My method of predicting the all-star game is slightly different from my method of predicting other major sporting events. The numbers the players have on their jerseys is irrelevant.
If I watch the game: the National League has a chance of winning
If I don’t (like I haven’t post-1996): The American League will win.
Bet on the American League. My weekly writer’s group still meets on Tuesdays.
My methods of predicting the All-Star Game are simpler:
1. Read the newspaper the next morning. Chances are the game will have been over before the paper went to press. Look at the box-score for a big hint.
2. Look to see if the Major Leagues are on Strike. If not, then there most definitely (99+% certainty) will be an All-Star Game. I empirically fitted this model to the data, and discovered an extremely reliable fit of the model to the data.
3. But, sadly, there are no Stars at an All-Star Game — the lights are so bright that you could not see a star, even if it was shining all night — maybe not even if it went super-nova on us. As long as the shining star is not Sol, though, we are okay, so we think.
1)
Your definition of prediction is a bit different from my own. Still, considering this year’s all-star game went 15 innings, method #1 may not have worked in every town. However…(and this is a big however)…if you read your news online, it doesn’t matter where you live, method #1 would have worked.
Since I didn’t watch the game, I had to use method #1 to find out if my prediction was correct.
2)
The existence of an all-star game doesn’t require a winner. Alas, there have been tied games, and this year’s almost ended up in that category.
3)
Finally…your suggestion in #3 that just because you can’t see something means it doesn’t exist is worrisome.
So it goes…