First, let me say great job of data mining. Appreciate the work. Very helpful to this discussion. The numbers are what they are, but they don’t speak for themselves. They have to be interpreted. And you’ve overstated the significance of these numbers in your interpretation, perhaps knowingly for emphasis, in order to make your point.
If what you are trying to do is determine whether a bus ride of 2+ hrs. is a more (or less) significant disadvantage than a shorter ride, simply using won-loss records won’t do that. For example, Team A travels 2 1/2 hrs. to play Team B. On a neutral field, Team B is 10 pts. better than Team A. They play the game, and Team B wins by 21 pts. The next week, Team A travels 15 minutes to play Team C. On a neutral field, Team C is 10 pts. better than Team A. They play the game, and Team C wins by 1 pt.
These data support, but don’t “prove,” that as for Team A, a long bus ride causes them to perform more poorly than a short bus ride, even though they lost both games. Of course, you’d need a lot more outcomes to be able to talk about “proving” something. But the principle is the same. If you want to find out how travel times affect performance, you have to measure the difference between the actual performance after travel compared to the expected performance over travel times. I think this illustration shows how difficult it would be to measure how much travel affects performance, since there are way too many variables involved.
Now it’s about time for someone to talk about “common sense” … which is often neither common nor accurate.