How long does it take an NFL head coach to reach his 1st Super Bowl?

[GoBills808]

26 minutes ago, folz said:

 

👍 Fair enough.

 

And yes, with Landry, it's fair to take him off my list (as an outlier---due to coaching  6 years prior to SBs). But you definitely can't equate him going to and losing an NFL Championship (in a 15-team league, the same year as SBI) the same as going to a Super Bowl. The Super Bowl actually existed that year and Landry wasn't in it.

So, I'm fine taking him out of the discussion altogether on both sides.

 

 

I don't know...it just seems in these discussions that the bar keeps getting set higher and higher for McDermott. The only coach who didn't win appear in a Super Bowl in 5 6 7 8 seasons, with the same team, with a top QB for more than 5 years (no discussion of how raw that QB was), when the president was a democrat, and Mercury was in retrograde.

 

I know, I'm being a bit over the top there, but it just seems that the more variables get added, the less useful the stat is in showing any kind of true trend or to be used as any type of predictor.

 

Look, we're all Bills fans. We all want a Super Bowl. We differ on our feelings about our head coach getting us there. It's all good.

It's not that complicated I think

 

No team that has ever started the same HC/QB combo for more than 5 years w no SB has seen them go on to win a championship

[oldmanfan]

10 minutes ago, GoBills808 said:

It's not that complicated I think

 

No team that has ever started the same HC/QB combo for more than 5 years w no SB has seen them go on to win a championship

Which only means the first time it happens will be the first time it happens.  It is not predictive necessarily.  I would say if your QB starts tuning out your HC that might be worthy of a change, but I don’t think it means you automatically make a change.

[GoBills808]

16 minutes ago, oldmanfan said:

Which only means the first time it happens will be the first time it happens.  It is not predictive necessarily.  I would say if your QB starts tuning out your HC that might be worthy of a change, but I don’t think it means you automatically make a change.

I mean yes but that's what predictive models do, analyze past data and trends to assess likelihood of future outcomes

 

Like we might double our expected annual rainfall next year but I'm not going to bet my crop on it

[GunnerBill]

11 minutes ago, GoBills808 said:

I mean yes but that's what predictive models do, analyze past data and trends to assess likelihood of future outcomes

 

Like we might double our expected annual rainfall next year but I'm not going to bet my crop on it

 

That isn't a good comparison though. Because there are reasons for rainfall levels and why they tend not to diverge much from the mean. The coach / QB not winning SB thing is an interesting pattern, but there is no reason beyond that pattern to suggest that there couldn't be outliers. 

 

It is an interesting stat to reflect on but I am not persuaded there is any causation. I don't think it makes the Bills any more or less likely to win the Superbowl in 2023.

Edited July 9, 2023 by GunnerBill

[oldmanfan]

13 minutes ago, GoBills808 said:

I mean yes but that's what predictive models do, analyze past data and trends to assess likelihood of future outcomes

 

Like we might double our expected annual rainfall next year but I'm not going to bet my crop on it

Predictive models take into account independent variables

[GoBills808]

1 minute ago, GunnerBill said:

 

That isn't a good comparison though. Because there are reasons for rainfall levels and why they tend not to diverge much from the mean. The coach / QB not winning SB things is an interesting pattern, but there is no reason beyond that pattern to suggest that there couldn't be outliers. 

 

It is an interesting stat to reflect on but I am not persuaded there is any causation. I don't think it makes the Bills any more or less likely to win the Superbowl in 2023.

Why isn't it a good comparison? We are talking about sets of data points modelling outcomes, they're alike in that regard

[GunnerBill]

Just now, GoBills808 said:

Why isn't it a good comparison? We are talking about sets of data points modelling outcomes, they're alike in that regard

 

Because one has external factors that make that pattern considerably more likely to repeat. With the other it could just as likely be coincidence, and none of the arguments put forward in this threat really take it beyond "there is a pattern". 

[GoBills808]

3 minutes ago, oldmanfan said:

Predictive models take into account independent variables

Your contention is that in this particular data set, the 5 year threshold for winning the Super Bowl is an independent variable?

Just now, GunnerBill said:

 

Because one has external factors that make that pattern considerably more likely to repeat. With the other it could just as likely be coincidence, and none of the arguments put forward in this threat really take it beyond "there is a pattern". 

You are saying there are no external factors on a football team/HC QB relationship/lockerroom politics/overall NFL landscape that might influence this 5 year threshold?

 

 

[GunnerBill]

7 minutes ago, GoBills808 said:

 

You are saying there are no external factors on a football team/HC QB relationship/lockerroom politics/overall NFL landscape that might influence this 5 year threshold?

