For the first 4 weeks, I’ve used Moneyline probabilities for wins. I also really like using Advanced NFL Stats Win Probability model. It’s based off data, versus public perception, which has its pros and cons. The ANS probability model for week 5 just came through.
What this does is add some colors to my tables… I put blue as ANS “People’s Upset” pick, yellow as Moneyline ones, and green are both ANS and Moneyline “People’s Upset” picks. Both showed Cincinnati over NE as a people’s upset, so I’m going to stick with that for my pick. For my dad’s pick of Chi over NO, the ANS model shows NO with a 66% win probability, so no upset there. Hence, I’m going to change my dad’s pick to Tennessee over Kansas City, since ANS gave it a 57% win prob and the moneyline wasn’t too far off at 40%. Only 14% of the public picked Ten. So here’s the new chart for Dad:

To recap: We now want Tennessee to “upset” Kansas City. I also hedged NO a bit, choosing them for 3 points instead of Chi.