Answer:
This is not a valid method because there is an implicit bias in the sampling method. This is because people at a basketball game are more likely to enjoy basketball than those who don't attend.
A better way to determine the actual most popular sport would be to randomly sample students during the school day. This would give you a better representative sample.