[doughhater]

I don't have a ton of experience with correlated parlays, but I believe that you have to look at this from a conditional perspective. IF the game goes under, THEN the dog is obviously a huge favorite to cover. IF the FAV covers, THEN the game is a huge favorite to go over. However, I think that this correlation only exists in such an extreme example (have you ever seen a 30+ pt favorite and a 33 pt o/u). Or perhaps the correlation just becomes weaker as the examples become more realistic.