How effective are college football offenses that spread out opposing defenses?

More and more teams are either integrating spread offense principles or formations (of any generation) or are adopting a version of the spread wholesale -- but that certainly doesn't mean that every spread offense has been effective. Simply sticking your best athlete at quarterback, moving him to the shotgun, and telling him to read a defensive end after the snap won't get your very far.

However, thanks to Bill's data on opposing solo tackles, I wanted to see if there was any relationship between effectively spreading a defense out and the quality of the offense itself. I graphed out Solo Tackle Percentage with Offensive S&P+ for all teams last season:

First thoughts:

• All four College Football Playoff teams are clustered near the center in terms of % solo tackles, but are on the upper end of Offensive S&P+. Florida State's offense was great last season -- 1.2 standard deviations above the mean for Offensive S&P+, but it was certainly the worst statistically of the four Playoff teams. It is also the one of the four that generated the most solo tackles.
• Even though all four Playoff teams are relatively clustered together near the middle, two of the four are clearly spread-to-run teams even though they didn't generate a significant number of solo tackles last season. Oregon and Ohio State are second- and third-generation spread offenses, but interestingly not all spread offenses necessarily spread defenses out in a way that leads to solo tackles. Baylor is a good counterpoint -- an efficient spread offense that did generate a large number of solo tackles.
• The association between spread offense (or really, generating solo tackles) and offensive efficiency is fairly low, as indicated by the mostly flat slope of the line of best fit. There are excellent offenses that produce a very high number of gang tackles like Arkansas and LSU, and that lead to a high number of solo tackles, like Baylor and UCLA.
• Obviously "spread offense" means a lot of things to a lot of people, so it's important to note that what we're measuring here -- the relationship between one measure of offensive success (S&P+) and opposing solo tackles -- is a very specific test, so it's not necessarily appropriate to generalize that spread/pro-style offenses are any more or less efficient as a scheme than the other. That being said, there was a statistically significant (at a 95% confidence level) relationship between solo tackles and Offensive S&P+ -- even if the r-squared value was extremely low. There's just a lot of variation in quality between spread offenses and between what we might call pro-style offenses.

(Note from Bill: the lack of correlation between the solo tackle figures and S&P+ is actually helpful in a way -- one of the concerns about it was that teams that simply don't have good tackle-breaking ball carriers are going to end up looking like spread offenses because it doesn't take multiple tacklers to bring anybody down. That there is a lack of correlation between these two variables certainly alleviates that concern.)