 

 

 

There are. But whereas the factors that affect climate and weather are the same underlying factors year on year - though not necessarily to the ssme degrees. 

 

But the factors are not anywhere near as consistent between each one of the HC/QB situations that have tried more than 5 times and failed. And that is why it could easily be coincidence. 

 

It is a pattern. I don't dismiss that. But I see no evidence of any causation.

Edited July 9, 2023 by GunnerBill

[HOUSE]

In Cleveland, 70 years 

[GoBills808]

5 minutes ago, GunnerBill said:

 

There are. But whereas the factors that affect climate and weather are the same underlying factors year on year - though not necessarily to the ssme degrees. 

 

But the factors are not anywhere near as consistent between each one of the HC/QB situations that have tried more than 5 times and failed. And that is why it could easily be coincidence. 

 

It is a pattern. I don't dismiss that. But I see no evidence of any causation.

I guess

 

I see it as a valid comparison. w weather we look at all different climates, elevations, latitudes etc and you still get pretty reliable data from the past to make determinations about what might occur in the future. In the NFL we have different teams from different eras with different HC/QB combos, different locker rooms and different ownership but all with the same 5 year threshold for winning a Super Bowl...that would be enough for me to make a determination and feel fairly confident

[Einstein]

16 minutes ago, GunnerBill said:

 

Because one has external factors 

Everything stat has external factors. But we (as humans) tend to only bring up those external factors when it suits us.

 

For example, why do we even mention INT totals for bad QB's? Weather, game plan, tipped passes, etc all are factors that could change this statistic. 

 

Another example is test scores. We  use standardized test scores to measure the performance of a student? After all, external factors such as socio-economic status, parental education levels, language barriers, quality of teaching, and even the student's health and nutrition can significantly affect these scores. That persons not dumb, they're just unhealthy!

 

And why blame weight gain on poor eating habits? Numerous other factors like metabolism, genetics, mental health, medications, and socioeconomic status also play a huge role. That person's not fat, they're just poor!

 

We can "external factor" our way out of every statistic to ever exist if we wanted to.

[Sammy Watkins' Rib]

 

48 minutes ago, oldmanfan said:

Which only means the first time it happens will be the first time it happens.  It is not predictive necessarily.  I would say if your QB starts tuning out your HC that might be worthy of a change, but I don’t think it means you automatically make a change.

 

It’s particular important because it tells us owners have a short leash on HC’s once the team has a franchise QB. 
 

That’s the history. That’s why the stat exists.

 

If more franchises were run like Philly when Reid was there and given 14 seasons with 10 consecutive seasons with McMabb as the starting QB (despite no Super Bowl victory anyway) we would likely have a handful of HC/QB combos with rings well past season 5 together. 

 

But Reid and McNabb were a very rare combo that stayed together for a long time. And still they never got their ring. 
 

Elway and Dan Reeves may have won a ring together eventually if they were together for all of Elway’s 16 seasons. 
 

As it was, they lasted 10 seasons together just like McNabb and Reid. 
 

I think McDermott and Allen are likely to follow the Reeves and Reid model and last 10 seasons together even without a ring. But staying together that long is often the exception. Would be hard to think McDermott’s seat would not be on fire come season 8 and 9. 
 

The key will be staying competitive in the postseason. Got to get back to a conference championship game again in at least one of the next three seasons I think to keep his seat cool.
 

 

15 minutes ago, GunnerBill said:

 

 

It is a pattern. I don't dismiss that. But I see no evidence of any causation.

The causation is simply the tendency for owners to become impatient. 
 

I agree, simply year to year the Bills or any other team have as good a chance to win the Super Bowl as any other year. 
 

What increases as the years go by is the chance the owner / GM pulls the plug.

 

Extreme patience seems to be about 10 years and even in those rare examples the ultimate goal of winning a Super Bowl was unsuccessful.

[Einstein]

27 minutes ago, oldmanfan said:

Predictive models take into account independent variables

 

Not the type of independent variables that you are speaking of.

 

Like my evidence in the first post, Auto Regressive models literally use past values and past prediction errors to predict future values. It's the entire purpose.

 

For example, in an autoregressive model, the estimate for a variable's value at a given time point 't' is derived from an amalgamation of its preceding values—designated as ('t-1', 't-2') and so forth. Despite their temporal antecedence, these values serve as the model's surrogate "independent variables", even though they are, in fact, antecedent observations of the very variable being forecasted. They are not a completely separate independent variable like you are talking about (weather, injuries, whatever).

 

Really an time series data model, like recurrent neural neworks. They are specifically designed to "remember" patterns over time, without excuses, which makes them great for univariate time series forecasting.

[Augie]

9 minutes ago, Einstein said:

 

Not the type of independent variables that you are speaking of.

