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

Just how important is "clutch", really?

Discussion in 'Miami Dolphins Forum' started by Pauly, May 30, 2016.

  1. resnor

    resnor Derp Sherpa

    16,329
    9,874
    113
    Nov 25, 2007
    New Hampshire
    Except passer rating is dependent upon the receiver catching the ball. Ergo, what the QB and receiver do is the single biggest factor in explaining team performance.
     
  2. roy_miami

    roy_miami Well-Known Member

    1,385
    560
    113
    Oct 11, 2013
    This 0-7 point win percentage certainly indicates something, I'm just not sure what that something is. Russell Wilson is supposedly clutch (13 comebacks, 18 game winning drives), he also seems to have intangibles, the team around him is supposedly great, great defense and his coaches are supposedly great. But his splits are:
    0-7: 15-16
    8-14: 7-2
    15+: 24-0

    So what gives?
     
    cbrad likes this.
  3. cbrad

    cbrad .

    10,659
    12,657
    113
    Dec 21, 2014
    I actually think the answer is he shouldn't be labeled "clutch" from a purely statistical point of view. Instead, he's just a very good QB but not necessarily "clutch" if you define it the way I've been doing. Here are his splits:
    http://www.pro-football-reference.com/players/W/WilsRu00/splits/

    The stats that would argue he's clutch would be 4th quarter passer rating being higher than quarters 1-3, the very interesting trend of performing ever better as you get closer to January, and as you point out his comebacks and GWD's (very impressive for just 4 seasons).

    But so many other stats aren't impressive or way worse than average, like how his passer ratings totally drop in 0-7 games or <4 minute trailing situations. So maybe this is a case where a statistical argument diverges from what most commentators would say.
     
  4. resnor

    resnor Derp Sherpa

    16,329
    9,874
    113
    Nov 25, 2007
    New Hampshire
    Sooo...if "clutch" is real, but Wilson, a poster child for "clutch", actually isn't "clutch," where does that leave us?
     
    Fin D likes this.
  5. cbrad

    cbrad .

    10,659
    12,657
    113
    Dec 21, 2014
    IMO it's just another (common) example of bias in human perception. The disclaimer to that statement is that the stats themselves might not be great measures of clutch, but I think they're decent enough to make an argument with, especially since no one can directly measure (or perceive) the amount of pressure in different situations.
     
    Pauly and roy_miami like this.
  6. resnor

    resnor Derp Sherpa

    16,329
    9,874
    113
    Nov 25, 2007
    New Hampshire
    Yeah, I just think the bias comes in when people want to believe that certain players are almost superhuman in certain situations.
     
    Pauly and cbrad like this.
  7. roy_miami

    roy_miami Well-Known Member

    1,385
    560
    113
    Oct 11, 2013
    Sometimes the stats test and the eye test are at odds, the difference between me and some others is I'm not so full of myself and my eyes that I just disregard the stats as garbage if they disagree with my eyes. Or vise-versa.

    Having said that I'm not convinced this 0-7 point split is an indicator of clutch. Is Andy Dalton clutch? This 0-7 point stat would say yes, but other stats say no.
     
  8. Sceeto

    Sceeto Well-Known Member

    13,501
    6,246
    113
    Oct 13, 2008
    New York
    You all are making me want to "clutch" a shotgun.
     
    resnor and Fin D like this.
  9. Pauly

    Pauly Season Ticket Holder

    3,696
    3,743
    113
    Nov 29, 2007
    We are talking about things that are far from superhuman, in terms of what 'clutch' can genuinely achieve.

    If 50% of games ar close, and your chance of winning a close game is increased by clutch to 57% (HoF average). If we take 50% of that increase is due to non-QB factors then a QB would win an extra 3.5% of close games.

    For a 10 year career that would be an extra 3 games won over what you would expect an average in the clutch QB to get over that time based on their rating.

    You have to shift the win% in close games up to levels only achieved by 4 or 5 QBs over a career in NFL history to get an extra win a season, and that is accepting all the shift in win% comes only from the QB.

    But if you told some people that having the best clutch performer in NFL history as your QB would get you an extra win per season at best on average they'd look at you as if you had grown a extra head.
     
  10. roy_miami

    roy_miami Well-Known Member

    1,385
    560
    113
    Oct 11, 2013
    Most, if not all good-elite QBs passer rating drops off in 0-7 games. That is actually one of the more clear indications of a good QB. In 15+ games Wilson has the highest passer rating in the league today, and its probably the highest of all time. I don't know where his 88 in 0-7 games would rank, but its probably average as poor QBs/teams tend to have their highest passer rating in that bracket.
     
    cbrad likes this.
  11. xphinfanx

    xphinfanx Stay strong my friends.

