Sunday Drive

You inherit a fourth quarter between two real NFL teams: you are trailing by one to eight points, there is between a minute and a half and three and a half minutes left, and the ball is somewhere between your own 15 and your own 40. You have one to three timeouts.

Everyone playing the Daily gets exactly the same situation, and the same choices resolve the same way for everyone. A new one appears every day at midnight Eastern.

You have six families to choose from, not a forty-play playbook: run inside, run outside, quick pass, intermediate pass, a deep shot, and a screen. Each trades yardage against risk differently — a shot gains the most when it connects and is intercepted far more often than anything else.

A field goal appears when you are close enough for one to be a real option. A spike appears when the clock is running and you have a down to spend. A timeout stops the clock without costing a down.

An incompletion stops the clock. A run that stays in bounds does not. Getting out of bounds stops it, which is why an outside run and a pass toward the sideline behave differently from an inside run.

The clock costs in this game are measured from real late-game football: a trailing offence inside five minutes uses about 23 seconds per play with the clock running and about 8 when it is stopped.

This is the part people learn the hard way. If you take the lead with meaningful time left, the other team gets the ball back and can answer. Scoring with forty seconds on the clock and no timeouts for them is a very different outcome from scoring with a minute and a half.

The drive ends in a real result — WIN 27–24, or a loss because they answered, or a turnover on downs. There is no grade out of a hundred.

Your first Daily drive is the one that counts, and it is final — there is no replaying it for a better score. Unlimited generates a fresh situation whenever you want one, held to the same quality bar, and does not affect your Daily record.

Team and player ratings are derived from publicly published NFL play-by-play data, then normalized within the season and shrunk toward the league average in proportion to how little data supports them. The outcome of each play is drawn from measured league frequencies for that kind of play in that situation, adjusted for the matchup.

These are Golfle’s own game-model inputs. They are not scouting grades, not official statistics, and not predictions about real games. The full source and licence list is on the data sources page.