Each year around this time, I try to figure out what's going to happen in the upcoming NFL season. Doing that in August isn't exactly an exclusive hobby of mine, but I try to focus on using metrics and measures that have historically been good predictors of future team performance. What happened last year matters, but not always in the way you might think.
In recent years, my predictions have been pretty accurate. I've split this column into winners and losers over the past three years, and over that time frame, I've identified 16 teams that were likely to improve the following year.