Tannehill to Miami?

Let's do this formally. If you run a t-test on wins in the years where you had Belichick + Adams but without Brady in NE versus Belichick without Adams and without Brady (i.e., the years 2000, 2008 and 2020 vs. 2021-2023) you don't get a statistically significant difference in win%. The p-value is 0.7956, meaning that there's a ~80% chance the win% comes from the same underlying distribution.

There's your statistical evidence that Adams' effect was not that large: it wasn't even statistically significant, which means you can't rule that there was actually no effect on win%.

Regarding Brady vs Bledsoe, using the larger sample size and overall passer rating is a far better metric than removing individual games, and clearly Brady > Bledsoe in 2000-2001 when you use all the data.

 

I don’t think looking at 2000 (their first year in NE) 2008 (unexpectedly losing Brady and still winning 11 games), and 2020 the first year without their 20 yr starter at QB who participated in their cheating is really telling us much.

Check this out:

“As the scandal broke, the NFL was investigating a possible violation into the number of radio frequencies the Patriots were using during the Jets game, sources told ESPN’s Chris Mortensen, who reported at the time that the Pats did not “have a satisfactory explanation when asked about possible irregularities in its communication setup during the game.”

Quarterbacks communicate with the sidelines via microphones in their helmets that pick up an NFL-monitored radio frequency. An NFL sideline official cuts off communications on this frequency 15 seconds before the play clock runs out.

O’Leary — who uses data crunched by a Las Vegas bookie and a Ph.D. statistician from China with no previous familiarity with Spygate — suggests Patriots “director of football research” Ernie Adams, a prep-school chum of Belichick from Phillips Academy in Andover, Mass., was the nerve center behind the chicanery.

Offensive plays would be called based on stolen signals and the information relayed straight to Brady’s helmet, O’Leary theorizes.

In this scenario, the extra frequency is critical, as it allows the team to do something in real time with the stolen signals, out of earshot of the NFL monitor, and change its plays accordingly.

If there’s an open channel during the play itself, you can also alert the quarterback to open receivers he may not see.

O’Leary repeats a rumor that Pats backup quarterback Doug Flutie once said he accidentally picked up Brady’s helmet during the 2005 season.

“He was amazed that the coaches kept right on speaking to Brady past the 15-second cutoff, right up until the snap,” according to O’Leary.

“The voice in Tom Brady’s helmet was explaining the exact defense he was about to face.”

https://nypost.com/2014/10/12/they-are-cheaters-spygate-the-nfl-scandal-that-started-it-all/

 

Look.. I know you want it to be true that Adams had this huge effect, but there's no evidence of it. I'm willing to grant you something behind the scenes probably went on that improved NE's chances of winning — your links are providing evidence of that. What is however absolutely clear is that whatever cheating went on its effect on win% wasn't that great. It wasn't even statistically significant.

That's not up for debate. The only significant predictor of NE's success during the Belichick years (i.e., we're not evaluating Belichick here, just Brady vs. Adams) is Brady.

I think this is a case where you're going to have to modify your beliefs in face of evidence to the contrary. Can't argue we should be using statistics then argue against something this solid.

 

I don’t think there’s any evidence that shows the cheating wasn’t significant… in fact I think it’s the opposite. Look at the Pats from 2000-2020. There are 2 years out of 20 where you’re getting your “with Adams and without Brady” stats from. 2000 was the first year and 2020 was his last. That’s not enough statistical evidence IMO to prove anything… especially when you consider the circumstances. In 2008 the Pats still won 11 games with a QB that hadn’t started a game since high school and came in and looked nearly identical to Brady. In 2002, and with Brady, the Pats won only 9 games and missed the playoffs all together.

 

Also, I don’t want it to be true. I wish I could tell my grandchildren they I got to live through possibly the greatest dynasty in football with the greatest QB ever. I love the Dolphins, but NFL football as a whole is greater. I don’t begrudge Rodgers; Manning, Brees, Mahomes… etc etc… I think Kelly and Bledsoe were great QB’s. I have no problems with calling opposing team’s players great. I can’t do that for Brady and the Pats. It’s tainted. We don’t know how much their cheating helped. We know their results while cheating, however. 3 SB’s in 3 years before being caught… no SB wins for 10 years after being caught…3 more SB’s under heavy scrutiny of cheating with lots of rumors, evidence, etc.

The Pats cheated 100%. We don’t know what it all entailed. Partly because the NFL itself destroyed evidence. But why do what they did if it had no real bearing on their win/loss columns?