 

Like my evidence in the first post, Auto Regressive models literally use past values and past prediction errors to predict future values. It's the entire purpose.

 

For example, in an autoregressive model, the estimate for a variable's value at a given time point 't' is derived from an amalgamation of its preceding values—designated as ('t-1', 't-2') and so forth. Despite their temporal antecedence, these values serve as the model's surrogate "independent variables", even though they are, in fact, antecedent observations of the very variable being forecasted. They are not a completely separate independent variable like you are talking about (weather, injuries, whatever).

 

Really an time series data model, like recurrent neural neworks. They are specifically designed to "remember" patterns over time, without excuses, which makes them great for univariate time series forecasting.

 

It’s sports, not math. 

[Einstein]

8 hours ago, Nextmanup said:

Why do you assume that if we move on from McDermott we will replace him with a clown show?

 

It's called framing an argument.

 

McDermott supporters have run out of runway with their original arguments due to another disastrous playoffs so they must now frame the argument as a "Stick with McDermott and at least win the regular season" or "LOSE FOREVERRRR".

 

It's not like a team could fire a successful coach (Pederson in Philly) and then immediately hire a coach that takes them to the Super Bowl (oh wait, Sirianni in Phily). That would be impossible.

 

5 hours ago, GunnerBill said:

 

Except the data does not support that.

What data supports that McDermott is a good in-game strategist?

[oldmanfan]

12 minutes ago, Einstein said:

 

It's called framing an argument.

 

McDermott supporters have run out of runway with their original arguments due to another disastrous playoffs so they must now frame the argument as a "Stick with McDermott and at least win the regular season" or "LOSE FOREVERRRR".

 

It's not like a team could fire a successful coach (Pederson in Philly) and then immediately hire a coach that takes them to the Super Bowl (oh wait, Sirianni in Phily). That would be impossible.

 

What data supports that McDermott is a good in-game strategist?

You framed the argument by deciding what you wanted as an answer and then setting up your analysis to give you what you wanted.  You deny it, but it is obvious.

[Einstein]

54 minutes ago, oldmanfan said:

You framed the argument by deciding what you wanted as an answer and then setting up your analysis to give you what you wanted.  You deny it, but it is obvious.

 

Yeah you have attempted to make this claim several times but have still yet to prove it.

 

That’s called an unsubstantiated claim. 

[oldmanfan]

39 minutes ago, Einstein said:

 

Yeah you have attempted to make this claim several times but have still yet to prove it.

 

That’s called an unsubstantiated claim. 

I don’t have to prove offering an opinion.  But I base my opinion on 40 years in research and as a reviewer for multiple professional journals.  You began your biased undertaking by starting with an assumption that has been obvious to all that have seen your posts:  that McD is bad and needs to be replaced.  So you made up an assumption (that many people are thinking McD can’t get a team to a SB) then chose that endpoint because you realized you could slant data to fit the conclusion you wanted.  Then when challenged started spouting off all kinds of esoteric statistical formulae.  
 

So if I were reviewing this for a journal, I would first ask why you framed your research question as you did.  I would ask why you did not for example chose winning a SB instead of getting to a SB, since winning one is the ultimate goal.  And because one could point to coaches such as Reid and Belichick.  So at first one asks whether the research question has any relevance.  Let’s assume your does.  Then you look at Materials and methods.  And here you never explain why you ignore or throw out data that could impact your analysis.  You’d have to look at whether critical injuries kept teams from winning a conference, whether there was a GM change that may have impacted results, and many others.  You provide no reason why you did not do so.  I would bet you’ll want to say it normalizes out, but without actual data on that you have no way to know that.  Finally I would get an independent review from colleagues in the Statistics Department to evaluate the statistical methods.  After 40 years I can tell you that the vast majority of the time (over 90%) they tell me the stats are wrongly applied.
 

I and many other reviewers would thus reject this paper because it is fatally flawed.  So while the only way I could “prove” your intent would be to do a Vulcan mind meld of the brain or something similar, my experience tells me all I need to know.  You go ahead and keep playing your let’s figure out how to bash McD stuff; most here realize what you’re doing and why.  McD may win, he may not.  At some point in his career he will likely get fired, because the vast majority of coaches in any major league sport do.  But it won’t be because of your “analysis”.  It will be because he either has an inpatient owner or because the team starts tuning him out.  There is no evidence for either at present.

Edited July 10, 2023 by oldmanfan

[GoBills808]

'Sean McDermott is unlikely to ever win a Super Bowl because no HC/QB combo ever has after their first five seasons together'...you need an independent statistics department review to make that analysis

 

'Sean McDermott is a great head coach because he has a high win%...applause