year to year. What you notice, however, is that the quarterbacks typically thought to be among the best in the league vary at a level significantly higher than other quarterbacks, in such a way that those better quarterbacks belong to group of their own so to speak, while the other quarterbacks belong to a distinctly different and lesser group.

    And so if you have one of those better quarterbacks, you can reliably expect that your quarterback will be performing in such a way that you need comparatively less help from the other parts of your team to make the playoffs, despite the fact that his performance may vary.

    And if you have one of the lesser quarterbacks, it’s fairly unlikely that his performance will vary in such a way that it puts him among the best quarterbacks in the league in any given year.

    This is all very easily explained by postulating simply that quarterbacks differ in individual ability, and that ability distinguishes them from each other in terms of their performance.
     
    AGuyNamedAlex likes this.
  18. KeyFin

    KeyFin Well-Known Member

    10,488
    12,821
    113
    Nov 1, 2009
    Brady, Brees, Rodgers and a few other dominant names are missing from this year's top-six so far. In fact, four of the top six QB's so far this season don't meet your criteria as "elite legacy players"...and that's with giving you Mahomes as one of those guys in year 2. What you've repeatedly said over and over in this thread simply isn't as applicable in 2019 as it was in 1990 or 2000.
     
    The Guy likes this.
  19. The Guy

    The Guy Well-Known Member

    6,598
    3,323
    113
    Oct 1, 2018
    Give it time.
     
  20. Pauly

    Pauly Season Ticket Holder

    3,696
    3,743
    113
    Nov 29, 2007
    Whilst that is interesting, we should wait 2 or 3 weeks before we start crowning kings. Passer rating is quite unstable below 200 attempts and starts getting more stable around 250 to 300 attempts. Once all the contending QBs have played a few more games we’ll have a clearer understanding of the true picture.

    Also the 49ers and Patsies have very good defenses this year, so that fits into the 50% of teams who make the playoffs do so more for the efforts of the rest of the team than being carried across the line by the QB.
     
    KeyFin and The Guy like this.
  21. resnor

    resnor Derp Sherpa

    16,329
    9,874
    113
    Nov 25, 2007
    New Hampshire
    No.

    Rating is a result not a cause. A QB could be doing what he needs to to do for the team to succeed, but the team doesn't succeed because other parts of the team fail. But that doesn't mean that the QB wasn't "functioning in such a way that they give their teams the probability of making the playoffs.". You guys act like that is a fact, but it isn't. Your stats are predisposed to eliminate QBs on losing teams, without actually grading the QBs actually performance.
     
    Unlucky 13 likes this.
  22. The Guy

    The Guy Well-Known Member

    6,598
    3,323
    113
    Oct 1, 2018
    Now here we go back round and round again about whether passer rating is a measure of quarterbacks’ individual ability. Agree to disagree.
     
  23. Unlucky 13

    Unlucky 13 Team Raheem Club Member

    51,930
    63,007
    113
    Apr 24, 2012
    Troy, Virginia
    I don't think that's Resnor's argument either way. He's simply saying that a QB can be one of the league's very best, while being a part of a losing team, for various reasons.

    Last year, Matt Ryan was in all ways one of the league's very best QBs, both statistically and subjectively. He was good almost all year long, and yet at one point, the Falcons were 4-9, before rallying with some wins vs bad teams at the end. They had the #6 scoring offense, with the #28 scoring defense, and were also 27th in running the ball. As good as Ryan was, he and the passing game simply couldn't overcome that. And you see things like that most seasons.
     
    resnor likes this.
  24. The Guy

    The Guy Well-Known Member

    6,598
    3,323
    113
    Oct 1, 2018
    But that quarterback’s performance was nonetheless associated with at least a 50% probability of making the playoffs that season. The fact that the rest of his team undermined that performance does not change that.

    Again the point is that only the top six quarterbacks in the league achieve such a 50% probability for their teams via their own individual performance, regardless of how the rest of their teams perform.
     
  25. resnor

    resnor Derp Sherpa

    16,329
    9,874
    113
    Nov 25, 2007
    New Hampshire
    It seems to me like you guys are using the result to predict itself.
     
    Irishman and Unlucky 13 like this.
  26. The Guy

    The Guy Well-Known Member

    6,598
    3,323
    113
    Oct 1, 2018
    The predicted variable is win percentage, and the predictor variable is passer rating. Those are two different variables.
     
  27. Unlucky 13

    Unlucky 13 Team Raheem Club Member

    51,930
    63,007
    113
    Apr 24, 2012
    Troy, Virginia
    Either way, I absolutely hate using stats of individuals to show wins (in any sport), and I also dislike trying to use stats to predict the future. For me, its a highly useful tool to show what happened in the past, and to lend insight into why.
     
  28. The Guy

    The Guy Well-Known Member

    6,598
    3,323
    113
    Oct 1, 2018
    If you were in charge of running an NFL team, it would behoove you to use such statistics. If you weren’t, the teams that were using them would have a distinct advantage over you. And it’s precisely statistics’ ability to predict the future, in terms of probability (i.e., there is no perfect predictive ability), that would give those other teams that advantage.
     
  29. resnor

    resnor Derp Sherpa

    16,329
    9,874
    113
    Nov 25, 2007
    New Hampshire
    But passer rating is a function of winning. A good passer rating is a result of successful plays, which requires more than the QB... Winning is a part of win percentage. So yes, you're using the end result to predict itself.
     