 

3 years: 2000, 2008 and 2020. Brady had 11 passing attempts in 2008.

Let's use games, not seasons as the sample size. So for Belichick + Adams it's 48 games and Belichick without Adams it's 51 games. That's more than enough.

The p-value for point differential is 0.8431 which as you can see is similar to the p-value for win%. The p-value for points scored is 0.9094 and for points allowed is 0.6497. So very clearly no significant difference with Adams. Obviously there's a significant difference for point differential, wins and points scored with Brady.

I'm only arguing the effect, nothing else. The effect was not big.

 

So we have heard that Brady was most likely an important part in the cheating. And again 2000 was the first year together. We know that stealing signs was a big part of the cheating. You can’t get signs until you play teams. They had no data to cheat in 2000. In 2008, how much was Cassel involved, if at all, in the cheating? In 2020, Brady was gone. Again, 1/3 of the system’s players was gone.
I don’t think the stats are good stats when these things are not factored in.

If I cheated at poker with a tainted deck of cards, and won many championships, but had that deck replaced with normal cards, and didn’t win anything, what was the difference? The deck of cards. Brady was a part of the deck of cards. A big part.

 

You've got this reversed. You're looking at reports of cheating, assuming it had a great effect, then trying to come up with any and every possible unobservable mechanism to keep the hypothesis alive that cheating must have had a great effect.

That's not the way to go about this. First, you look for evidence there was an effect that needs to be explained. Only if you find one do you go down that speculation route. There is no evidence any of that cheating had any significant effect. And Brady won a SB and had comparable stats going to Tampa while Belichick with or without Adams floundered.

Evidence is clear on this: it was primarily Brady, not Adams (or Belichick for that matter).

 

“If the numbers don’t match what is seen on film, something is likely wrong with the numbers.”

You implied that your eyes will cause you to ignore the numbers. I’m on board with ignoring numbers that aren’t good enough to show the whole picture. I’m onboard with using your eyes and numbers. But if the numbers are good, and your eyes are showing something different from the numbers, then the eyes are most likely wrong. You’re insinuating the opposite.

 

Not really. They cheated and did it for a long time. Most likely that cheating was way more involved than we know. Without knowing the true extent of the cheating there’s really no way to figure out how much it helped. And why do it if it didn’t matter? Occam’s Razor is a legit thing, as you know. And that tells me they cheated, it was greater than we know, and it helped considerably or else they wouldn’t have done it.

 

Occam's razor in this case says: if you don't see a statistically significant effect, then most likely there isn't one!

As far as why they'd cheat if it didn't matter that much, that's easy to answer. There are many highly competitive individuals who care about minute details. I do that in software and hardware design, scientific research papers etc. I'm talking about minute details like how well commented code is, how flexible code or hardware design is to future changes in requirements, even tiny things like font size on images etc. All those minute details add up, but not to that much. It just gives you a slight competitive edge, nothing more, but that's worth it. You were in the military right? There are a lot of minute details I'm sure they cared about that don't amount to too much but are worth doing because it gives you a slight competitive edge.

There's no major mystery here to explain. It was primarily Brady.

 

You’re forgetting that the data you’re using is flawed.

2000: First year with no data to cheat

2008: 1/3 of the cheating trifecta gone suddenly and unplanned

2020: 1/3 again missing.

I just don’t see how the data available can prove one way or another how much it helped.

What we do know, however, is that they cheated. Why cheat if it wasn’t a significant help? It’s not like they did it for a couple of years and stopped. Wouldn’t that have been the case if it wasn’t significantly a big help?

A slight competitive edge doesn’t make up for the risks involved. The juice wouldn’t be worth the squeeze. There might be a competitive edge in capturing and torturing your enemies, but the risks involved are greater than the edge it may give, so it’s not done. Not to mention the moral implications.

 

No it doesn't work that way. First of all there's no justification for ignoring the Cleveland data, and including it gives you the same result but with larger sample size. Second, you can get data during the season for cheating. There's no reason why an effect wouldn't have been seen in year 1. Third, it's important to look at the effect of Adams without Brady, and in that analysis there's no such thing as saying 1/3 (Brady) is missing. Fourth, if Brady is that crucial but he needs Adams and/or Belichick, then he wouldn't have had similar success in Tampa Bay.

The statistical analysis is very clear on this. I can even run a regression model if necessary (unnecessary though given what has already been shown), but the result is going to be it's primarily Brady. We could even put an interaction term with Brady*Adams or Brady*Adams*Belichick in there, but what that model will do is "note" (mathematically) that there's no significant difference when Brady goes to Tampa Bay.

