Quote:
Originally Posted by DJ's left nut
Short answer?
Yards surrendered has more predictive value.
Points scored can be dependent on things like a turnover short-circuiting a drive or an ill-timed penalty from the opposition. Granted, some teams build their defenses around exactly that, but it's not the norm.
So when you're looking for a rearview statement of how a team's defense has done, points scored is obviously the best metric. However, when you're trying to analyze how a team may do going forward, yards surrendered is a more constant statistic with more predictive weight.
If the Tigers have routinely given up a lot of yards this season, they're probably going to give up a lot of yards in most every game they play. However, they won't necessarily get a lot of the turnovers, sacks, etc... that have stalled drives for them in every game this year. So even though they've not been a bad scoring defense, there's very little predictive weight to that.
The best example I can give you is the 2003 Chiefs. The Chiefs were actually an average defense on the year when discussing points allowed. The NFL average was 20.83 PPG and the Chiefs gave up 20.75 PPG - a tick below the NFL average. However, in total yards, the Chiefs were the 4th worst in the NFL that year, giving up more than 10% more YPG than league average (356/gm vs. 318).
And as we all remember, the Chiefs got worse and worse as those breaks stopped coming. Teams learned to avoid turnovers and just plug away at the defense. Suddenly the Chiefs PPG ended up right in line with their YPG and the Chiefs gave up 27.5 PPG over their last 8 games (including the Colts game).
If you're looking at numbers and trying to predict what a teams going to do going forward, yards surrendered by a defense is actually probably the best stat to look at. It's counter-intuitive and it took me years to get there, but it really is the best predictor of future results.
The wild card here isn't necessarily the number itself, but the fact that the Tigers may have been sandbagging against lesser opponents and as a consequence have an unduly inflated yards allowed figure.
|
You make good points about the predictive quality of yards surrendered, but I'd still argue that yards per play, yards per attempt and yards per rush are better indicators than just total yards. Those things will hold up regardless of number of possessions and total plays.
For example, what's better:
A defense that faces 85 offensive plays per game (40 pass, 45 run) and surrenders 450 total yards (5.3 YPP) or a defense that faces 70 total plays and surrenders 400 total yards (5.7 YPP)?