    10,823
    2,214
    113
    Nov 1, 2009
    Clutch is very important without it your vehicle is either stuck in gear or won't go into gear unless your lucky enough to be able and grab the next gear by maintaining the correct RPM to engage the gear you need while maintaining momentum.

    [​IMG]


    #1 is the clutch : )
     
    resnor likes this.
  12. cbrad

    cbrad .

    10,659
    12,657
    113
    Dec 21, 2014
    Great point! That changes the interpretation of the 0-7 drop-off.

    Let's start by assuming that passer rating really is measuring QB ability. If that assumption is true, then we should expect on average QB passer ratings in 0-7 games to be similar for both good and bad QB's because you only get a close game if for THAT game both teams (and thus, on average both QB's) played at similar strength.

    To test that claim, I took all QB's that started in 2015 and looked at the career ratings of only those that started at least 8 games for 4 or more seasons (since RW did that.. this btw leaves out Luck because he started only 7 last year). I found 22 such QB's. The average 0-7 ratings for those QB's is 87.5 with a 4.2 standard deviation. All ratings lie between 80.5 and 96.9, so guys like Brady (88.6) or RW (88.3) are performing around average (87.5), same as not-so-good QB's.

    But as you point out, if passer rating does measure QB ability, then of course better QB's will more often have better ratings than their opponent, meaning that most large point spread games occurred when they won by large amounts. That means passer rating for good QB's in high point differential (positive or negative) conditions should be much larger than for bad QB's, which is something you've pointed out before. Oh, and yes you're right that RW ranks last among those 22 QB's in drop-off.

    Importantly, it means the causal effect isn't just "higher pressure" -> "change in performance". That of course exists. What you're saying is that "change in performance" -> "change in pressure", but in different ways for good vs. bad QB's. That is, if for whatever reason a good QB plays worse than average he's more likely to create a higher pressure situation by getting in a close game, while the opposite is true for bad QB's (they help create close games if they play better than their average level).

    The drop-off in 0-7 passer rating includes both effects but is dominated by the effect you're describing, meaning the raw drop-off for 0-7 games is actually a weak clutch stat on its own. This suggests we should do what I did for Brady for all QB's: see what percentage of the time the team that won a 0-7 game was ahead at the end of the 3rd quarter. Your argument also decreases the relative value of looking at drop-off from overall average to <4 minute trailing situations and increases the relative value of looking at <4 minute ahead vs. <4 minute trailing etc..

    Very nice roy_miami. You removed one piece of statistical "evidence" against RW being clutch, so he's more clutch based on stats than I thought yesterday haha, but you still can't ignore the massive drop-off in the <4 minute trailing situations if he's like resnor said supposed to be a poster child for clutch.
     
    roy_miami likes this.
  13. Pauly

    Pauly Season Ticket Holder

    3,696
    3,743
    113
    Nov 29, 2007
    Well this also goes back to something I touched on in my Tannehill -v- Lasor thread.

    There is a difference between passer rating when the tean is ahead/tied and when they are behind. For Tannehill when you look at his splits by game situations in 2014 and 2015 he had very similar stats in each situation, what changed was how often he was behind. It's no surprise to me the Cam Newton and Andy Dalton also had career years when their defenses improved.

    If we look at 2015:
    Leading:
    50.2% rushing percentage
    3.9 yp rush
    92.1 passer rating

    Tied
    43.2% rudhing
    4.0 yp rush
    90.0 rating

    Trailing
    32.3% rushing
    4.3 yp rush
    84.7 rating

    (NB in my intiial Tannehill/Lazor thread I combined tied and ahead)

    So when teams are ahead their rushing is less effective in terms of ypr but their passing is more effective. When they are trailing their ypr increases and their passer rating goes down.
    Occams razor says that when defenses are geared up to stop the run then passing becomes more effective and when defenses are expecting passes that passing becomes more difficult.
     
  14. Fineas

    Fineas Club Member Luxury Box

    18,217
    23,523
    113
    Jan 5, 2008
    None of the stats you cite are for passer rating when down by 21. But I tweaked the search form you linked to for down by 14+ in the last 4 minutes and the NFL average passer rating in that situation in 2015 was 94.1.

    http://www.pro-football-reference.c...location=DL&pass_location=DM&pass_location=DR

    And that is when you include a bunch of backups and scrubs getting garbage time. If you limit it to starting QBs, it comes to 102.3. So yeah, NFL QBs put up big numbers and better passer ratings when down big in the last 4 minutes.
     
