## Superbowl and Falsifiability

Posted by allzermalmer on March 20, 2013

There was a Superbowl a couple of months ago. Now keep in mind that with the rules of NFL and Superbowl, no Superbowl can end in a tie. One of the two teams must win the game. But let us take a look at falsifiability and unfalsifiability.

Suppose that the Superbowl is only between the Baltimore Ravens and the San Fransisco 49ers. You must predict who will win the game.

We can make the claim “The team with the most points will win”. Now this statement is true and cannot be false. This statement is necessarily true. “The team with the fewest points will lose”. This statement is also true and cannot be false, so it is necessarily true. We cannot show that this statement is false, since it is true, so it is unfalsifiabile.

We may now take a further step and predict that either the Ravens will win. This is falsifiable, since we can watch the game and see if the Ravens won or lost. Further, we make take the opposite position and predict that the 49ers will win. We know that either the Ravens win or the 49ers win. Only one of these options can be right.

But we may take a further step and predict who will win the game and what the score will be. We can predict that the Ravens will win and that they will win 14 to 7. We may also do the same with the 49ers. But now we are getting more specific with our predictions. It makes it easier and easier to show that our prediction is wrong. The score could be that the 49er won 14-7, which shows the prediction of Ravens winning 14-7.

Let us further assume that the total amount of points that can be scored in the game is 40 points, and there are a total of 2 teams. We may use the formula of computational probability, nCr = ( n! ) / (r! (n-r)! ). n= total number of subjects, r= number of objects in arrangement, != factorial. Through computation of (40!)/(2!(40-2)!)=780 different possible combination of the =score. This means we can make 780 different predictions. For example, Ravens win 14-7 or 49ers win 14-7, and etc. Maybe the Ravens win 21-14 or the 49ers win 21-14. There are 780 different specific predictions we can make which contain (1) the team that wins of the two, and (2) the score.

Now if we pay attention, we start from the most general and further move down from less general to more and more precise. We can start with a general prediction of “The team with the most points will win”. It is true but it is not falsifiable. We may further move down with a more precise prediction of “The Ravens will win”, which is falsifiable. We may further move down with a more precise prediction of “The Ravens will win 14-7”. Each stage down we go, the claim is more and more falsifiable. It eliminates other possible outcomes, and can be shown if one of these other possible outcome were to be obtained.

We have two general claims that we know are necessarily true and not informative about who will win the game between the Ravens and 49ers, which is 100% you will get the answer correct. We have two further general claims which are not necessarily true and are informative. This claim is general and can be shown to be right or wrong, but it 50% you will get the answer correct. When we further move down to what the score will be, then we have 0.00128% of getting the answer correct. We have 780 possible correct predictions, and only one of them can be correct, so we only have a 0.00128% of getting the correct prediction.

We may make a further specification that you must pick the winner, the points, and how they obtained those points. For example, the only ways to obtain points would be (1) 6 for a touchdown, (2) 3 for field goal, (3) 2 for safety, (4) 1 for extra point. This now means one may say, “The Ravens will win 28-14, with Ravens getting 3 touchdowns, 3 field goals, 1 extra point.” Things are becoming more and more specific, which makes it easier and easier to show it is false.

Suppose we have these claims: (1) All Ravens on Earth are Black, (2) All Ravens on the Northern Hemisphere are Black, (3) All Ravens in North America are Black, (4) All Ravens in the United States are black, (5) All Ravens in California are Black, (6) All Ravens in Los Angeles are black.

These claims are based on the most general of them all and is moving down to more and more precise general claims. We know that If all ravens on earth are Black and all Ravens in Los Angeles are not black, then we know that all ravens on earth are not black. But if all ravens on earth are not black, that does not mean that all ravens in Los Angeles are not black. All the Ravens in Los Angeles can still be black.

## Leave a Reply