Wow is that not true. You can't test the concept that a coin flip is 50/50 with only a few tosses.
Good post about the away team points diff (PD).
The test criteria is relevant to test requirements.
If you said the coin will always be heads, then I only need 1 fail to prove otherwise.
There have been claims that PD predicts playoff winners. The primary claim by another poster, IIRC was 18 0f 20 in his scenario. That's 90%. Using his PD>100 criteria, the 2018 data without the bye advantage was 0 of 2. Including the bye advantage winners it was 4 of 6 with the 4 being the 4 "predicted" by the bye advantage alone. That's a 6 game data set where PD>100 was applicable and the results were not 90%.
Even if we treat a complete season of data as 1 "coin flip" (we really had 6 coin flips) then with a 90% expected success rate it's likely that randomly choosing 2018 and getting a fail, that the concept is not going to result is a 90% prediction success rate. Obviously I would need 1 more fail in 10 years of data to say with certainty that it fails the 90% criteria.
My point was to show the concept that PD alone predicts playoff wins concept is likely due to people confusing cause and effect (i.e. Teams often run more in the 4th quarter when they're winning. It's unlikely that correlation of 4Q rush attempts means that high 4Q rush attempts is a "cause" of winning).
By definition teams with high winning percentage are most likely to get a bye in the wildcard round and home field in the division round against teams that had to play the wildcard round.
PD, like win percentage should correlate to the teams getting the bye and home field against teams that played the extra game.
Going into the playoffs we know which teams have a bye and and should know which teams will have home field in the first 2 rounds.
The bye week and home field are likely a "cause" in regards to winning.
Any correlation of PD with winning is likely the "effect" of teams with higher PD likely (by design) getting the bye and/or home field.
Summary:
High win percentage teams tend to have higher PD. High win percentage teams tend to get the bye and home field. Teams with the bye and home field tend to win those games.
Does PD alone tell us anything other than teams with high PD tend to have high win percentage and high win percentage teams tend (by design) to get the bye and home field in the first two rounds?