    Pauly and resnor like this.
  15. Fineas

    Fineas Club Member Luxury Box

    18,217
    23,523
    113
    Jan 5, 2008
    In an attempt to try to isolate some numbers that I'd be more comfortable calling "clutch," I did a search for the period 2012-2015 for the last 2 minutes and when down 7 or less or up 3 or less. I used down 7 or less to include games when a score would tie it with a regular extra point and up 3 to reflect that it is still a close game and the leading QB still has some "pressure" to not screw up. The numbers are all over the place and do not jibe with most people's perceptions of who is clutch. In fact, they are kinda funny that way.

    The most QBs with the best ratings in these situations are:

    Blaine Gabbert (138.2)
    Cam Newton (117.9)
    Christian Ponder (106.8)
    Teddy Bridgewater (106.1)
    Tony Romo (99.1)

    Gabbert is pretty much a scrub, as is/was Ponder. Romo is often considered a choker. Few people would have Bridgewater on a "most clutch" list.

    Similarly, many of the icons of "clutch" fared quite poorly:

    Tom Brady (65.4)
    Russell Wilson (58.0)
    Joe Flacco (57.0)
    Eli Manning (66.1)
    Matt Ryan (49.3)

    Brady had a huge dropoff in these clutch situations. Whatever Flacco learned when he "learned to be clutch" in the playoffs a few years ago he apparently forgot. Eli's Super Bowl "clutchness" appears to be non-representative. Matty Ice seems to be hot garbage when the pressure heats up.

    Now, to be clear, I do not believe Blaine Gabbert and Christian Ponder are more clutch than Brady, Wilson, etc. Because I don't believe clutch exists. These are all still pretty small sample sizes and random occurrences come into play. But the data certainly does not show that the guys people think are clutch are really better in those late game, pressure situations.
     
    Pauly and resnor like this.
  16. Finster

    Finster Finsterious Finologist

    3,087
    2,038
    113
    Jul 27, 2013
    Do you still have the sheet you made it on, because I would like to see the individual QB numbers.

    Are they true avgs, and what I mean by that is, did most QBs have around an 88 rating, because if that isn't the case, then it's not really a true avg.

    Eleven 0s and eleven 100s comes out to an avg of 50, but that would be a very misleading avg, since none of the 22 were even close to 50.

    Mathematical avgs and actual avgs can be different, in this case, you would have to do 2 avgs, eleven avg 0 while the other eleven avg 100.

    So my question would be, did most of the bad QBs avg in the low 80's while the good QBs avg in the high 80s or above?

    Also, avg that many starts over 4 years eliminates some bad QBs who've lost their jobs, which can also skew the avg.
     
  17. cbrad

    cbrad .

    10,659
    12,657
    113
    Dec 21, 2014
    Yeah, you're ignoring post #614. Better QB's tend to have close games when they play worse, while worse QB's tend to have close games when they play better. I didn't take that into account at first, but after roy_miami's comment, that really drives home the fact that you can't compare overall passer rating with <2 minutes in -7 to +3 conditions. You need to compare overall <2 minute ratings with <2 minute -7 to +3 conditions, and Brady performs about the same there (no drop-off).

    However, in such a comparison for <2 minute stats, you will run into sample size problems. Brady overall has 119 attempts, and in your -7 to +3 condition has 90 attempts. Both are OK sample size wise, but the 29 attempts for all cases outside of -7 to +3 isn't. That is, the variance for 30 attempts will swamp most differences between stats.
     
  18. cbrad

    cbrad .

    10,659
    12,657
    113
    Dec 21, 2014
    Have the data still but not the mappings to the individual QB's (would have to look that up again). Average rating for the 22 is 88.7 with a standard deviation of 7.27. Average rating in 0-7 conditions is 87.5 with a standard deviation of 4.21, and average rating for 15+ condition is 91.8 with a standard deviation of 14.6.

    Here's something real interesting: the correlation between overall ratings for those QB's and 15+ ratings is a whopping 0.956!! That basically means looking at 15+ performance or overall ratings is the same in terms of relative abilities of the QB's.

