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

Shocking Danny stat

Discussion in 'Miami Dolphins Forum' started by dirtylandry, Jul 6, 2017.

  1. cbrad

    cbrad .

    10,659
    12,657
    113
    Dec 21, 2014
    Yup should be double benefit. So either Peyton was worse than he should have been at home or better than he should have been away. Which one is accurate I can't tease apart.

    Oh and for reference, Peyton's final dome vs. outdoor difference was 5.1 points (that link I used did its analysis from 1998-2013.. so the numbers Finster quoted are the updated ones: 99.1 for dome and 94 outdoors):
    https://www.pro-football-reference.com/players/M/MannPe00/splits/
     
    Last edited: Jul 15, 2017
  2. cbrad

    cbrad .

    10,659
    12,657
    113
    Dec 21, 2014
    Yes what you say about domes is correct. That was my point #1. You're forgetting point #2, which is that dome teams are at an extra disadvantage playing outdoors relative to outdoors teams playing in domes. That's borne out very clearly in the win% stats.

    So even though we're interested in a single QB you need to do it the way I did it (or something similar). And I'm counting 17 away games Peyton played in domes or retractable roofs out of 130 total, so you can see how confounded those stats are with home vs. away.

    Just so this is clear, I'm interested in this because I need to know whether I need to adjust the stats posted in post #145 for Peyton (or Brees if we add him) based on dome vs. outdoors, and it looks like I don't need to.
     
    Last edited: Jul 15, 2017
  3. jdang307

    jdang307 Season Ticket Holder Club Member

    39,159
    21,798
    113
    Nov 29, 2007
    San Diego
    If he had his numbers in a dome away might be another data point if the sample size isn't too small
     
    cbrad likes this.
  4. cbrad

    cbrad .

    10,659
    12,657
    113
    Dec 21, 2014
    Yeah.. like I just told Finster I counted 17 away games out of 130 total in domes or retractable roofs (have to count retractable roofs because Indy changed to that in 2008), so sample size is way too low.

    EDIT: I mistook games for attempts lol. OK, there may actually be enough data for this. Let me take a look.

    OK jdang good catch! Also, turns out there are 21 away games in domes or retractable roofs, not 17. I forgot about St. Louis and somehow missed one more elsewhere.

    Anyway, Peyton's away passer rating in domes or retractable roofs is 103.6, which is a bit above where it should be if you go from outdoors rating to dome rating. Many ways to interpret this (never edited a post this many times LOL) so I'll just leave the interpretation up to you haha.
     
    Last edited: Jul 15, 2017
  5. danmarino

    danmarino Tua is H1M! Club Member

    15,360
    20,983
    113
    Sep 4, 2014
    People who use wins for individual performance also know little to nothing about football. Peyton's and Brady's postseason stats are nearly identical even though Brady played on much better teams...and cheating.
     
  6. danmarino

    danmarino Tua is H1M! Club Member

    15,360
    20,983
    113
    Sep 4, 2014
    Again, Brady and Peyton's stats, in the post season, are nearly identical.
     
  7. danmarino

    danmarino Tua is H1M! Club Member

    15,360
    20,983
    113
    Sep 4, 2014
    I couldn't find stats over his entire career, and I don't have the time to work them right now, but there are plenty of seasons (and I didn't look at them all) were Brady's home vs away passer rating was 4 or more points difference. (I saw a few seasons over 10 points difference) Also, Brady's passer rating on grass and turf shows a big difference. And I think the Pats changed from natural grass to artificial turf sometime in 2005.

    Dome TEAMS do tend to lose more when playing outside, but I don't think the 99 to 94 passer rating difference of Peyton was the reason the Colts lost more.
     
  8. resnor

    resnor Derp Sherpa

    16,329
    9,874
    113
    Nov 25, 2007
    New Hampshire
    Wouldn't it be better to look at outdoor teams that Peyton faced twice a year? See what his rating was against them at home and away?
     
  9. Finster

    Finster Finsterious Finologist

    3,087
    2,038
    113
    Jul 27, 2013
    Those are career numbers for Peyton, Brady's career numbers are 98.1 at home and 96.3 on the road, a 108.9 rating in domes, a 100.4 rating under the roof and a 96.6 rating outside.