Evidence is clear dude.

And again, many highly competitive people put effort into things that give them only a minor competitive advantage. This assumption that the only reason to cheat would be if it gave you a huge advantage is completely wrong. Many humans don't behave that way.

 

Yeah, I was wrong there. Missed that post. Cbrad pointed that out too.

 

There’s plenty of reason to ignore the Cleveland data. There was never any accusations or evidence of cheating then.

And I just can’t see how using data from 2000, 2008, or 2020 can be used what with the circumstances of each year. There’s plenty of legit reasons to call those years anomalies.


Here’s another way to look at those first 6-7 years:


“In the second game of the season, Drew Bledsoe was lost to injury, and in comes second-year player Tom Brady.

A kid from Michigan that was taken in the sixth round, and did not look anything like the part of a Hall of Fame player. He was skinny, looked out of shape, and no NFL team thought that he was worthy of being drafted in the first five rounds.

The Patriots turn things around, and qualify as the second seed for the playoffs.

Is it possible that a second year player, drafted in the sixth round, with no game experience, could lead his team, on his own, to the second seed from one of the worst?

Could knowing the defensive signals have had any input in his progression?

The Patriots hosted the Oakland Raiders in a snow storm. The game was won on an overtime field goal, sending the Patriots to Pittsburgh to play the Steelers.

I am not going to say, for one second, anything negative about the call that resulted in the tuck rule. It is a rule that is on the books, and the call that was made was correct.

But, knowing the defensive signals, would the game have gone into overtime if they didn't? Does anyone else find it funny that ALL of the Patriots points were scored in the second half of the game. After halftime adjustments were made?


Had the Patriots not won that game, they obviously would not have won that Super Bowl.

Part of the spygate issue was the video taping of the Rams' walkthrough practice prior to the Super Bowl.

After much speculation, the man that was in charge of videotaping for the Patriots, Matt Walsh, told the NFL that there was no videotaping that happened, but that himself, and other members of the Patriots staff, did witness the walkthrough of the Rams.

Did they take notes? What were they watching? Why were they watching?

Halfway through the second quarter, Ty Law intercepted a pass from Kurt Warner and returned it for the games first touchdown.

At the end of the first half, Tom Brady led the Patriots on a 60-yard scoring drive, that gave the Pats a 14-3 lead at half time.

Because of the formation, did Ty Law have some idea what play was being run? Was that information known from what was watched by Mr. Walsh?

Did the Patriots have some idea of what the Rams were doing that made Brady be able to drive the field, when he had not been able to prior to that?

New England won by three.

Since it has been admitted that there was taping, then if any of the above mentioned plays were the result of "spying" then the Patriots would not have been the Super Bowl Champions.

2002 the Patriots do not make the playoffs.

Super Bowl XXXVIII.

The New England Patriots were playing the Carolina Panthers. With just under four minutes to go in the third quarter, the Patriots took over and had an eight play, 71 yard drive that led in a score.
In the last 1:08 of the game, Tom Brady went 5-5, not counting the offensive pass interference call, that led the Patriots to the victory.

Was it possible that any of the plays at that crucial drive Brady knew what the defense was doing?

2004

New England took a 21-game winning streak into Pittsburgh to take on Big Ben Roethlisberger and the Steelers in week seven.

The rookie and the Steelers took the Patriots apart, 34-20. The Steelers defense held the Patriots to 3-9 on third downs. The Patriots were held to 248 yards of total offense, and only five rushing yards.

In the AFC Championship game, also in Pittsburgh, things were quite different.

The Patriots picked the Steelers defense apart, to the tune of 41 points. Forget the fact that Pittsburgh was the No. 1 defense in the NFL that year—No. 4 against the pass, and No. 1 against the run.

Could it be that the Patriots were just so much better that day? Or could they have know what Pittsburgh was doing?

With all the speculation over what happened in the first part of this decade, if Pittsburgh wins the Super Bowl this year, both the Steelers and the Patriots will be tied with three Super Bowls each.

After really looking at the situations, is it not only possible, but probably, that had New England not cheated, that they may not have even won ONE Super Bowl? Had the Patriots not cheated, Pittsburgh may have had four right now.”

https://syndication.bleacherreport....eally-give-the-patriots-an-advantage.amp.html

 

Even if the cheating gave only a “slight” advantage, and as you’ve alluded to slight advantages are often sought after, wouldn’t that be enough to win games? And if cheating was the reason for these “slight” advantages can’t we say that the cheating was the greatest factor in winning? If it put them over the top then cheating was the biggest factor in winning.

