Dave George over at PBP wrote this piece (LINK) with an interesting question about HC's and their playoff records/performances.
"For all of his success as the NFL’s winningest coach, Don Shula’s overall record in the playoffs was barely above .500 at 19-17, and that’s with Hall of Famers Johnny Unitas, Bob Griese and Dan Marino at quarterback for most of those games...
What we’re trying to say here, and what every Dolphins fan should understand, is that the playoffs are an exceedingly cruel and unpredictable environment, and that goes for everybody.
Adam Gase will try to communicate that to the Dolphins this week as they prepare to face Pittsburgh in Miami’s first playoff game since 2008. Some things, however, just have to be experienced to be believed.
“It’s another level,” Gase said Monday. “They’ll know it. They’ll know it when they hit the field. Trust me. “ ...
Consider how much Gase already has seen in his NFL career.
A playoff win in overtime with Tim Tebow, back when Gase was the Broncos’ quarterbacks coach. Two sudden dead ends when Peyton Manning was quarterbacking the Broncos but they lost their opening playoff game just the same.
Or how about 2013, the first of two seasons for Gase directly collaborating with Peyton on play-calling? Denver rolled all the way to the Super Bowl that year with the league’s No. 1 offense in all the big categories only to get smashed 43-8 by Seattle...
On the other sideline Sunday at Heinz Field will be Steelers coach Mike Tomlin. He’s taken Pittsburgh to a couple of Super Bowls, sure, and won one at the age of 36, but overall Tomlin is 6-5 in postseason games. What’s more, half of his playoff teams have lost their opening game.
It’s another level, all right, and with confidence levels rising and falling with each crucial snap. No reason Miami should be counted out of anything, whether there’s a ton of playoff experience on the roster or not..."
Good reminder, but to me that's just as much reason to temper excitement as to raise it. That is, to count the Dolphins neither in nor out. We'll just have to wait and see.
Page 1 of 2
-
-
So that means we're going to win the Super Bowl right?...
Sent from my XT1254 using TapatalkColmax likes this. -
We are walking into a bit of a buzz saw here, the Steelers haven't lost a game in about 2 months, the last lost was Nov 13 in a shootout with Dal, 35-30.
What's left of our crappy LBs vs Bell?
Whats left of our crappy secondary vs AB and Ben?
MM will have to play lights out for us to win, but I do think he does give us our best chance. -
-
Thanks for posting that! Anything can happen, but hopefully it will temper everyone's expectations.
-
That means out of 41 seasons 23 of those seasons a 1st seed has won the SB, or 56%. Second seed is worse but decent: 4 each for AFC and NFC, meaning that a 3rd-6th seed in either conference wins it only 1 out of 4 times.
As far as our situation.. only once has a 6th seed in the AFC won the SB: Pitt in 2005. Oh.. and interestingly, never has a 5th seed in the AFC won the SB.Unlucky 13 and danmarino like this. -
In other words, there's a 19% baseline chance the Pats win the Super Bowl this year. There's an 81% chance that one of the other 11 teams wins it. The numbers are nowhere as dominant as people make them out to be condemning the lower seeds. -
My reasoning was this: in the playoffs I thought teams were more evenly matched, meaning that the probability a 1st seed would win the SB would be closer to random chance (= 1/8 probability for 3 wins in a row = 12.5%) than it actually is. So random chance would have predicted the top 2 seeds would win 25% of the SB's. I thought maybe the actual figure was 30% or even 35% but 56%???
Can't speak for anyone else but that is way above expected. Also, it was way below expected that an AFC 5th or 6th seed won it only once in 41 years.
So I guess it's all relative to expectations.Last edited: Jan 3, 2017P h i N s A N i T y likes this. -
cbrad likes this.
-
In any case, if one is going only by those historical numbers, we can say the Patriots as AFC 1st seed are 8x more likely to win the SB than we are.danmarino likes this. -
btw.. just for reference, here's the entire list of the number of teams based on seed that won the SB since 1975 where "A" refers to AFC and "N" refers to NFC.. numbers after A and N are seeds:
A1 = 8
A2 = 4
A3 = 1
A4 = 4
A5 = 0
A6 = 1
N1 = 15
N2 = 4
N3 = 1
N4 = 1
N5 = 1
N6 = 1
.. really weird how that AFC 4th seed has so many SB wins eh?P h i N s A N i T y, resnor and danmarino like this. -
Nevertheless, I think one can make your argument stronger by just looking at the salary cap era. As you might expect the #1 seed shouldn't be as dominant in the salary cap era and that's what you see from 1994:
A1 = 4
A2 = 3
A3 = 1
A4 = 3
A5 = 0
A6 = 1
N1 = 6
N2 = 1
N3 = 0
N4 = 1
N5 = 1
N6 = 1
#1 seeds have won 10/22 = 45% of the SB's since 1994, instead of 13/19 = 68% before the salary cap era (from 1975). And the #6 AFC seed would have a probability of 1/22 = 4.5%. Then one might also note that while the salary cap era started in 1994 teams didn't suddenly change overnight in strength, which is why a #1 seed won the SB in 1994, 1995, 1996, 1998 and 1999 (so 5 out of first 6 years after the salary cap started). So there are good reasons to think the probability of us winning the SB based solely on history is more like 5% or so instead of 2.5%.Last edited: Jan 3, 2017KeyFin likes this. -
-
Back in the day, roulette tables at Vegas casinos did not have a board to keep track of previous winning numbers. Then one day, some brilliant SOB came up with the idea of tracking those numbers and displaying them to the patrons of the casino. That gentleman is a hero to all casinos because he understood what most people do not: statistics are not predictive.