    Peyton's home and away, 99.8 at home, 93.3 away.
     
  10. cbrad

    cbrad .

    10,659
    12,657
    113
    Dec 21, 2014
    Yeah.. same kind of suggestion jdang made (that I for some reason didn't even consider). It's interesting info, and I wouldn't mind getting it, except for the fact that this kind of info simply cannot tease apart whether Peyton was better than average away vs. worse than average at home, or worse than average in domes vs. better than average outdoors, etc..

    There will always be at least two different hypotheses that fit the data so I'm not yet inclined to go get the data you suggest because while it's nice to have an extra data point, it can't resolve the question of where Peyton was better/worse.
     
  11. resnor

    resnor Derp Sherpa

    16,329
    9,874
    113
    Nov 25, 2007
    New Hampshire
    Seems to me it would be clearer, looking at the same opponent, and looking at how he did at home in the dome vs away ay an outdoor stadium. Looking ay all opponents brings other factors in, like quality of team, etc. Looking only at non-dome teams that they played both at home and away would simplify it.

    In my simple mind, at least.
     
  12. cbrad

    cbrad .

    10,659
    12,657
    113
    Dec 21, 2014
    Because every home game is a dome game and every away game is an outdoors game you can't tease apart whether it was a "home" vs. "away" effect or "dome" vs. "outdoors" effect with what you're suggesting.

    Regardless, it's not that difficult to get the data lol so here it is. Keep in mind I can only use stats while Peyton was at Indy, excluding his time at Denver. First 4 years are against AFC East and last 9 years are against AFC South.

    Peyton's combined passer rating (so you add up all component stats and then compute a single passer rating.. this is how they do it for his "career" stats so it's the same methodology) for home/dome games vs. divisional opponents was 98.9 and his away/outdoors games was 89.7.

    This is a bigger difference than his overall career dome vs. outdoors games, which makes sense given that some of those dome games were away and some of the outdoors games (while in Denver) were at home.

    btw.. this approach might help answer one type of question (what happens when you keep opponents constant) but adds another confounding factor lol: division games tend to be tougher. That is, the home team tends to win a higher percentage of non-divisional games than divisional games (football outsiders says that from 2002-2013 division home win% is 55.6 and non-divisional home win% is 58.6 and non-conference games home win% is 59.9%). But that's another story.. anyway there's your data.
     
    resnor likes this.
  13. resnor

    resnor Derp Sherpa

    16,329
    9,874
    113
    Nov 25, 2007
    New Hampshire
    Yeah, I just prefer to see whay his difference was against the same teams. It brings a couple other factors in, but to me, it makes more sense to look at that data. Like, I don't necessarily care about his rating against one outdoor team that he played once. Doesnt seem as god as knowing that against a team at home in a dome he three for a 113 and that same team away, outdoors, he threw for an 88.

    Would be interesting to see his stats in away dome stadiums. Then we could compare to away outdoor stadiums.
     
  14. cbrad

    cbrad .

    10,659
    12,657
    113
    Dec 21, 2014
    Oh that data I have handy (did this already for post #164). Peyton had a 103.6 rating in away games in domes, but only a 90.2 rating in away outdoor games. This is consistent with his overall 93.3 away passer rating because those are mostly outdoor games.

    Anyway, maybe now you understand why I had that convoluted response in post #164: these stats show that the effect of "dome" vs. "outdoors" isn't consistent in "home" vs. "away" conditions, meaning you can interpret this in many different ways. I'll let you choose the one you want haha (as long as it's consistent with the data).

    Also a disclaimer: it's easy to get the data but hard to know whether the game was played indoors or outdoors, especially with retractable roofs. So when I say I found 21 away games in domes, that's my best guess and I might be off by 1 or 2.
     
    resnor likes this.
  15. jdang307

    jdang307 Season Ticket Holder Club Member

    39,159
    21,798
    113
    Nov 29, 2007
    San Diego
    You just said Ryan Tannehill knows nothing about football.

    I thought you were a fan of his. I guess I was mistaken

    People who only looked at career stats don't know anything about football.
     