 

And what about the Super Bowl against the Seahawks? Did it not look like Butler knew that play was coming? He beat the WR to the spot. Of course he may have seen this play on legit tape, but really? A rookie CB that had started one game that season?

After all the evidence supporting the Pats cheated and knew the plays can we really think that’s the case?

 

How do you know they weren't cheating in Cleveland and just didn't get caught? Since we're talking about unobservable mechanisms one can easily use that same argument in reverse. And while you can come up with all kinds of reasons why 2000, 2008 or 2020 might be "different", you could do the same for 2021-2023, which is where you didn't have Adams. That's not how you should decide whether to remove those data points as outliers in statistical analysis.

The statistical argument for not treating those years as outliers to be removed is simple: there's no significant difference observed between those years and 2021-2023. In other words, to justify removing those data points as outliers you'd need the very evidence you don't have. In general one shouldn't remove things you think are outliers in statistical analysis and instead modify the analysis (e.g., instead of testing for a difference in means, which are sensitive to outliers, you could test for a difference in medians, which aren't, but that doesn't change the results here).

Yes it's enough to win games. I could get behind an argument that suggests NE might have only won 3 SBs without it. But their overall win% would probably have been similar (maybe at most 1 less win per season?). So saying the effect is "slight" doesn't mean I'm dismissing it. Like I said, I think Brady is likely responsible for 3-4 extra wins per season, but maybe not all the 5 observed. My main point is that the evidence is clear as to what the primary reason for NE's success was: Brady. That the success was the best ever by any franchise is where I think there's an opening to suggest something else might be helping them. But IMO they're at least a 3 SB team with Brady and Belichick over 20 years. That's close to expected statistically for that level of elite QB play and top-level defense.

I think that one play was more a brain fart by Pete Carroll than anything else. Seahawks win that SB if they just run the ball on that play. The preparation for NE was good, yes, but that preparation could easily be due to available game tape. The reason NE won that SB is mostly luck that the opposing HC called one of the stupidest plays ever in a SB.

 

There was no accusations or evidence for cheating in Cleveland. If we’re going to add stuff for the sake of adding stuff then nothing will be true.

The reason to remove the year 2000 makes sense. You can’t drive the car until you build the car. And again, there is no evidence or accusations that the Pats did what they were caught doing in 2001 (6 years later).

And the evidence shows that Brady was heavily involved. So him not being there in 2008 is a good reason to throw it out.

Same with 2020. Brady was gone.

the commonality between those three years and 2021-2023 is that major pieces of the cheating puzzle were removed. Do we not think removing any one of the three major players isn’t good enough to consider them an anomaly?
What we do know, however, is that when all THREE were there they did things almost unheard of.

Are you claiming that Brady wasn’t the reason for the Seahawks SB win?

 

You realize that a lot of the things you've been suggesting have no actual evidence associated with them, like what precisely was on the tapes that were destroyed or who precisely was hushed, etc. That info is important for determining how serious the cheating really was. My Cleveland comment was just following in the vein of this type of speculation of unobservable mechanisms. My entire argument works ignoring Cleveland of course.

And again, for 2000, 2008 and 2020, from a purely statistical point of view you can't suggest removing them from consideration unless there's at least statistical evidence they're outliers (during non-Brady years), and they're not.

And yes I'm not crediting Brady with that one SB win against the Seahawks. That "win" goes to Pete Carroll lol. Own goal.

 

There’s enough evidence in the fact that the tapes were immediately destroyed to suggest it was more than what was reported. Add to it the smoke from all the NFL execs, players, and coaches and it’s a pretty good assumption that the cheating was far worse.

The sustainability of that organization for 20 years is damn near impossible in today’s salary capped league.

Anyhow, for me, there’s just no changing my mind (with what we know and don’t know) that those 20 years are tainted. I’m not saying that BB or Brady weren’t phenomenal at their respective careers, I’m questioning the level of their successes. And that’s based solely on their proven cheating. As you’ve even said, at minimum the cheating gave them a slight advantage. And when we’re talking about a game of inches, that’s enough of an edge to put them over the top.

 

I remember running the numbers on the Pats win% home -v- away a few tears back and will post a link to it when I find it again.

The Pats breakdown was that they won at the same rate at hone and same rate away as other teams with an overall winning % similar to the Pats.

