I think you have to make a wager when you have an edge in information. Last night, I took the under 8.5 after reading about info on the home plate umpire being 12 of 17 to the under. He obviously has a big strike zone and forces batters to swing at bad pitches, which results in more strikeouts, weak ground balls and pop ups. Did not win because of the rainout. Look for an edge on a side or total before making a bet!! Public on Sunday night games will play the over, so a bigger advantage to the under player last night.
I'm not sure knowing a small sample trend that is widely available constitutes having an edge in information. If the information is easily attainable, it's likely already factored into the line. A way of checking this is looking at the line history and there was no movement towards the under.
Your second point has some merit, especially if you have square shops in your rotation. I had my proxy/beard bet u8.5 at Sportsbook because it was +100 when the consensus no vig was -105. Pinny was shaded even more at -107