About this website
Every week of the American football season, one game is more interesting than the schedule makes it look. This site picks that game.
The underdog
Before each game, bookmakers publish a number: how many points they expect the weaker team to lose by. That team is the underdog. A three-point underdog is barely behind. A seventeen-point underdog is not supposed to be competitive.
The number appears once on each page, as a plain sentence. It plays no part in choosing the game.
How the game gets chosen
Every game has an underdog. Sixteen games in a week means sixteen of them, so the question is never who the underdog is. It is which one is worth your afternoon.
It is not whoever is most likely to win. That picks the closest game on the board every week, and two evenly matched teams make a fine game but rarely a story. Teams expected to lose by three or fewer win about 45 times in 100, which is not an underdog in any sense worth the word.
So the site reads the record book instead, and looks for these:
- A drought against this opponent. How the two have gone lately, how long since the underdog last won, and the longest run of defeats inside that.
- A drought at this stadium. How many years since the underdog last won in that building.
- The last time they met. Who won, by what, where, and how recently.
- A coach against his old club. Whoever runs the underdog having run the favorite before.
- A run to end, or one nobody noticed. Either side arriving on a winning streak.
- A record that disagrees with the scoring. The better record belonging to the team with the worse points difference.
- Wins by a whisker. A record built mostly on one-score games.
- How old the rivalry is, and what the result decides.
Each one found is worth something, and the highest total takes the week. A game with nothing worth a headline is not picked at all, however close it might be.
One rule about hopeless games
Games at seventeen points or more end as blowouts about 45 times in 100. A game that big only leads when no other game is available, and when that happens the page says so.
Where the facts come from
Every result, venue, coach and score since 1999, and the current season as it goes. Nothing is claimed from before 1999, which is why a rivalry is described rather than counted: two of these teams have met more than two hundred times and the records here see only the last twenty-odd years.
Who writes it
Nobody. Each piece is assembled from the facts found above, and it may only say what those facts say. It can connect them and phrase them. It cannot tell you why something happens, what a team needs to do, or what a trend means. If a sentence cannot be traced back to a fact in the list at the bottom of the page, it does not get written.
That list is the receipts. Every line in it is checkable, and none of it is a forecast.
Written before, never rewritten
The pick is made Tuesday morning and does not change, even if the news does. Everything on the page was true that Tuesday: records are records going in, streaks are current as of then, and nothing refers to how the game turned out.
When the game finishes, two things are added and nothing else changes: the score on the card, and the result at the bottom.
What the result means
The underdog loses most weeks, so most of these report a loss. They are graded on the watch rather than on the upset. Losing by three after leading into the fourth was a good afternoon and reads like one. Losing by twenty-eight was not, and reads like that.
The weeks
Weeks run straight through. The regular season is 1 to 18 and the playoffs carry on from there, so the Super Bowl is Week 22. Each week stays put once posted, so the tabs fill up as the season runs. All times are Pacific.
There is a pick every week, the Super Bowl included. Even in that one game somebody is favored and somebody is not, which is all this site needs.
