OK, I know you're joking, but here's something that's been bothering me since last night:
All of the analytics models argued pretty strongly in favor of going for it, and yet everyone (me included) think it was dumb. So what do we think the analytical models are missing here?
It's not enough to say, "Well, the analytics can't account for every situation", because I'm guessing most of us would have trouble finding
any situations where we would recommend going for it on 4th and 1 from our own 18.
You also can't just say, "MOMENTUM!" and then drop the mic. First of all,
momentum is a myth (in terms of having any predictive capability), but also, the way the game actually played out would seem to rule out momentum being a significant factor. The Chargers D responded to the crushing disappointment of the failed conversion by ... forcing a three-and-out on the subsequent series. Meanwhile the offense staged a 15-point 4Q comeback. Even if momentum exists, it certainly wasn't what cost the Chargers the game.
To me, the most obvious critique of Staley's decision can be seen in what did happen: the defense did its job, held the Raiders to a quick three-and-out where they only gained four yards, and LV was still able to score three points on a chippie FG. In other words, the cost of not converting was an almost guaranteed 3 points for the opponent. So why wouldn't the models account for that?
With Belichick's infamous call vs. the Colts, the benefits to the Pats were far more apparent. They were leading late in the game, so if they converted they had a chance to run out the clock and keep the ball out of Peyton's hands. In last night's game, if the Chargers had converted, they would have had 1st and 10 from the 20. What are your expected points from a drive in that scenario, and how does that compare to the 3 points they were handing the Raiders?