  16. danmarino

    danmarino Tua is H1M! Club Member

    15,360
    20,983
    113
    Sep 4, 2014
    Let me guess, RT has said something along the lines of "wins make good QB's"? Or something similar? If you believe him, or any QB when they say that, then I have a bridge I want to sell you. And jeez!...there you go bringing up RT again. Can't you keep him out of your mouth? I bet he's in there more than he is in Lauren's. ;)

    Comparing career stats, especially amongst players who played in the same era, is extremely telling in regards to performance. Of course there are a few other things, longevity, etc. a person could use, but team wins isn't one of them.
     
  17. Finster

    Finster Finsterious Finologist

    3,087
    2,038
    113
    Jul 27, 2013
    Wins are a part of it, the better the QB, the better your chances of winning, but wins should be at least commensurate with team skill, if not, the QB will lose his job, and if you win more than what your team talent would suggest, you're probably going to get a big fat contract.

    So wins do factor in, but wins have to be gauged vs team talent.
     
  18. resnor

    resnor Derp Sherpa

    16,329
    9,874
    113
    Nov 25, 2007
    New Hampshire
    Yes, one time Tannehill said he wanted to be judged on wins.
     
  19. cbrad

    cbrad .

    10,659
    12,657
    113
    Dec 21, 2014
    I've said before there's no good way of teasing apart how much the QB contributes to wins, but you can figure out how many wins per season you need to explain either by: 1) strength of the rest of the team, or 2) greater than average influence of the QB on wins.

    Which of the two, or which combination of the two, is actually going on one can't figure out. Still.. having an idea of how MANY wins you need to explain through #1 or #2 is useful. The way to do it is to compare actual win% to expected win% for each season given the best-fitting line between passer rating and wins. This is what you get for average over a career:

    Brady-led offense = +2.24 wins above expected
    Peyton-led offense = +1.02 wins above expected
    Marino-led offense = +0.53 wins above expected
    Montana-led offense = +0.92 wins above expected
    Rodgers-led offense = -0.58 wins below expected
    Young-led offense = -0.65 wins below expected

    Ascribing all those wins above/below expected to the rest of the team fits for maybe 2/3 of the data but not all. You'd expect Brady to have a high number because the team around him is great and Marino much less, but why is Young negative?? Either Young's team wasn't that good or Young for some reason just was less influential on win% than an average QB. OK maybe that narrative fits, but then Rodgers?? Was his team THAT bad? Or are we overestimating his influence despite great stats?

    Anyway.. you get the idea.
     
  20. Finster

    Finster Finsterious Finologist

    3,087
    2,038
    113
    Jul 27, 2013
    How did you determine expected wins?
     
  21. cbrad

    cbrad .

    10,659
    12,657
    113
    Dec 21, 2014
    Take the passer ratings of all starting QB's in a year, plot them against win% adjusted to a 16 game season (so if a QB started 14 games and won 10 the calculation is 10/14*16 = 11.43), then find the best-fitting line through the data points.

    So on the x-axis you have passer rating and y-axis you have expected wins over a 16 game season. Look on the x-axis for the passer rating your QB got in that year and see what point on the best-fitting line it corresponds to. Now take the difference between the two (actual number of wins - expected on the best-fitting line). The average of those differences across a career is what I posted.


    The most surprising result is Rodgers. You look at the z-scores I posted in post #145 (or more easily seen in jdang's repost in post #150) and the statistical GOAT based only on those z-scores is probably Rodgers, which jibes with intuition. But this analysis shows that his play isn't translating into wins, unless you assume he doesn't have much influence on win% or he's just been on a well below average team for most of his career. Neither seems easy to accept, but maybe his play doesn't translate as well to wins.. who knows.
     
  22. jdang307

    jdang307 Season Ticket Holder Club Member

    39,159
    21,798
    113
    Nov 29, 2007
    San Diego
    Not in the playoffs when you have sudden death. Manning feasted on the first round or two and throws up a 30 rating, or a 50 rating and they get knocked out. We've gone through this and I think cbrad crunched the numbers.

