One of the (many) reasons I favor going deep on 2nd and 2 is that a pick 40 yards downfield isn't really any worse than a punt. It's far from optimal; if you have a drive going and you give up possession that's obviously bad, worth about 4 potential points by itself, but it's only as bad as giving up a TD if it happens at the line of scrimmage.

Heh, well, I usually take the view that my mind is a far more entertaining toy than my body, but it seems Norway has more stills per capita than any place but AL. I did more drinking in two weeks of Christmas holidays than I normally do in a year, but, strangely in a country where 70% of people told one pollster they don't believe in ANY personal god, the place pretty much shuts down over Christmas (as well as Sundays, and most of Holy Week).

It's an interesting proposal, but seems too much like bringing the bounce pass to the NFL. Why should a one hopper picked up and run for a TD be an incomplete if the receiver's 5 yards past the line but a game winner if he's 5 yards behind it? Don't forget that, while it may help his PR, as far as the official NFL stats go, a TD pass thrown anywhere is credited to the receiver, not the QB; I think that's enough to make up for giving them 15 PR points when they throw a pass for -5 yards.

But, really, the nice thing about this and so many other things in THGoF is that they demystify the system, and once we understand it we can tweak it in ways everyone can understand and fine tune a better system more people can accept, or at the very least articulate their objections to it and the alternatives. I don't mind crediting QBs with what amount to short running plays if they initiate them, provided they get the blame for bad plays as well as the credit for good ones. Probably the biggest thing that needs to be done is cut the value of completions in half, but without getting into how we'd like to modify the existing PR bonuses, I think it's past time we factored in things the League ignores. With that in mind I've concocted what I call a "QB Rating System" because it evaluates overall performance as a quarterback rather than purely as a passer. Obviously I don't have a convenient metric for playcalling; the League doesn't track audibles and there's no way to know which plays come from the sideline and which from the QB, but ball handling and rushing have been too long neglected by the PRS, and here's how I propose that be remedied:

The simplest thing is to also count:

rushing yards,
rushing TDs like passing TDs
lost fumbles like interceptions

Each of those actions counts as an attempt (fumbles recovered by the QBs team count as attempts for the yards lost), but only positive yards count as completions. Be sure not to count TD yardage/attempts twice (it would be a lot better if we could find interception return yardage, but team data is the best that's likely there).

You can subtract 2/7 of a TD if the QB is sacked for a safety and/or add 2/7 if he runs or passes in a 2 point conversion, but both are rare enough to have little impact. You can also double the fumble penalty (or half the interception penalty) since interceptions are usually down field but QB fumbles near the line. Everyone's seen blindsided QBs lose the ball; the PRS penalizes 100 points per pick, even 40 yards away with no return (i.e. a punt), but since LoS fumbles cost about three times as much in terms of potential points doubling the penalty is, if anything, generous.

This does NOT automatically favor runners, who tend to be sacked more. It mainly penalizes otherwise good passers who hold the ball too long in a rush, giving up sacks or, sometimes, fumbles, almost invariably lowering ratings. However, really good runners (like Vick and McNabb) can make up ground, because their scores fall LESS.

As the authors note when The Hidden Game of Football dissects the system, any RADICAL leader change would make a new system uselessly suspicious. Under this one, QBs good and bad remain just that, but there's definite shuffling. In fact, it helps (though by no means eliminates) one problem plaguing the PRS for years: Comparing modern passers to '40s and '50s era T formation QBs is hard; comparing them to older “passers” who might throw a dozen (and run forty) times in a game is impossible. For a smaller comparison, Steve Young’s still second overall for best season rating:

Young 324/461 35 TDs 10 INTs Sack: 31/163, Rush 58/293 7 TDs 3/4 TOs

Old: 112.79 New: 105.87

Manning 336/497 49 TDs 10 INTs Sack:13/101, Rush 25/38 0 TDs 2/5 TOs

Old: 121.10 New: 113.46

This is typical; both ratings drop about 7 points (a particularly good/bad runner can be ±5) but Youngs famous ability to run for TDs and yards is good for an extra half a point, not nearly enough to challenge what remains about an 8 point lead. Young would make up more ground, but suffers from nearly three times as many sacks for 50% more yardage (in my system sacks are essentially incompletes that costs yards) as well as losing nearly twice as many of his fumbles.

Incidentally, another benefit to this approach is that it factors in every single offensive action that can occur. That means you can rate not only QBs, but ANY ball carrier (receivers can have a completion percentage, too, and it's a far more meaningful statistic for them) or even whole offenses. It's not nearly as good for defenses; so many scalars are necessary to produce stats for comparison that almost any other approach would be preferable.

Really, the only beef I have with the West Coast offense in itself is the risk/reward ratio. If teams want to dink and dunk their way to the end zone that's fine by me, but how many times have you seen a DB jump one of those quick outs and have nothing between him and the end zone but empty air? If you want to spend a dozen plays getting to the end zone, more power to your ball control clock consuming offense, but the smart way to do it is on the ground. Given the far greater likelihood of a passing rather than rushing turnover I want the vast majority of my teams passes to be for either first downs or touchdowns. I don't mind vastly increasing the risk coefficient, but I insist on proportionately inreasing the reward coefficient. You need enough short passing to keep the defense honest, but shouldn't make it your focus.