Ok, for those who don't know, I calculate a Pts/Drive stat for offense and defense. It's based on 7pts for every scrimmage (rushing or passing) TD scored for offense or allowed for defense. So, special-teams and turnover scores as well as safeties are not counted. Obviously, made FG's count 3. Drives are every possession which ends with a TD, FG attempt, punt, failed fourth-down, or lost turnover. So, safeties and end-of-half possessions don't count as drives (unless there's a score as time expires).
Here are the game by game pts/drive for Clemson in 2018:
View attachment 6493
While I didn't calculate end of year PPDvPwr5 for all of Div 1 last year, Clemson's PPDvPwr5 averages are obviously incredible. Those 3.25 offense and 1.06 defense ppd stats against pwr5 for the season are each typically top 5 averages. After NC State, these averages go to 3.44 and 0.96, resulting in an average differential of 2.48. An average differential of 1.00 is typically top 10, IIRC.
For those who act as if Duke is still Duke, you might want to see how Clemson's O performed against them compared to their average against Pwr5 opponents last year.
If you look at our performance against Clemson, then, it appears that they scored 18% better against our defense than their average, and we scored 65% better against them than they allowed on average. Obviously, that didn't mean much when they were so good.
I suspect that the general wisdom which ranks us so low this year is that our D was limited by talent and our O was benefited by scheme. With that assumption, we are not projected to be better on either side of the ball.
However, I think that we've heard a lot to suggest that our D was limited by coaching (let's just leave it at that), and that Collins/Thatcher have had a lot of success producing quality D's. So, I'm looking for decent improvement on D, even in our first game. I do think that our our O was benefited by scheme, but I don't think our talent is that bad. I'm hoping our offensive efficiency only drops off a bit.
Clemson's PPD averages are just insane. I think it will again take them some time to get rolling on both offense (and I know they started last season with 2QBs) and defense (losing starters in front 7, but they played a lot of guys). So, I'm going to guestimate that CU's baseline O is 2.90 PPDvPwr5 and D is 1.25 PPDvPwr5.
So, if our D is 20% better than average rather than 18% worse, then we'll hold them to ~2.4 ppd. If our O is only 20% better than average rather than 65%, then we'll score 1.5 ppd.
As a result, if we each have 14 drives, then the final score would be CU 34 - GT 21. However, we also get a Juanyeh Thomas KR score and also somehow limit them to only 13 drives because of the end of half end of game, so that makes it 31-28. On top of that, we go for 2 after all four TD's and win 32-31.