NFL Spread as Straight-Up Predictor
It makes sense that if you are picking straight-up NFL winners, then take the favored team. Is this accurate? Well, yes but only about 2 out of 3 times. Thats like going 11-5 every week in the office pool. Maybe good enough to win over a year but not good enough to win the week.
I am developing a system for picking straight up NFL winners and in the process have gleaned a few interesting items. There have been 224 games through week 15 with one tie; the Bengals - Eagles game. So of the 223 games with winners and losers I see the following:
2008-2009 Season Favorites The Home favorite is 105 - 49 (68%) The Visiting favorite is 47 - 22 (68%) Therefore, the home field advantage is taken into account in the spread already.
In Division Favorites The Home favorite is 39 - 19 (67%) The Visiting favorite is 18 - 6 (75%) Home favorite is not impacted by being an in-Division or out-of-Division game, but the Visiting favorite is getting an extra 7% accuracy. However, the pool of data is low. Lets see what happens when we add in last season. ** 2007-2009 Season (almost 2 full seasons)** Favorites The Home favorite is 210 - 100 (68%) SAME The Visiting favorite is 120 - 49 (71%) SAME ** In Division Favorites** The Home favorite is 69 - 47 (59%) LOWER The Visiting favorite is 48 - 14 (77%) Confirmed The Visiting favorite IS getting more accurate by 7% to 9%. The Visiting favorite wins 3 out of 4 times.
My next study will take into account geographical locations like west coast vs east coast teams.