    If he has a perfect game in the wildcard and then another good one in the divisional round and throws up a 30 burger his passer rating will look good. Someone who doesnt know much about football will say, yah but he had a 100 rating in the playoffs! Not his fault!

    In 2003 he had a 106 rating across three games, 65% completion, 8.92 ypa. That's stellar.

    But then you see in his first two games he had a 158.3 rating, a 138.7 rating and then a 35.5 rating and 1 td, 4 INTS, and a 5.04 ypa.

    2004 a 145 rating then 69.3.

    With only 27 games to analyze its beyond lazy to just look at passer rating.

    Against the chargers in 2007 he had a good rating. 97.7. 68% completion, over 8 ypa. 3/2.

    Those two INTS? A pick six in SD's territory. Then another pick deeeeeep in SDs territory. They lost by 4. Defense gave up only 21. Manning gave up 7 and their offense scored 24.

    Just by looking at passer rating you'd believe it wasn't Mannings fault.
     
    Finster likes this.
  23. Dolphin North

    Dolphin North Well-Known Member

    366
    387
    63
    Apr 30, 2017
    I just watch the guys play the game mostly. Stats are something I look at from time to time but I just watch them play now. That's not to diminish what you did here, I actually appreciate it and I like looking at the numbers to see how they line up with what I feel when I watch players play. Numbers do matter too and I recognize that they need to be looked at. I'm actually a math guy, just not as completely so when I watch sports. Nice job.
     
    cbrad likes this.
  24. Finster

    Finster Finsterious Finologist

    3,087
    2,038
    113
    Jul 27, 2013
    What I don't see here Brad is the state of the team each player plays on, because that's the only real way to give out expected wins.

    For example, the Pats should have 12 wins as an expectation, but the Browns should have 4 wins as an expectation, the expected wins of any team rely heavily on the rest of the team.

    The Pats win about 12 games a year, so how is Brady rated so high?

    A-Rod and Dan have taken FLAWED teams to the playoffs, that had no business in the playoffs, I would say the those two guys are above 2 games per season.
     
    resnor likes this.
  25. cbrad

    cbrad .

    10,659
    12,657
    113
    Dec 21, 2014
    Yeah the point is we don't know what the expected wins of a team is independent of the QB so you have to do this differently than just taking the average wins of that team (confounded by the QB).

    Here's a graph of Expected Wins vs. Passer Rating for 2016:
    [​IMG]

    Each blue dot is the number of wins that QB had over 16 games or would have likely had if they played 16 games. Brady is the upper-most dot there where passer rating was 112.2.

    The red line is the best-fitting line to those data points and the equation is in black. As you can see the expected number of wins given Brady's passer rating is 11.5 wins while Brady's expected wins over 16 games in 2016 is 14.67.

    That's one year. Do that for each year for every QB on that list and look at the average difference and you get the numbers in post #179. At least this approach tells you what you should expect given the average influence a QB in that year would have on win%.
     
  26. Finster

    Finster Finsterious Finologist

    3,087
    2,038
    113
    Jul 27, 2013
    I understand Brad, and this is good work, but this is expected wins, *based on QB rating, it's not expected wins of the team.

    Because of Rodgers QB rating, he is expected to win 12 games a year, but he doesn't play on a 12 wins a year team, and because of Brady's rating he is expected to win 10 games a year but he is on a 12 win team.

    So you can clearly see how this chart doesn't reflect reality, and this is without getting into opponents, it's easy to make a logical assumption that SD would have won more games had they played in the AFCS, but they don't, so the team win expectations should reflect that.

    It's subjective, no doubt, but when you see a team that is lacking in talent, competing for playoffs every year and the QB is a world beater, the answer is pretty obvious.
     
  27. cbrad

    cbrad .

    10,659
    12,657
    113
    Dec 21, 2014
    It does reflect reality. The question is what the reason is for the discrepancy. And like I pointed out in post #179 there are multiple interpretations. The question is which one is correct.

    You asked why Brady is so high. That should be obvious. He plays on a great team. That red line tells us how many wins a QB should have if he played on an AVERAGE team with whatever passer rating you want. Brady plays on a way above average team so it makes sense his differential is high.