    Also something interesting: the correlation between overall ratings for those QB's and 0-7 ratings is very high at 0.766. That means that while good and bad QB's are performing at similar absolute passer rating levels in close games (from the previous post), better QB's tend to perform ever so slightly better than worse QB's in close games, which makes sense otherwise they wouldn't win more.
     
    Finster likes this.
  19. Fineas

    Fineas Club Member Luxury Box

    18,217
    23,523
    113
    Jan 5, 2008
    But even when better QBs play worse than their norm, they aren't necessarily worse than the inferior QBs they are playing. Indeed, I believe you and Pauly (or at least one of you) has repeatedly made the claim that in close games both QBs are playing at roughly the same level. I don't believe that, but if you are right about that, Brady's 65.4 rating is still bad. Presumably, even if Brady is having an off-day and finds himself in a close game against an inferior opponent, he would "turn it on" at the end of the game. Except he doesn't, at least according to the numbers. And isn't the failure to rise to the occasion when the pressure is on in a close game the antithesis of clutch? I'm not just talking about Brady's "dropoff" from his normally high levels, I'm taking about his absolute numbers in these situations, which simply aren't good.

    I see no point in looking at his less than 2 minute stats in games that aren't close. That's garbage time.
     
  20. cbrad

    cbrad .

    10,659
    12,657
    113
    Dec 21, 2014
    Yeah, post #620 shows the correlation within close games in favor of the better QB's. So what's happening is that the range of QB ratings vastly decreases in close games, but better QB's tend to play overall better in those games.

    Now in the specific case of Brady, we know where that's occurring. Both for regular season and playoff games, he's ahead at the end of the 3rd quarter in close games at most half the time, yet wins 69% of those games. So Brady does "turn it on" at the end.
     
  21. Fineas

    Fineas Club Member Luxury Box

    18,217
    23,523
    113
    Jan 5, 2008
    None of this is surprising in the least. Indeed, it is what I have been saying throughout this thread.

    QBs pretty much perform at the same levels overall, in close games and in non-close games. I'm other words, game situation and pressure are largely non-factors.

    Better QBs perform better than bad QBs in close games because they are better QBs. They don't elevate their games under pressure at all. They don't show any real indicia of clutchness overall. On a QB by QB basis, as one would expect from any sample, there are guys who in a given sample size will play better than their norm and other guys who will play worse. But as those sample sizes get bigger, I would expect them to normalize to the player's overall ratings.

    And again, the teams win, not the QB. A lot goes into winning that the QB has absolutely no control over. He has little influence on the run game, no influence on his own team's defense, no influence on special teams and very limited influence on coaching decisions.
     
    resnor likes this.
  22. Finster

    Finster Finsterious Finologist

    3,087
    2,038
    113
    Jul 27, 2013
    In regards to "clutch", there is one stat that would be very hard to get, and that is, did the QB make a throw when he had to, or did a QB miss a throw when it was an easy comp.

    Did a clutch QB, while not having a great game, hit on the 3rd downs, and TDs when he needed to, and did the avg QB, who was having a good game, miss in those same situations?

    Those would be very hard to get, that would entail reviewing a **** ton of film, lol, but that is also the crux of the difference in clutch, or non-clutch QBs.
     
    cbrad likes this.
  23. cbrad

    cbrad .

    10,659
    12,657
    113
    Dec 21, 2014
    Well I said it multiple times already, but the data is consistent with multiple hypotheses, including the idea that clutch exists. No way to pin this down until you can directly measure pressure.

    My bias simply comes from thinking professional athletes aren't "special" and are responding to increased levels of pressure similar to most humans. To be consistent with that pressure-performance curve, you're saying pressure may change but the change is small relative to changes in the pressure-performance curve. No way to test this right now.
     
  24. Fineas

    Fineas Club Member Luxury Box

    18,217
    23,523
    113
    Jan 5, 2008

    Yes, the game has changed over the years. But Stabler's career 5.9% INT% is still markedly worse than what you have now shown to be the average for his era. So even compared to his peers in his era, he wasn't a very good decision-maker. I could give several examples of crappy QBs from his era who threw INTs at a much lower rate, but it seems unnecessary in light of the data you provided, which makes my point perfectly well and undermines yours.
     
  25. cbrad

    cbrad .

    10,659
    12,657
    113
    Dec 21, 2014
    Yeah.. you guys are slowly pushing me to write a program that parses play by play so that we can look at every imaginable condition that can be measured with that. Problem with the parsing is I have to parse English phrases, and I just don't want to spend that time right now, but who knows.. maybe in the future I (or someone else) will do it and we can answer your question (and yes, that's the question that needs to be answered).
     
