Through Monday night, there had been 32 NFL games played. 18 of those games had stayed under the total, 13 went over, and one hit the number on the nose. Betting the under would, up to this point, be profitable. Of course, it's not as easy as that - or else we would all be rich. By delving deeper into the numbers we can see what there is to be learned to make us more successful total bettors. Here's the breakdown:

Week-by-week: In the first week of the season there were 11 unders in 16 games, likely reflecting the oddsmakers opinion that the public will happily take the over regardless of how high it is set out of the gate. Last year 10 of the first 16 games went under, so this is a consistent trend. The second week is also remarkably consistent between last year and this year. Last year there were six unders and two pushes in the 15 games played on Sunday. This year we saw the same six unders and just one push. Though it would be tempting to assume that the oddsmakers adjust their approach from week to week to force these outcomes, the numbers don't play this out - the average total was 40.75 in the first week, and was higher (42.3) in week two.

If you believe that history repeats itself then you will be interested to know that there were five unders and a push in week three last year.

The over machines: Six teams have gone over in both of their games this season - Dallas, New England, Cincinnati, Cleveland, Oakland and the Giants. It doesn't take much detective work to figure out how those six teams ended up on this list. Dallas, New England, and Cincinnati have the three most productive offenses in the league, and Cleveland, Oakland and the Giants are all in the bottom five for defensive effectiveness. Cincinnati is also helped by the fact that they have a bottom five defense. There may not be a total high enough to put them under - both they and their opponents, the Browns, went over the game total by themselves on Sunday. The recipe for consistent overs is pretty simple - the ability to score a lot of points, and an inability to keep the other guys out of the endzone. Only New England doesn't fit that pattern, but there offense is working so efficiently that they are able to significantly exceed the expectation of oddsmakers.

Always under: Showing that it is harder to score points than people think it is, there are 10 teams so far that have been unable to exceed their total in two tries - Tennessee, Minnesota, Indianapolis, San Francisco, Atlanta, Buffalo, Kansas City, Chicago, Jacksonville and St. Louis. Though these teams have achieved their unders in different ways (some through competence, and some through total incompetence), there are some obvious trends that several teams share. Minnesota, San Francisco, Atlanta, Buffalo, Kansas City, Chicago, and Jacksonville all have quarterbacks that have, either through inexperience or perhaps a lack of talent, not yet gained traction and found a way to put up a lot of points. St. Louis and Indianapolis are potent offenses that have had one or two less successful outings. In Indy's case, their one weaker offensive game against Tennessee was coupled by a defense that has significantly exceeded expectations. Along with the Colts, Jacksonville, Minnesota, Chicago, New England, Tennessee and San Francisco all have defenses in the top half of the league and have kept their opponents off the board. For a team to consistently go under it needs a combination of two or more factors - an incompetent offense, a powerful defense, or unrealistic expectations. These 10 teams all check out.

It will all come back to the middle: Though the under has been coming up more often so far this season (56.7 percent of games played), recent history has shown us that this will come back closer to the middle, and sadly to a point where neither side is profitable by itself. Last season, 51 percent of games that weren't a push ended up over. They year before it was 52 percent of the games that went under, while 50.4 percent were under in 2004. The number this year could fall within a final range of 48 to 52 percent, and there is no way to know for sure how it will turn out, but it is highly unlikely that it will fall out of that range. That means that we will be stuck having to actually handicap the games again this year instead of just blindly betting them for profit.