    In fact, for all except Rodgers the numbers make sense. Marino is relatively low suggesting he played on a worse team than Brady or Montana, etc.. assuming of course these QB's had similar influence on win%.

    As I said initially, Young MIGHT be explained by pointing out that while he had a great team, he was such a "cog in the machine" that his average influence on win% was no better than the average QB (NOTE: there's a distinction between average influence on win% and whether you're an above average QB.. those are two different things).

    The only real mystery is Rodgers. It's either: 1) he's playing on a well below average team with a lot of influence on win%, OR.. 2) his style of play puts up huge stats (he is the statistical GOAT it seems) but it doesn't influence win% as much as some of the other QB's on the list.

    I'm starting to suspect #2 is correct. It HAS to be one of the two (well there's a spectrum in between of course).
     
  28. Finster

    Finster Finsterious Finologist

    3,087
    2,038
    113
    Jul 27, 2013
    See when you site Brady, you are contradicting yourself, why is he so high, because he's on a great team, well then, why is he so high?

    You are basing this on an avg team, Brady doesn't play on an avg team, so how then is this reflecting reality? It's actually altering reality.

    Is Rivers supposed to carry the brunt of such a poor team? In your chart he would, he would be in the negative, but in reality he may be the reason they won a few, so in reality he's in the plus, but on your chart he would be in the negative.

    This chart is not taking into account the absolute biggest factor in win expectation, the state of the team.

    The chart only pertains to QBs on avg teams, since you have that as a baseline, it doesn't pertain to QBs on good or bad teams.

    Also, how was "avg team" determined? Because you can't use record to determine that, because a bad QB on an above avg team can make them avg, and a good QB on a below avg team can make them avg.

    You have to be able to look at the team without the QB and assess the talent level, lets say it's a 5 win team, but on the strength of a great QB with a 110 rating they win 8, that's plus 3, but on your chart that would make him minus 4.

    This is how it doesn't reflect reality, because in reality, every team in not avg, which should actually speak for itself.
     
  29. cbrad

    cbrad .

    10,659
    12,657
    113
    Dec 21, 2014
    Yeah you're definitely not understanding the graph if you think there's a contradiction with Brady. Brady's result reflects reality the MOST.

    Let me try again. The red line represents what you get with an AVERAGE team. ONLY the red line represents that. Nothing else there is based on "average" as you suggest. All those blue data points are actual (or projected if the QB didn't start 16 games) data points of the teams.

    So.. if you have a given passer rating, then that red line shows (in 2016) the expected wins for an average team. If you're above the red line, as Brady is, then either: 1) you were on an above average team, or 2) your QB had above average influence on win% It's that simple.

    (and technically point #2 only holds when you have above average rating.. for below average rating an above average influence on win% would push you below the red line)

    So Brady's data points make the most sense. He's that high because his team is great, and possibly because he is too. No contradiction. If you wish, Young's result is the harder one to explain but it can be done. Now if that's understood, you can go back and revisit the mystery with Rodgers.

    Rivers btw wouldn't be a mystery because we all know he's on a bad team. Are you going to argue that on average Rodgers has been on a well below average team? I can't argue that.
     
    Last edited: Jul 17, 2017
  30. resnor

    resnor Derp Sherpa

    16,329
    9,874
    113
    Nov 25, 2007
    New Hampshire
    Brady has almost a .500 win record when he throws for below an 85 rating. That's insane.
     
  31. cbrad

    cbrad .

    10,659
    12,657
    113
    Dec 21, 2014
    It's slightly above .500 at 52.5%.

    Thing is, that's a misleading stat because league average passer rating when he started in 2001 was 76.2 and it's only recently gone above 85 so this is including games where his passer rating was well above league average.

    Nevertheless, when compared to other QB's like Peyton during the same period (43% win% for games where his passer rating was <85) you do see the effect of the team.

    The approach I used in post #179 is much better because it adjusts for year and league average. And you do see Brady has the highest value among the QB's I'm comparing to suggesting he's getting the most support from the rest of his team.
     
    jdang307 likes this.
  32. Finster

    Finster Finsterious Finologist

    3,087
    2,038
    113
    Jul 27, 2013
    Again, you contradict yourself with Brady, he is on a great team that would win 11 games without him, so how is it he gets all this credit for winning games? It's contradictory, meanwhile ARod is working with a patchwork team and manages 10 wins and that counts against him on your graph.