    Finster likes this.
  26. Fineas

    Fineas Club Member Luxury Box

    18,217
    23,523
    113
    Jan 5, 2008

    I haven't gone through to verify your 69% stat, but as I said previously, I think a lot of that may have to do with Belichick, Vinatieri/Gostkowski, other Patriot players etc. You dismissed that out of hand, which I think it absurd. It would be inane to suggest that Brady is wholly responsible for that result. The Patriot defense and special teams, which account for well over half of the Patriots play, are clearly responsible for some of that. And I think it would be ridiculous to say coaching plays no role (or even a small role) in winning of close games. If I remember correctly, you said there were 91 close games in Brady's career. If the Pats won 69% of them, they won 62 of those games. If they had only won 50%, they would have won 45-46. So there's a 16-17 game difference over a span of about 16 years, or about one game a year. Obviously, some amount of variation from 50% would be expected based solely on random chance. I could well be that the Pats had "luck" on their side and by random chance would have won 53-54 games (or maybe more) on that basis alone. That might cut the difference down to 8-10 games over that 16 year period. Having Vinatieri/Gostkowski could easily have been worth 4-6 games over that time period. That would leave 2-6 games left to be answered. Personally, I have no problem believing that Belichick was worth 2-6 wins in a 16 year period. I'd have no problem believing the Patriot defense was worth 2-6 games over that time period. Brady was probably worth some games too, but how many is a question that is hard to answer. But to claim that Brady and Brady alone was responsible for the 16-17 wins above "average" is just ridiculous.

    And again, no, Brady does not turn it on at the end. His 4th quarter stats are the worst of any quarter. And his late, close game stats aren't good.
     
    resnor likes this.
  27. Fineas

    Fineas Club Member Luxury Box

    18,217
    23,523
    113
    Jan 5, 2008
    But professional athletes absolutely are special. They are the best of the best of the best at what they do. And what they do is inherently something that is done under pressure -- tens of thousands of fans, millions more watching on TV, multi-million dollar contracts at stake, etc. So they are already a group that is self-selected based on how they perform under pressure. And they are doing things that involve huge amounts of muscle memory and concentration that they have practiced for tens of thousands of reps over many years. Everything they are doing is in the peak performance portion of the curve because that is what they have trained for and are accustomed to. Now maybe that would change if one upped the pressure to an incredible degree. For example, if the ramifications for a bad outcome in a game or on a given play is death or death of a close family member, that might be a different story. You might see these professional athletes starting to melt under that pressure and perform much differently. But that's not how it is.
     
    resnor likes this.
  28. Fineas

    Fineas Club Member Luxury Box

    18,217
    23,523
    113
    Jan 5, 2008

    You haven't proven anything. I have repeatedly pointed out the flaws in your position and your stats. And I have repeatedly posted stats to support my points. As for the points I made that prompted your BS response, it was not an assertion of fact, it was a hypothetical. There is nothing unusual or crazy about the assumptions, especially considering that I explicitly said there will be some variation from team to team. But if you disagree, why don't you post some stats that prove that, overall, teams that win 75% of their games aren't up after 56 minutes in roughly 75% of their games? Show me the evidence that there is no correlation (or a weak one) between being up after 56 minutes and winning games.
     
  29. cbrad

    cbrad .

    10,659
    12,657
    113
    Dec 21, 2014
    lol.. now you're sounding like some posters that create a strawman argument. Go find where I said ALL of the result was on Brady, or even insinuated that. What I dismissed was YOU acting like Brady had nothing to do with the 69% win percentage in close games when he's ahead at most 50% of the time at the end of the 3rd quarter.

    What's ridiculous is saying Brady's isn't a major factor in that result, so yes Brady does "turn it on" at the end according to that.

    What you have is ONLY a hypothesis. No evidence (as in data that's consistent with one hypothesis but not the other). Just accept that you can't prove what you're claiming. Neither can I.
     
  30. resnor

    resnor Derp Sherpa

    16,329
    9,874
    113
    Nov 25, 2007
    New Hampshire
    Only on this site do we argue that Brady posting his very worst ratings late in games is the Brady "turning it on. "
     
  31. cbrad

    cbrad .