While I don't doubt the Pats cheated, what the win% breakdowns tell you is that there are 3 scenarios
1) the cheating provided no benefit
2) all NFL teams cheat so it put the Pats on a level playing field with the rest if the NFL
3) The Pats were such masters at cheating that they were able to benefit from it at the same rate away in stadiums/cities they didn't control as they did at home. Remember a lot of the allegations refer to things such as bugging rooms and spy cameras at practise facilities which are functionally impossible for the Pats to do at 30 something other venues.
Or maybe even
4) they didn't actually cheat.

 

Note that doesn't answer the question of whether they won more than expected, but that does dovetail with what I was showing above, namely that you don't see statistically significant effects of any hypothesized "cheating" when looking at regular season game records. That means that whatever the effect was, it's not huge even if it might make a big difference in number of SBs won.

I think you have to ask the question of whether there's an anomaly in winning 6 SBs over 18 years. Different ways of estimating how many should be expected, but since we're talking about the effect of a QB among other things I used defense adjusted z-score passer ratings, and a team with an elite QB + top level defense has an expected ~3 SBs over that period, or once every 6 years, which seems intuitive.

Now use the binomial theorem with a 1/6 probability of winning. The probability you'd win 6 SBs across 18 years is only 2.1%. In other words you'd expect what NE did once in every 47 franchises with top-level QB and top-level defense over an 18 year period. That to me seems too unlikely to have occurred in 58 years of SB-era NFL history. That's where the opening lies: what's the effect of a slight advantage that can win you that one extra game in the season that might not reveal itself in larger scale regular season statistics.

 

Here is the link to the old thread with numbers and discussion
https://thephins.com/threads/its-no-harder-to-win-at-ne-than-anywhere-else.93568/

Definitely "spot cheating" on one game a year in the playoffs would be much harder to pick than consistently doing it throughout the year.
They may test their system on a handful of plays in each game, which may not be enough to alter the overall win% but be enough to verify their system.
The argument against this is human nature and greed when there are big prizes up for grabs every year.

 

Similar to teams in any given year or similar to teams who also did it for 20 years straight? I suspect it’s the former because I can’t think of any other teams in that span who consistently won as much as the Pats.
As cbrad alluded to, and I agree with, the cheating they did (and it’s proven that they did in fact cheat) wasn’t a panacea for undefeated seasons or never losing. What it did do, and what one of my recent posts shows, is put them at a distinct advantage in a lot of games. A few plays here and there, especially in the post season, where even a small advantage can turn the tides so to speak, is the difference between getting to (and winning) a Super Bowl.

As I mentioned earlier, a “game of inches” can be determined in small fractions. One incompletion, reception, or INT, for example, can change the outcomes of any game. The Giants Super Bowl winning “helmet catch”, the Seattle Super Bowl loss INT, and the Rams loss due to just enough yards for FG’s were the determine factors in those games. The proven cheating by the Pats had an effect on their post season success. We can argue to what extent, but we can’t argue that their cheating benefitted them.

 

Tannehill had Philbin and Gase running the offense for his full tenure in Miami with worse OL play and inferior skill position players than Tua, but somehow Tua's early struggles are all because of Flores. Imagine Tua playing his rookie season w the best WR on he had being Brian Hartline. Friggin hilarious.

 

Google the 2020 Dolphins roster lol.

 

The Dolphins as a whole were mediocre under Flores, they are still mediocre but more exciting w McDaniel. That has more to do w personnel than coaching unless you think having Tyreek Hill wouldn't make ANY QB/offense more effective.

 

The 2016 team had 5 first round picks across the OL. Devante Parker, Kenny Stills, and Jarvis Landry all in their primes at WR. Not the greatest WR unit by any measure, but significantly better than the 2020 and 2021 units.

That 2016 offense was infinitely better than the 2020 and 2021 offenses.

 

Yeah, the OL play is provably wrong.

@Unlucky 13 posted this:

Games missed to injury by OL starters among winning AFC teams in 2022 and 23 combined:

Miami: 49
Jacksonville: 25
Baltimore: 17
LA Chargers: 17
Kansas City: 8
Buffalo: 7
Pittsburgh: 4
Cincinnati: 1

This also doesn't include the partial games lost by various Fins as they went out early.


Tyreek has done better in Miami than he did in KC.

 

2016 was Tannehill's fifth year, how is that relevant to what happened in Tua's first two seasons?

 

I was responding to your post that implied the Tannehill had a worse offense his full tenure than Tua did in 2020-2021.