Einstein

In my opinion Allen became a franchise QB in year 2, when he had 29 TD’s. But keep in mind (since I sense that this is where you’re going) that the data reflects total years. Not years since a coach gets a franchise QB. My Belichick comment was simply explaining the variable.

It’s sad that math is so elusive nowadays that it’s seen as foreign and copied. We know the historical percentage of coaches who made it past year 7. My post above spelled this out for you already, but given that only 5 coaches in the last 40 years have made it to the SB past year 7 of their coaching career, the simple probability historically = (Number of coaches past year 7 who made a Super Bowl) / (Number of Super Bowl coaches). 5 / 48 or approximately 0.1042 (10.42%). but this doesn’t take into account some coaches made it several times, and other coaches made it one time and there are more than 48 opportunities but because of the duplicate coaching data it doesn’t take that into account.

And the data shows that McDermott likely never will.

Actually, that’s exactly what i’m doing (on a simple level). Do you not understand how probability works? I’ll teach you. It’s simple math. For example, the probability of landing a heads or a tails on a two sided coin in 50%. Because there are two faces and 1/2 = 0.5. Or how about the probability of rolling a 6 on a six-sided die? The total number of outcomes is 6 (the die can land on 1, 2, 3, 4, 5, or 6). The number of favorable outcomes is 1. So the probability is P = 1 / 6 ≈ 0.1667. I’m sure you’ll say “but variables”, and you can find conditional probability with variables as well. For example… say Josh Allen completes a pass 65% of the time (making it up). However, his completion rate increases to 75% when the Bills are in the lead and we lead in 40% of our games. So, how can we figure out the probability the Bills are winning the game if Allen completes a pass? Well, we can define L as the Bills currently having the lead, P as Allen just completing a pass, and we want to find P(L | P)… (the probability the Bills are leading when Allen just completed the pass). We know from our data set that P(L) = 0.40 ... P(~L) = 1 - P(L) = 0.60 (probability the Bills are losing). P(P | L) = 0.75, the probability of Allen just completing a pass given the Bills are in the lead. P(P | ~L) = 0.65, the probability of Allen completing a pass given the Bills are losing or tied. Then using the law of total probability we can find Allen’s total probability of completing a pass in ANY scenario (winning or losing): P(P) = P(P and L) + P(P and ~L) = P(P | L) * P(L) + P(P | ~L) * P(~L) = 0.75 * 0.40 + 0.65 * 0.60 = 0.30 + 0.39 = 0.69. Lastly you can use Bayes theorem to find the probability the Bills are leading when Allen just completed that pass: P(L | P) = [P(P | L) * P(L)] / P(P) = (0.75 * 0.40) / 0.69 = 0.434. So, if Allen just completed a pass (given the data set - and my data set was much larger), there is about a 43.4% chance the Bills are currently leading in the gwmeZ. I think. This is rich, coming from a poster that likes posts full of factual inaccuracies (literal statistical wrongs) because they have a (wrong) conclusion that agree with his own. You don’t care about facts - you want your feelings placated.

Yep. My analysis of the data is strikingly accurate with all of the other (similar studies) done on this topic. I understand @GunnerBill assertion that it doesn’t prove causation, and he is right that there is no absolute, but the evidence is so darn consistent that you would really have to squint to pretend that it’s not alarming.

The data indicates the probability of the complete opposite. But again, I realize that this is not your strong suit.

PS, for everyone who is interested in this line of research. Five Thirty Eight (the polling website) did an entire research column on an offshoot of this very subject. Rather than just looking at coaches, they looked at head coach and quarterback duos. The result? They determined that HC/QB duo’s have only 5 years. Very close to my 4.2 result. https://fivethirtyeight.com/features/why-coaches-and-qbs-should-divorce-after-five-years-of-not-winning/amp/

Im glad I was able to teach you the word that you meant to use. But no, using 40 years of every Super Bowl attending coach is not cherry picked data. It is as extensive as you get. Yeah, this part proved that you’re not a scientist nor a researcher… To complain that a 40 year data set is invalidated due to a team (that is not even included in the data set) had a player almost die on the field? It’s not only ignorant of data study, but it is also laughable.

Pretty much.

The hypothesis doesn’t invalidate the data. Your ruining your scientist ploy. All data in studies are selected. There is no other way to narrow the subject from the entire universe to the topic at hand. I think what you meant to say is the data is cherry picked. Which, you would again be wrong, considering I used every Super Bowl attending coach for 40 year. It doesn’t get more large data set than that. As said before - What is actually happening is that you do not like the conclusive results of the data, but you know that with an extremely large data set of 40 years, there is very little that you can do to discredit the data. Therefore, you resort to ad hominem.

You don’t seem to be. You still have no basis for this assertion. What is in all actuality happening is that you do not like the data, and with an extremely large data set of 40 years you know that there is very little that you can do to discredit the data. Therefore, you resort to ad hominem. It’s called presenting a full data set (I included both mean, mode, and range) to allow people to see the full data.

You said that you are a scientist with experience in research. Surely you understand that the mode does not intimate the expected result. You already wrote this but, again, you have no basis for this assertion. Not a single shred of evidence. While on the off-topic subject, what you said is complete nonsense. The pre-bias of the researcher is not grounds for dismissal of a study UNLESS the pre-bias directly affects the results. Show me a researcher without bias and i’ll show you a liar. Its called a hypothesis. Researchers tackle subjects that interest them, and subjects that interest people have inherent bias. Numerous studies - including massively funded, accepted and peer reviewed studies - were originated because the research team had a hypothesis that they were attempting to prove was true. If the data set as large as this one (40 years) and shows a consistent result, it is accepted regardless of intent. It would be peer reviewed and the data would stand on its own. In the vast majority of situations, the coach and GM are hired concurrently. And for many successful coaches, they ARE the gm. Belichick, Carrol, Reid (who hand picked his own GM), Payton (who is said to have control over personnel), etc. .

There are 16 current head coaches who have been in the playoffs. McD is 13th among them, in playoff win percentage.

The fact that there are so many different variables and the data is still so consistent is what makes it so remarkable. 4.2 years is the average. You have no basis for this assertion. This is a 40 year data set that proves the assertion. The data disproves this.

Belichick didn’t have a franchise QB in Cleveland.

The fact that there are so many different variables and the data is still so consistent is what makes it so remarkable.

I see that data analysis is not your strong suit.

You would rather be the bridesmaid for a long time but never the bride… I just don’t understand that rational.

Great question. The answer is that the numbers are likely to be very close to the same, as many of the coaches have both won and lost Super Bowls.

Reid made it in his 6th season.

And when it does it typically happens within 7 years.

Great ideas for further research.

That’s fair.

Agreed, but those numerous variables and unique circumstances seem to point to a remarkably consistent conclusion when examining the data over a long span of time. Can a coach past year 7 be the outlier? Sure. But that requires faith in a significantly low percentage of success (8%).

What the data shows is that point in time typically happens within a coaches first 7 years.