Win% vs. yards/game in 2018 college football:
View attachment 5497
The # of yards per game your offense produces explains about 30% of the variability in your win%.
Win% vs. Pts/game:
View attachment 5498
Pts/game explains about 50% of your win% (duh, how much you want to bet that pts allowed/game explains about 50% as well?).
So clearly scoring points means more than just getting yards.
To no one's surprise, average Margin of Victory is a great predictor of Win%:
View attachment 5499
MOV can explain 86% of the win% variation. But it is not perfect. NW has a +1 MOV and 64% win% (88% in B1G). ND had a +13.2 MOV and 92% win%. Iowa had a better MOV, 13.4, but only a 69% Win%.
It is just common sense that points matter more than yards. But, of course yards tend to predict points.
View attachment 5500
Yards/game explains 80% of the variance in pts/game. You can see Iowa is above the curve, scoring as many points as a team that averages about 40-50 more yards per game.
So to the point at hand, OF COURSE points matter more, so we shouldn't dismiss Iowa scoring effectively just because their yards/game is less than stellar. But if they gain more yards, they would be expected to score more points.
What we should really be doing is judging how many points they score per drive compared to how many they would be expected to score based upon starting field position and strength of defense. Luckily,
football outsiders does that for us. Based upon this metric, Iowa is 48th in the country. In the B1G, they are behind (in order): OSU, Mich, Purdue, Indiana, Nebraska, PSU, and Wisconsin. So Iowa is 8th. Lots of room for improvement.
The good news is that in terms of
defensive efficiency, they were 7th nationally, and 2nd in the B1G (behind MSU).
And if the offense gets more yards, they will tend to put the defense in better field position, yielding even fewer points (of course, punting is huge here as well).