    Unlucky 13 likes this.
  30. Unlucky 13

    Unlucky 13 Team Raheem Club Member

    51,930
    63,007
    113
    Apr 24, 2012
    Troy, Virginia
    The extreme usage of stats in this way has sadly made baseball much less enjoyable to watch, even as my own team has been successful lately. Every team is nearly the same, the game has become stale, and most of them play a rather dull style of ball. My hope is that the same does not happen to the NFL.
     
    resnor likes this.
  31. Unlucky 13

    Unlucky 13 Team Raheem Club Member

    51,930
    63,007
    113
    Apr 24, 2012
    Troy, Virginia
    Exactly. Its almost to the point of saying that the team who scores the most points will have more wins. For me, there are so many other things that you can look at below the surface that can tell you which offense has done things repeatably well, and which ones are fools gold.
     
    resnor likes this.
  32. The Guy

    The Guy Well-Known Member

    6,598
    3,323
    113
    Oct 1, 2018
    If that were true, then passer rating and win percentage would correlate perfectly, and they don’t.

    We went through this already, but it doesn’t seem to have registered. Once again, when a quarterback has a high passer rating, in an offense that doesn’t score a comparatively high number of points, what is passer rating measuring?
     
  33. Pauly

    Pauly Season Ticket Holder

    3,696
    3,743
    113
    Nov 29, 2007
    Well you don’t use statistics to predict the future, you use probabilities.

    Statistics is is the art of trying to find out what the probability was after the experiment was run.
    Probability is the art of trying to predict what the statistics will be before the experiment is run.
     
  34. cbrad

    cbrad .

    10,659
    12,657
    113
    Dec 21, 2014
    Wait a second.. whatever other arguments you want to make, this statement is false. A win or a loss occurs at the end of the game, after all the individual components of passer rating have been recorded. Cause and effect is clear here: the components of passer rating come first and THEN comes the win or loss. So no, there's no attempt to predict the result from itself.

    Now.. it is true that passer rating depends on point differential, but it does NOT depend on wins. You have to be technical about this, otherwise you'll be arguing things that are physically impossible for no reason.

    Oh, and one more thing: it's not a priori obvious to what degree passer rating and win% are related. No one knew that until people looked at the stats. Just like many were surprised at how little rushing efficiency contributes to win% compared to passing efficiency (Y/A in both cases). So these things have to be discovered and aren't somehow the result of circular reasoning.

    Of course all that is totally independent of to what degree an individual QB's ability contributes to passer rating. No one really knows so it ultimately boils down to an opinion.
     
  35. resnor

    resnor Derp Sherpa

    16,329
    9,874
    113
    Nov 25, 2007
    New Hampshire
    For example? What is high? It's very difficult to have a high rating without scoring points. If you have any positive QB stats, then you're seeing the result of successful plays.

    You can't understand how passer rating is a result of winning. The conversation can go no further.
     
  36. resnor

    resnor Derp Sherpa

    16,329
    9,874
    113
    Nov 25, 2007
    New Hampshire
    The RATING IS A RESULT. The individual parts are NOT SOLELY THE RESULT OF THE QBS ACTIONS. Winning teams have more successful plays than the opposing team. As such, the QB on the winning team, WILL ALMOST ALWAYS have the higher rating. Therefore, using passer rating to predict win % is essentially using win% to predict itself.
     
  37. cbrad

    cbrad .

    10,659
    12,657
    113
    Dec 21, 2014
    Hmm.. two issues with those statements:

    1) Statistics actually refers to an entire methodology as well as an entire discipline (defined by the methodology) so in many usages of the word it will be true that "statistics is used to predict the future" (happens all the time really). You're referring to something that is more often called "data" rather than something different from the method of analysis (which IS statistics).

    2) Statistics isn't an art, it's math. And probability isn't an art, it's a measure.

    Just pointing this out because it's actually somewhat important in this debate to be technically correct.
     
  38. cbrad

    cbrad .

    10,659
    12,657
    113
    Dec 21, 2014
    Everything except the last statement is true, but the last statement is false. There is a clear cause and effect relationship here you cannot ignore: the events that determine the components of passer rating occur before the win or loss. So no, there is no possible way to argue that the win is predicting the passer rating (or that we are using a "result to predict itself").

    If we were looking at past win% to predict future win% (e.g., the distribution of win% across the league) then yes the "result is predicting itself" because it literally is the same stat. Similarly if we used past passer rating to predict future passer rating (e.g., the distribution of individual passer ratings for a given QB across years) then yes the "result is predicting itself". But using passer rating to predict win% is NOT using a "result to predict itself". This is physically not possible resnor.
     
    Last edited: Oct 9, 2019
    The Guy likes this.
  39. Vertical Limit

    Vertical Limit Senior Member

    12,162
    5,057
    113
    Nov 25, 2007
    Case Keenum got the start... i think we are getting a demoralized Washington team.. they have played all 3 quarterbacks in a ****ed up rotation.. at least Miami knew that Rosen was eventually going to get the start.. washington draws straws on who gets to throw the ball next game..
     
  40. Pauly

    Pauly Season Ticket Holder

    3,696
    3,743
    113
    Nov 29, 2007
    With (2) I was using art in it’s meaning of being a skill, like in phrases such as “the art of war” or “the art of the deal”.

    Regarding (1) saying “statistics is used to predict the future” is a short hand version. I would say that a more correct forming of that sentence would be “statistics are used to determine likelihoods from past events, which are then used to generate the probabilities of future events”.
     
    Irishman and resnor like this.

Share This Page