Campbell's numbers dont lie

[GTripp0012]

Originally Posted by CRedskinsRule

GTripp,

I looked at those, and it seems less clear. If you say a TD pass is +10 yds, your basically saying that a TD is equal to one Firstdown. That's not a fair value. I think the -45 for an int is probably a good number, to say an INT costs 4 first downs, or half a possession on a sustained drive. But a TD really has to be considered as valuable as a full possession, because the other team would now be one possession behind your team.

Not sure how all the numbers mish mash, but in this case, I think they are doing it more for historical ranks than actual game time value. I might agree in that regards

I think the implied is this:

You can already judge a quarterback by his YPA figure. If Brett Favre has a career YPA of 7.0 (which he does), and 9565 attempts (which he does), then you can already predict his TD total using a simple regression with that information.

Now, it turns out that Favre has a career TD rate of 5.0 which is better than roughly 64% of quarterbacks who played over the exact same timeframe as Brett Favre. So, unsurprisingly, Favre is a more valuable player than a hypothetical player who played over the exact same timeframe and threw the same # of passes for the same number of yards. We can all agree that Vinny Testaverde is not the same as Brett Favre, even though he played over the same timeframe, and had similar attempt and YPA numbers up through 2002.

The 10 marginal yards comes from the assumption that you can already account for passing yards by looking at passing yards, but you're trying to determine how valuable that one yard between the one yard line and the goal line is relative to the rest of the field. I think PFR's analysis falls apart a bit because a TD pass off of play action from the one might actually only be worth ten yards of field position (of course, the same pass on third down might be worth three times as much as a one-yard TD pass on first down), while a TD strike from the 25 yard line might be a passing play that would have gone for 40 or 45 yards if the endzone hadn't truncated the play.

I happen to think the 20 yard estimate is a lot closer to the true value of the average TD. But, admittedly, that's based on "feel" and not a lot of evidence. If we establish the value of the average INT = 45 yards, it's tough to say that the last yard between the one yard line and the end zone should be considered equivelent. We know that it's greater than one yard...it's the toughest yard to get. The major key is to seperate "passing touchdown" from "offensive touchdown".

I'm going to add on to this post later, but right now, I'm changing locations.