Einstein

I have been invited to speak at an upcoming conference in Las Vegas, but I must admit that fashion and style are not my areas of expertise. Can you suggest any local shops or boutiques where I can find assistance in creating a stylish ensemble? All I have right now is the office-standard mix of blue/black/brown suits.

Haha, yeah. There is a chance that Allen drags McDermott kicking and screaming to a Super Bowl like Manning dragged Dungy.

Technically the sample includes several hundred, since the probability is calculated by comparing all coaches to the coaches who made the Super Bowl. This is called "shifting the burden of proof." The burden of proof lies with the accuser - you. You are responsible for presenting evidence and arguments to support your claim that I used sampling bias. When you shifts the burden of proof onto the accused, you are essentially demanding that the other party prove their innocence instead. Which is… insane. You can’t show a single shred of evidence that I used sampling bias because there is none. Which is why you continue to regurgitate that comment with any proof.

Again. Where is the proof? You have not shown a single shred of evidence that I did what you are accusing me of. You’re just regurgitating the same unproven assertion. You can variable any meaningful stat into extinction. For example. Why count INT’s? There are many variables for why a QB may have thrown one.

I agree. I think you and I more alike than many others here, even if our methodologies differ.

Not at all. I mentioned in my original post that I think McD deserves this year to move the needle and advance to the Super Bowl. Firing him at this moment would not be productive. The point of this thread is to ascertain the leash that McDermott should have going forward. So let me ask you - how much longer would YOU give him? Indefinitely? Would you be willing to spend Allen's entire career waiting for McDermott to figure it out and not make mistakes in the playoffs? 2 more years? 3 more yeas? What, in your opinion, should be McDermott's leash?

Ok, this is progress. Before you claimed confirmation bias (but with no proof). Now you're claiming a calculation error. Where is the exact calculation error? Woh! I'm not upset at all! I know it's not your fault that you and others don't understand. Now we are back to confirmation bias. Where is the proof? You say this, but provide not a single shred of proof. Because my goal was to ascertain McDermott's chance of making a Super Bowl. That is the next step. He has made an AFC Championship game, the next step is to at least make the Super Bowl. Though winning would be great as well. Who wants to tell him? lol!

It's really not. I respect your architectural knowledge greatly but you're out of your domain here. For example, when our company conducts focus groups, we do it in groups of 10 or 12 people. Because that is all you need for a sampling of the entire United States demographics. To be fair, it is typically done in 5 rounds, so it's more like 50 to 60 people, but it is still a small (to you) sample size. Another example is Nielsen and their "people meters" process where they only collect data from 0.03% of US households yet estimate the data to over 120 MILLION. Augie told me to post my tax return. I did, for fun. Then several posters became jealous of my financial situation

Now that’s true. I still chuckle about you thinking simple calculus needs to be peer reviewed 😂. I wasn’t the only poster who called you out on that nonsense.

A very consistent coincidence over 40 years of Super Bowls… I’m still open to you proving this, but every time I call you out on it, you simply repeat “sampling bias!”. That’s not proof - that’s simply a broken record. .

Do what exactly? Are you asking to go back 50-60 years instead of just 40?

All valid questions that can form a hypothesis.

Because we are trying to identify when coaches are most ready to make a Super Bowl run. To be honest, I don’t think the results will be much different when looking at Super Bowl wins, but I can run the analysis for you if you’d like. A studies data simply provides the answer. This is followed by individuals proposing various hypotheses to explain the phenomenon. For instance, one hypothesis might suggest that as coaches extend their tenure in the NFL, teams decode their strategies and understand their gameplay, thus causing a decrease in their effectiveness over time. However, this is just one possible explanation and may not necessarily be the correct one.

When an expert engages in conversations with laypeople, it is not uncommon for the latter to feel confident in their own correctness and perceive the experts as being mistaken. I have grown accustomed to this dynamic.

Too many variables. Can’t predict.

Haha didn’t intend to make you feel bad. Most people do not understand probabilities and odds. The evidence of which is glaring throughout this thread.

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

If that dice landed on the same number 35 times you may be on to something with that dice being weighted toward a certain outcome.

Understood, but It is 40 years of data and it is not n=1. When those 50~ results come from a pool of over several hundred coaches, it has some meaning as it shows who out of that pool of several hundred (the real sample) was able to make it. But I do understand that there will always be *some* reason for results we do not like to be discredited. I blame lab error for my cholesterol numbers.

This is true. We just have to hope that McDermott is an outlier.

Sorry but your math is wrong. Odds (for example, the ones you listed at 10 to 1, 5 to 1, 4 to 1) do not translate directly into percentages (10%, 20%, 25%). Odds and probabilities are related, but they're not the same thing. Odds are a ratio of the probability of an event happening to the probability of it not happening, while probability is a ratio of the probability of an event happening to all possible outcomes. In other words, 10 to 1 odds correspond to a probability of 1/11 (9.09%), 5 to 1 odds correspond to a probability of 1/6 (16.67%) and 4 to 1 odds correspond to a probability of 1/5 (20%). So we have 0.9091 * 0.8333 * 0.8 = 0.604 then 1 - 0.604 = 0.396 (39.6%) The probability of making the Super Bowl one time in the past 3 years (given Vegas odds) are 39.6% .

I disagree with the counter point but I can understand it and I think there is some validity to it.. Every person can do with the data what they wish (including ignoring it).

Bingo. That’s how (almost) literally every forecasting or grading calculation in sports works. It’s all past data. The people getting upset simply don’t like the conclusion, so they grasp at straws to find ways to discredit it. I’m starting to realize that people on this forum simply don’t understand how probabilities work. We find probabilities from PAST EVENTS. That is literally how it works! We don’t have a magic fortune teller ball. We use past data! - Financial analysts use historical stock data, market trends, and other indicators to estimate the probabilities of future stock price movements or market trends. - Transportation planners use historical traffic data, road conditions, and other variables to estimate the probabilities of traffic congestion in different areas and at specific times. - Seismologists, meteorologists, and other scientists use historical data linked to monitoring systems to assess the probabilities of earthquakes, hurricanes, floods, and other natural disasters occurring in specific regions. Probabilities require strong data unless you have a very simple set (such as a 50:50 coin), If you prefer, you can substitute “probability” with “forecasted probability”.

Indeed, that's why information such as this thread can prove valuable. We've seen from historical data that only five coaches have reached their inaugural Super Bowl after their seventh year of coaching in the past four decades. It's hard to anticipate the trajectory of a career, but if McDermott ends up making it to the Super Bowl, he would (after this year) be an anomaly within a vast data set.