    Do you know why the ARod stats don't make sense to you? Because they flat out don't make sense.

    His record in Dec, Jan, = 28-11, 75 TDs to 19 INTs with a 103.8 rating, on 3rd down he has a 108 rating, 99TDs to 26 INTs for his career, and you think he has a lesser impact? Where's his D? They were 21st in scoring last year, Run game? Their leading rusher had less than 500 yds, What about his Oline?

    They won 10 games and made it to the championship game, with poor D, a wishy washy run game and a subpar Oline, ARod is the only reason they were there, he is the most impactful player in the league, bar none.

    If your chart doesn't bear this out? You need to go back to the drawing board, ARod is a one man wrecking crew, this is common knowledge.
     
    resnor likes this.
  33. cbrad

    cbrad .

    10,659
    12,657
    113
    Dec 21, 2014
    Yeah Finster you're simply not understanding what I did. If you really think it's true that Brady is on a team that would win 11 games without him then you'd AGREE with the Brady result!! And where is the credit I gave Brady?? The analysis suggests Brady is probably getting more help from his surrounding cast than any of the other QB's.

    Anyway, I'm certainly open to other approaches because the Rodgers result is a bit weird, but any alternative has to be data-driven. When you say NE would win 11 games without Brady that's pure conjecture (going forward obviously.. yes you can bring up Cassel's 11-5 but Brady-less NE in 2000 went 5-11, etc.. so we need more data points). Maybe it's accurate, maybe not, but it's not a data-driven analysis (hopefully we find out sooner rather than later!!).

    btw.. one thing that would support the Rodgers result is that humans aren't good at predicting win probability. That is, we're not good at figuring out how much different events during a game actually contribute to increasing or decreasing the probability of winning the game. And win probability is really what this analysis is getting at, not the stats previously quoted (passer rating, YPG, TD's, TD:INT ratio). So our intuition might fail in this case.

    Well who knows.. not that important. Anyway, good discussion regardless.
     
    Last edited: Jul 17, 2017
  34. danmarino

    danmarino Tua is H1M! Club Member

    15,360
    20,983
    113
    Sep 4, 2014
    And that's why cherry picking a few games will never tell the entire story and will make a person look silly when trying to push their agenda. I can find similar, if not worse play, from Brady in the postseason.
     
    resnor and Fin D like this.
  35. Finster

    Finster Finsterious Finologist

    3,087
    2,038
    113
    Jul 27, 2013
    Well when data doesn't meet with what is a know fact, obviously the data is wrong, which as you know, happens more times than it's right.

    It's very obvious that the data on ARod is wrong, therefor all of the data may be wrong, this is, if I'm not mistaken, your own take on obtaining this data, which is good work, as I said, the attempt to find the truth in science is always good work.

    WD-40 is called that because it's the 40th attempt, the great thief Thomas Edison said that the 1000 times he attempted to make a light bulb weren't failures, it's just that he figured out 1000 ways not to make a light bulb.

    Personally I think that snipit is one of the best things he's done for science.

    Look at Garrapolo last year, 3-1, you want to site BB's first year in NE, but that's hardly fair in any way shape or form, it's 17 years ago and he hadn't had the time to build his team yet, what are the records of back ups since then? They seem to win no matter who's at QB, they are 14-6 without Brady since Brady took over in 2001.

    Your chart is saying that Brady's presence is worth 2+ wins per year, now if that's true then ARod should be worth more than that, because he's way more important to his teams success than is Brady, and that's easy to see just watching the 2 teams, it jumps off the screen.

    Look at the year ARod got injured, they were 6-2 when he got injured, they finished 8-7-1, 6-2 with, 2-5-1 without, from one of the better teams to one of the poorer teams when ARod is taken out, that's impact, and real empirical data.

    So real life data doesn't match your chart, the Pats win with or without Brady, GB has no chance without ARod, if you do a passrushing chart that says Cam is better than Watt, you have to realize that the chart is showing false data.

