One of the distinctive elements to the NFL is that the season is so short -- sixteen games does not provide a lot of history to go by when evaluating a team. At the halfway point of the regular season you have only eight games played to assess, which means that a rash of good or bad breaks along the way can quickly lead to false impressions of a team’s true ability.

As a point of contrast, after sixteen games had been played in this past year’s Baseball season, of the six division leaders at that stage (Boston, Cleveland, Seattle, NY Mets, Pittsburgh, San Francisco), only one actually made the playoffs. The eventual champion Anaheim Angels would have been considered a poor team by NFL standards with a 6-10 record!

The brevity of the NFL schedule is in part perhaps why underdogs have historically been a smarter play, since getting an accurate read on a pro football squad can be hard. Clearly it’s not realistic to expect the players to participate in more games, they already suffer enough physical damage as it is! As handicappers though we need to be aware that the quick look at a matchup using winning percentage on the year can often be misleading.

A reasonable starting theory then for finding some good wagering spots would be to consider how we might locate teams with less than stellar records that are actually decent teams with some amount of hidden ability - the Anaheim Angels of football.

Could it be that there are signs in a team’s stats that could guide us to under-rated clubs ready to break out with some strong efforts?

The following table lists several factors deemed important to a team’s overall ranking from the public and linemakers, and the spread performance as an underdog playing qualifying teams with below brilliant records (from week 5 on):

First off, all the cutoffs above are arbitrary and were decided in advance of running the actual data tests. We could have used +3 net points instead and so on, but rather than try and backfit the data to optimal ranges, we began with some assumptions/opinions as to what would represent a positive statistic in a certain area and went with that. The <55% breakout was designed to capture the exactly .500 teams without including too many above the line, while the <45% is designed to only include teams with a (significantly) losing record.

Of course at first glance the above numbers are impressive - to hit 63% against the NFL point spread is good going in anyone’s books, and most of the <55% categories above produce anywhere from 10 to 15 plays per year on average, and 5 to 10 plays per year in the <45% breakouts.

We didn’t include the 2002 results so far in our dataset, but chances are very good that the numbers above would be better than the ten-year average shown given the dominance of underdogs this season.

There is a lot of logic behind the results, since we know -

• underdogs have historically been much stronger plays than favorites in the NFL
• with so few games on the schedule, teams can quickly become over or under rated on the basis of their final scores and won-lost record alone
• a team with a positive statistic from the list above could be presumed to have some level of talent and capability to win future football games

Still, it’s hard to believe that the basic premise is anything but sound: look for teams with good statistics in key areas that for whatever reasons have not translated into a successful won-lost record. Then back these teams when they are underdogs!

At the time of this writing (games completed through week ten) the following “good stat, bad record” teams exist:

Net Points - Kansas City, Jacksonville
First Downs - St. Louis
3rd Down Conversion % - Arizona, Minnesota, St. Louis
Net Rushing Yards - Kansas City, Minnesota

The astute reader may after looking at the above come up with a further line for research on this subject - how have teams with “good stats” in more than one of the key areas fared? The answer unfortunately is that the results are inconclusive - the sample sizes were small for our first pass data-crunching, going to teams with two strong areas reduces our basis for analysis even further. At a glance the spread performance as an underdog with multiple positive factors appears to hold up fine for teams with 45% to 54% winning records, but slips for teams with <45% winning records.

In conclusion, keep an eye out for teams that may not have shown results in the win column, but have put up some numbers elsewhere.