If you're playing roulette and the board shows that the previous 10 rolls have been on red, most people assume that somehow moves the probability. Let's assume the previous 100 rolls have all been red. Whats the probability of the next roll being black? It's still 50%.
The moral of the story here is that history can not be trusted and using past statistics to predict future outcome is 100% garbage. Statistics are not predictive. -
I've never understood people looking at historical stats to apply to upcoming games. I couldn't care less if Team A is 3-15 over the past 9 years against Team B, that really has no bearing whatsoever on their upcoming game.
Fame likes this. -
And you're comparing apples to oranges here. Everyone understands the "coin flip" scenario. A football game isn't close to a coin flip or roulette table. -
For this debate.. I think the predictive power is moderate. I mean.. there's a reason the #1 seed in each conference has won so many more SB's. They are generally the best teams in the conference that year! You also have home field advantage AND they play one less game than the WC teams. So there's a mechanistic explanation behind this. With roulette wheels the physical mechanism essentially ensures each slot has equal probability. I mean that's how it's constructed!! Definitely not so with most applications of stats to sports we've discussed here. -
I don't mean to sound like a broken record here, but truly ANYTHING can happen with the 12 teams remaining. Will the Fins win it all? Probably not. But we're as good as anyone with our "next game up" mentality and we've consistently beaten teams that Vegas said we'd lose to.cbrad likes this. -
-
You can make a guess using past outcomes, but that is merely the assumption that the trend would continue. It is not a prediction.
By the way, if you go to Vegas I would suggest you avoid the roulette table. -
The simple fact of the matter is you can not make predictions based solely on past data measured on entirely different teams in different circumstances over different years. If the Pats suddenly lost Tom Brady to an injury do you think the past 15 years of playoff data would ease the concerns of Pats fans? -
I can predict, due to past outcomes, the future outcomes of #1 seeded teams in the playoffs. Why? Because the "home factor" (among other things) influences football games. How do I know this? Because a lot of data (hundreds of football games) proves this to be true. Home field, weather, injuries (as you talked about in you last post to cbrad) all affect football games. Injuries, weather, home field do not affect roulette.
Now, as I wrote earlier, we currently have no way of predicting a football game with 100% certainty, but we can (or at least I can) predict with more certainty playoff outcomes using past playoffs.
By the way, if you want to bet $10,000 on the Dolphins to go all the way...I'll take that bet. If you don't, why not? I mean, you have just as great a chance of winning picking them over, say, the Patriots. Right?cbrad likes this. -
The probabilities/odds are simply useful in estimating how difficult their task is (and for some others.. for betting purposes).Last edited: Jan 4, 2017danmarino likes this. -
Same with football: the question isn't whether two situations are exactly the same. The question is simply how much predictive power you can obtain from what is similar (because there are almost always SOME things that stay the same or approximately the same).
The structure (determined by the league) of the NFL playoffs is similar this year when compared to previous years in the sense that the WC teams will have to play an extra game to win a SB compared to a #1 seed. That and the fact that the #1 seed is guaranteed home field advantage means there is a built-in disadvantage for teams that play in WC games. And as I showed before, changing the structure from pre-salary cap era to salary cap era affected the percentage of #1 seeds that won the SB (their dominance went down, as expected).
These aren't all "unpredictable" the way you're describing it. The example you gave with the roulette wheel is one of the rare cases you find in real life where you can expect all possibilities to have the same probability. As soon as you can show one possibility has the highest probability, that's the one you pick to "predict" what occurs.
The history of science and technology speaks VERY loudly against your claim that you can't make predictions from past data if the circumstances are different because not ALL circumstances are different!danmarino likes this. -
As interesting as the ones about who wins the Super Bowl are, I'm actually much more interested in a similar one about who makes it to the Super Bowl (since one team from each conference does every season), and who makes it to the conference titles games, as thats almost always the sign of a very good team.
And speaking more to our own situation, I would argue that with the current playoff and divisional setups, getting the #1 or #2 seed is more important to Miami than almost any other club. While others may (and will) disagree, our club just does not generally play well in cold weather, and especially in January. The Dolphins are therefore less likely to win as the #5 or #6 seed, even in the first round, than most other teams.
I'm also of the opinion that one of the reasons that the AFC has had less parity than the NFC since around the time that the league started the regular season later, an pushed back the start of the playoffs, is due to geography and weather.
The AFC has nine teams that play outdoors in cold weather climates, two that play in mild climates (and TEN, who I include here, can be cold sometimes), two that play in domes, and three that play outside in warm weather.
In the NFC, only five teams play outdoors in cold weather, three play outdoors in mild climates, six (soon seven) play in domes, and two play outdoors in warm climates.
I plan on digging into the playoff histories of the clubs deeper soon, but at a casual glance, its clear that in the expanded playoff ear of 1990-present, the warm weather clubs in the NFC had done much, much better than the ones in the AFC.danmarino likes this. -
-
If you were watching a football game and before the broadcast the announcers took a moment to talk about the fact that the (hypothetical) Dolphins have beaten the Browns in each of their last 10 meetings, would you then conclude that the Dolphins, based on this information, had a greater chance then of beating the Browns in their 11th meeting? Based purely on past games, ignoring the current situation, would you conclude the Dolphins were overwhelming favorites? -
A1 22
A2 10
A3 1
A4 6
A5 1
A6 1
N1 24
N2 9
N3 3
N4 3
N5 1
N6 1
Don't have the comparable data for getting to the conference championships.. you'll have to do that yourself haha!
Unlucky 13 likes this. -
Page 1 of 2