    There is not one facet of QBing that ARod is not superior to Brady, including clutch and impact, this is a fact that is backed up by cold hard stats and the eye test, which is why it confounded you, because you actually knew better, it doesn't reflect reality.
     
    resnor and danmarino like this.
  36. danmarino

    danmarino Tua is H1M! Club Member

    15,360
    20,983
    113
    Sep 4, 2014
    For example:

    2006 Playoffs (2005 season): Pats lost to the Broncos 27-13: Brady had a 55 comp% with 1 TD and 2 INT's and a 74 passer rating. He threw his 1st INT with around 1:00 to go in the 3rd quarter with the Broncos winning 10-6. The Pats are on Denver's 5 yard line and about to score to take the lead. Instead, Brady throws a pick and it's returned to their own 1 yard line. In the ensuing play Denver scores and makes it 17-6. Then at the end of the 4th quarter the Pats are once again inside Denver territory and Brady throws an INT to effectively end the game.

    2007 Playoffs (2006 season): Pats win against San Diego even though Brady had a 52 comp% with 3 INT's a Y/A of 5 and a 57 passer rating. 1st INT is in the 2nd quarter when the score is 7-3 Chargers up. The Pats are on the Chargers 35 yard line and Brady throws the pick. The 2nd INT happens in the 10th minute of the 3rd quarter with the score 14-10 San Diego. His 3rd INT came in the 4th quarter with the score 21-13 Chargers. However, Reche Caldwell strips the defender and the Pats recover. They go on the score a TD and 2 pt conversion and tie it up. Then the kick a FG a few drives later to win. Brady tried to choke and give it away, but his team wouldn't let him.

    In the next game of the same season the Pats lost to the Colts 38-34. The Colts had the 32nd ranked defense while the Pats had the 2nd ranked defense. Brady had a 79 passer rating with 1 TD and 1 INT against that putrid defense. Granted, Manning also had a 79 passer rating, but he was playing against a great defense and still had over 100 more yards passing with 1 passing and 1 rushing TD.

    And there are many other similar games from Brady with passer ratings of 66, 49, 57, 62, 78, 56, 68. And the win-loss record was 4-3 for those games.
     
    resnor likes this.
  37. danmarino

    danmarino Tua is H1M! Club Member

    15,360
    20,983
    113
    Sep 4, 2014
    jdang liked your post because I don't think he understood it. lol
     
  38. danmarino

    danmarino Tua is H1M! Club Member

    15,360
    20,983
    113
    Sep 4, 2014
    Great post, but cbrad agrees that Brady gets more help from his team.

    And Garoppolo was 2-0 and Brissett was 1-1. So, the Pats win even with their 3rd stringer coming off the bench and 1 week to prepare.
     
    resnor likes this.
  39. cbrad

    cbrad .

    10,659
    12,657
    113
    Dec 21, 2014
    There's one thing you need to do and it's to stop repeating this falsehood. You don't understand the analysis which is fine. The analysis shows that the offense Brady leads is winning 2+ wins more per year than his passer rating would suggest. To anyone that understands what that means, it suggests his surrounding cast is most likely responsible.

    Anyway, as I've admitted the Rodgers thing doesn't make sense, but it's not that the data is wrong. That IS the data. It's that passer rating isn't capturing what we see occurring. And it's not THAT much of a surprise considering that passer rating isn't perfectly correlated with win%. It just works well most of the time.

    And no you can't cherry pick away 2000 Brady-less NE. I mean they go from 5-11 without Brady to 11-5 with Brady and a SB win in a matter of a year. Just like you can't cherry pick away 11-5 Brady-less 2008 NE or 3-1 Brady-less 2016 NE. Regardless it's not enough data points to know what Belichick without Brady can do.
     
  40. danmarino

    danmarino Tua is H1M! Club Member

    15,360
    20,983
    113
    Sep 4, 2014
    IMO, going 5-11 to 11-5 had little to do with Brady. The Pats defense went from 17th to 5th and to me that was the biggest difference. Hell, even BB thought Bledsoe was better than Brady or he wouldn't have been the starter.

    Plus...cheating.
     
    resnor and Fin D like this.

Share This Page