Lastly, if you look at how the Cardinals won, it was by intercepting the QB five times. But they only had 13 in the entire regular season. They average less than one per game. So perhaps they defense is playing really well. But its more likely that Carolina's offense was just really bad.
The toughest things to figure out are absolute measures of a team. I think I would have to have a great understanding of a team and watch the game three times at a minimum to figure out how good a team is. If the Cardinals defense is say, a 60 on a 100 point scale, but Carolina and Delhomme play at a 20 level, the Cards will really do well . But if Carolina plays at a 60 but the Cards at a 40, it will look worse for the Cards.
So the two concepts are the *normal skill* or average, median, mean, typical, or baseline, performance level of an a team unit. Then there are certain factors applied to it, for motivation, correct game planning, fatigue, familiarity between teams and players, etc. These factors can boost or reduce a teams number.
As an analyst one has to figure out the baseline level and then the true level. And these concepts can be difficult to discern because there is no easy way to measure them. A defense can only be graded relative to the offense it is playing against, which also has a myriad of its own factors. No home-boosted offense plays a home-boosted defense.
These are the problems I have to solve. The first is just watching a game several times. I've never watched a game more th an once. What if I watched it five times? What if I watched every game that a football team played all season five times?
Then I would have a tremendous understanding of that one team.