…the next pick to the people who saw you pick the “winner”. Now half of those people see one team, the other half see you pick the other team, and whoever saw you pick the winner thinks you’ve got a 100% accuracy rate over two games. You could do that for a while and then offer to sell your pick for the Superbowl. Starting with a big enough group in the beginning, this might be really lucrative.
But is it legal?
You forgot about ties. They’re rare, but they happen, and in this scenario they work like the 0 in Roulette - they fuck over your nice and comfy 50/50 chance.
And as others already mentioned: I’m pretty sure that whole scheme wohl just be plain fraud.
According to Wikipedia, there were 7 tied games in the last 8 years (2017 - 2023).
Those 7 games appeared in 5 out 8 seasons, which results in a chance of 62,5% to catch a draw within a season.
Or with 272 games per season, the draw chance per game is 0,3%. So no draw chance is 99.7%.
Hence you get a chance of 1-(99.7%^272) = 56% to catch at least one draw per season.
This assumes the 2023 season is finished (without a draw), which I think is wrong. So the odds for a draw could be slightly higher.
You could make a more informed split to increase your odds. Say a weak team plays against a strong team, and official sports betting offices rate the chances 30/70. Instead of splitting the two groups 50/50 you now split them 30/70 as well.
Emails (gmail at least) can also dynamically display information. So you could just change a wrong guess to a right one after the fact.