rhickok1109 wrote:Rupert Patrick wrote:ChrisBabcock wrote:This just popped into my head and I'll admit I'm just throwing a dart with this one.
Within the next 20 years we'll have our first NFL regular season that has every team's record between 6-10 and 10-6.
Won't happen. Ever. Even in true parity, where every team was equal in every single way, due to the law of averages, it is very, very unlikely that it would ever happen. For one thing, there is a point somewhere around 25 percent where on any given Sunday any team on average can beat any other team due to unforseen flukes, such as a gimme last second field goal is shanked, or one team has a statistical advantage over the other team but still loses the game, or an injury to a key player that causes the favored team to win, or some player on the underdog team has the game of his life, or a tipped pass that turns into the winning score, or any one of a hundred other reasons. Sometimes, it is something as small as a missed block that allows a long touchdown run.
I know this 25 percent range for flukes is true because no matter how smart the so-called experts are, none of them can pick the winners of all regular season pro football games before the fact at better than a 75 percent clip over the course of the regular season. No matter how you pick them, no matter what type of system you use, there is a wall where you cannot pick more than 75 percent of the winners week after week. With that being said, if you are picking good vs. bad games, a good team (say, New England) against a bad team (say, Cleveland), those probabilities are much higher. At that point, the probability of picking the winner goes up to about 90 percent. But in any given week, there are usually three or four games which are total toss-ups, and it is how you pick those which really decides how good you are at picking games.
I think it tells more about how lucky you are at picking games

That is true. This is why I hate picking the winners of games. I do pick the winner of the Super Bowl every year, and it is difficult, and even studying the teams very closely I can pick up things that I see that led me to state that teams who were rated as underdogs in the game should have actually been favored to win, as in the case of Seattle over Denver in Super Bowl XLIX and Phiadelphia over New England in Super Bowl LII. Even with that, I was wrong about predicting the Panthers to beat the Broncos, and I picked Seattle but the Seahawks-Patriots game was decided by a fluky play at the end.
Getting back to the original statement about all the teams in the league being between 6-10 and 10-6, I decided to examine the issue at a statistical level. Looking at teams in the 16-game era, between 1978 and 2017, removing strike seasons, and dropping out all teams with tie games on their records, approximately 55.275 percent of all teams finished either 6-10, 7-9, 8-8, 9-7 or 10-6. The probability of all 32 teams finishing between 6-10 and 10-6 in a given season should therefore be .55275 to the 32nd power or 5.76668E-09 or 173,410,061 to 1. That's sort of a rough estimate.
Even the probability of a season with all 32 teams between 4-12 and 12-4 is over 100 to one.
"Every time you lose, you die a little bit. You die inside. Not all your organs, maybe just your liver." - George Allen