    10,659
    12,657
    113
    Dec 21, 2014
    His drop in 4th quarter rating relative to 1-3 quarter ratings is exactly average. We've already shown why the drop from overall passer rating to <4 minutes or <2 minutes isn't a strong clutch stat because good QB's are in close games when they play worse than average, so you have to first condition on the game being close. And when you do that, you get the 50% to 69% result where it's ridiculous to claim Brady's wasn't a major factor in the result. Yeah, that's evidence of "turning it on".
     
    jdang307 likes this.
  32. resnor

    resnor Derp Sherpa

    16,329
    9,874
    113
    Nov 25, 2007
    New Hampshire
    Dude, you're smart, and way better with numbers than me. However, having your play decline, regardless of how it compares to other people, in crucial moments, is absolutely NOT turning it on.

    Also, you have yet to show that defense, coaching, and special teams weren't the major factor in those wins. Vinatieri won several big games at the end with field goals, I don't know how many "regular games" he was responsible for winning.
     
  33. resnor

    resnor Derp Sherpa

    16,329
    9,874
    113
    Nov 25, 2007
    New Hampshire
    Also, in the last 12 years, Brady has only had a season rating below 90 three times. His other the below 90 were his first three years starting.

    How does his drop look discounting the first three years in the basis of developing still?
     
  34. jdang307

    jdang307 Season Ticket Holder Club Member

    39,159
    21,798
    113
    Nov 29, 2007
    San Diego
    Because Vinatieri led the team to put the ball into field goal range, right?
     
    cbrad likes this.
  35. resnor

    resnor Derp Sherpa

    16,329
    9,874
    113
    Nov 25, 2007
    New Hampshire
    Because throwing screens to Faulk and 5 years slants to Welker, and having them turn them into 10-15 yard gains is about Brady, right?

    And you realize we're talking about the win, right? If Vinatieri kicks the game winner, he gets credit for the win, not Brady. Just like if Brady throws a td, Vinatieri doesn't get any credit.
     
  36. cbrad

    cbrad .

    10,659
    12,657
    113
    Dec 21, 2014
    Hey.. all I'm asking is that you give the QB a decent amount of credit. Doesn't even have to be the majority of the credit. But don't act like he should get little to no credit.
     
  37. Finster

    Finster Finsterious Finologist

    3,087
    2,038
    113
    Jul 27, 2013
    I'm sure you also noted that my second sentence was heavily influenced by the interesting theory that good QBs are playing sub par in close games and that avg and bad QBs are playing well in close games that you came up with, which also makes a lot of good common sense, once presented.

    I love those type of situations, where once laid out, it makes a ton of common sense, but all that good common sense managed to stay hidden, until extensive research "stumbles(no slight meant here)" upon it.
     
    cbrad likes this.
  38. resnor

    resnor Derp Sherpa

    16,329
    9,874
    113
    Nov 25, 2007
    New Hampshire
    Well, this started out discussing "clutch," not who gets credit. But, giving some credit is far, far different from "major factor," or assigning most of the credit for wins to him.

    I agree, he gets credit for moving the ball downfield, as do the receivers and backs, who often did the most work.
     
  39. resnor

    resnor Derp Sherpa

    16,329
    9,874
    113
    Nov 25, 2007
    New Hampshire
    Couldn't it just as easily be the average QB is playing average, while the elite QB is playing badly? If one QB is playingplaying badly, the other guy just has to play decent.
     
  40. Fineas

    Fineas Club Member Luxury Box

    18,217
    23,523
    113
    Jan 5, 2008
    What you said was "And whatever the contributions of the coach, defense, kickers, etc.. there is NO way you cannot give a big dose of credit to the QB for this 50% to 69% jump." Yet, you have no evidence to show that any of it was Brady, let alone a "big dose." I don't doubt that Brady was a part of it, but I don't know that it is a big dose at all. As I explained, random chance could easily explain 8-10 of those 16-17 "extra" wins. The kickers have been big. During Brady's career, Vinatieri and Gostkowski have combined to make 11 of 12 FGs (91.6%) from 40 yards or more in the last 2 minutes of 7 point games. That's pretty big. During that same time period, non-Patriot kickers made 62% of such kicks. If the Pats had an average kicker (or one making those at the rate the rest of the league's kickers made them), they would have missed 3-4 of those kicks and likely lost those games. During Brady's career, the Pats have allowed an opponent's passer rating of 63.4 in the last 4 minutes of +/- 7 pt games. Ya think the Pats' defense may have played a big part in winning some of those close games? Etc.

    So we don't know how much of a factor Brady is/was. But we know he wasn't alone responsible for that 69%/50% differential. Or even close to it.
     
    resnor likes this.

Share This Page