First 2 weeks totals

Over the summer I decided to do some research. I wondered how fast offenses and defenses come together to start the season. Did one have the advantage over the other? It seemed to me that a good place to look would be unders and overs. So I looked at totals for the first 2 weeks of the season, going back 6 years. Below are the numbers:

----------------Week 1----------------------Week 2
------------under -------over -------under -------over -------total under ---total over ---Win % ----units won
2009-------8-------------8-------------7------------9-------------15---------------17--------------------------- -3.7
2008 ------9-------------7------------ 7------------7-------------16---------------14--------------------------  +0.6
2007 -----10 -----------6 -----------10 -----------6 ------------20---------------12--------------------------- +6.8
2006 -----12 -----------4 ------------7------------8-------------19----------------12-------------------------- +5.8
2005 ------9------------6------------12------------4-------------21----------------10------------------------ +10.0
2004 ------9------------7------------10------------6-------------19----------------13-------------------------- +4.7

-----------------------------------------------------------------------110---------------78-----------58.5------- +24.2

In the 12 weeks that were checked, 8 of them showed that under the total was the way to go. All in all, 8 winning weeks, 4 losing weeks ( I'm counting the 2 tying weeks as losses of course.) In fact, every season for the last 6 shows that if someone blindly bet the under in every game for the first 2 weeks, they’d have won. Over those 6 seasons, this blind system hits at 58.5% for a profit of 24.2 units. Of course, the catch is making 16 bets a week blindly can lead to devastating losses if (and when!) you run into a bad week. Sixteen blind bets is too risky.

However, it can help point us in the right direction. Simply put: You should tend to look for games to bet under. This is a concept that I haven’t seen discussed much. When you pick a game at random, it’s 50-50. Imagine a box with 100 marbles in it, 50 black and 50 white. Every time you reach into the box and pull out a black marble, you win your bet. A white one, you lose. When someone handicaps a game, what he’s doing is finding reasons that make him believe that the odds are better than 50-50. As long as he can pick 53% or better, he’ll win. As he does his research, he finds reasons to bet on a certain team. When he gets to a certain point, say where he feels that his pick has a 60% chance of being right, he makes his bet. In his mind, the selection isn’t 50-50. The selection is 60-40. So, in effect, after all is said and done, he’s reaching into a box with 60 black marbles and only 40 white ones.

So, using what we know about the first 2 weeks of the season, we can say, at least for the past 6 seasons, the box of marbles isn’t 50-50. The box has 58 black marbles in it, representing unders, and 42 white ones, representing overs. What this means is, if the trend continues, betting on the under is much preferred for the first 2 weeks. Betting on the over would lead to serious trouble. Would you want to reach into that box looking for a white marble, if there were only 42 white ones out of 100? Clearly, given this information, you’d bet a game under or not at all in the first 2 weeks.