Also measured by W's & L's, although some people in here don't seem to think so.
Wow. Let's try to break it down for you.
W-L reflects team strength, including especially offense and defense but also special teams. If you use W-L as a measure of just the offense or just the defense, then you are betraying an ignorance of the game. I'm not sure why you failed to understand that when you were corrected last time.
W-L is, again, a matter of scoring more points than your opponents not yards. So, one of the simplest ways of translating W-L in to a measure of offensive and defensive efficiency is to simply measure points for versus points against. This stat is typically listed as Scoring Offense (points for/game) and Scoring Defense (points against/game).
However, this stat suffers two important limitations: (1) it includes scores by special teams and defenses as if they were scored by the offense, and (2) if used for comparing teams, it assumes that all games reflect basically the same number of possessions or drives/game. Like all raw stats, it also doesn't account for differences in strength of schedule.
That's why I use the points/drive versus pwr5. My points/drive stat counts only offensive scores and FG's, and by requiring more than 2 pwr5 opponents the SOS is accounted for better than not at all.
Nevertheless, these stats are still raw. They measure how many points does an offense score, or a defense allow, per drive. By subtracting this defensive ppd from this offensive ppd for each team, I have a pretty good indicator of team strength. It has had a fairly good correlation with polls based on W-L, validating it as a measure of ranking the offensive and defensive units independently.
In other words, if you think that W-L is a good indicator of team strength, then you should value my ppd vs pwr5 stat as a good indicator of offense and defense independently.
I don't cherry pick stats.
I prefer overall (total O or total D)
You like to refer to "cherry picking," but, again, I'm not sure you understand what that means. It means just finding and using the stats that fit your conclusion rather than using appropriate stats for the question at hand.
For example, if you use W-L to talk about how well our offense has performed, that's cherry-picking because W-L reflects more than just offense. Similarly, if you use yards/game or yards/play stat to address a question about scoring, that's cherry-picking. If you go back and look at the post that prompted your response, you'll see that I wrote (including your emphasis): "we were practically scoring at will against a
top 25 defense from last year." I would like you to see that I made a claim about our
scoring. So, the appropriate (not cherry-picked) stat will be scoring defense, and ppd vs pwr 5 is even better because not all games have the same number of drives.
