Perhaps the BEST B1G Forum anywhere, here at College Football Fan Site, CFB51!!!

I like this discussion, and in some spots, I disagree with OAM, others with bwarbiany.

The riskier offense thing doesn't make all that much sense to me. The option teams turnover rates fluctuate all over. Shoot, even Wisconsin's does, and that's despite relying heavily on a run game that involves no pitches. 

The question as to why would the option be less effective at Bama is because Bama has access to higher end QBs and WRs. The best offense is throwing more, but having a level of physicality in the back pocket Something like Navy works because it's weird and because it's just really well run. If you can run a spread that leans pass as cleanly as Navy's offense, you can have the most efficent offense in the country. Bama running the option would probably have more of a ceiling with where passing is (the option also relies on a lot of reading big linemen, which solves a problem Bama doesn't have, since it can just block them).

But if you can't get good to goodish QBs, you see the ceiling of passing offenses fall. Sometimes you can get around it with parred back playbooks and excellent execution (Leach). 

The tempo part is more interesting and where bwarbiany isn't super correct. Lower tempo is good for underdogs. It's good because if three big weird things happen, that can swing an eight possession game more easily than a 15 possession game. This actually fed into some of the more historic OSU upsets because they always played slow, so a few weird things could befall them (it also meant they needed a slew of weird things to help them, which often happend with Tress). The analog is this. If you pay 1-on-1 against a better player, you'd prefer to play to 3 rather than 15. You might be able to luck into three points first, but the longer it goes, the more things slide toward the average outcome of you bleeding points.

But on top of that, there's another factor to tempo: extremity. If you are comfortable at extremes and can pull opponents into that, you create conditions where you feel better with the flow of the game. If Army can force Michigan to have the ball 6 times, Michigan isn't super happy about it. Likewise, if you are Kliff Texas Tech and can get 18 possessions in a game, you can in theory make someone else uncomfortable. The issue is that if a normal team can push you off your slow and fast approach, you're out of your element AND at a talent disadvantage, which means you lose like you were supposed to. 

Anywho: OSU lost to Purdue in 2017 because its good offense played bad, it's bad defense played bad and Purdue's OK offense played out of its mind. 

(It's interesting to look at this on the HS level. Talent is usually king, but a QB is this interesting factor. If you have a kid who can throw and receivers that can catch, your ceiling is higher. But lots of schools don't and try anyway. Others will go option or single wing because they know they can get the 1-3 athletic kids the ball and teach the other 8 to block much more often than they can get kids good at throwing and catching)

Yeah, "just because I think so" isn't persuasive. 

I already clarified that I don't have empirical evidence that pass-heavy spread offenses are inherently higher variance / more risky than traditional ball control offenses... But it makes logical sense. If I'm wrong, I'm wrong. 

so, i wanted to "test" the theory of "high risk" offenses.

so i took last 5 years average turnovers. here are the numbers (note, a few schools, like idaho, uab, etc, are NOT included due to not competing in div1 for entire 5 year period).

rank - school - 2019 - 2018 - 2017 - 2016 - 2015 - - - total average



looking at those, it does seem to have a correlation between conservative offenses being lower and "higher risk" offenses being higher. and there are outliers as well.

also consider, this needs to be taken with more than a little grain of salt. it's missing quite a few data points that should be considered. like if the risk of turnovers is worth more points. talent level. total plays ran. effect on defense and total plays faced against. etc. it'd be an interesting study, but i don't have time to do it right now.
but it'd be interest to see if we can come up with a decent data set that we think might give us a logical result. any suggestions?

Raw turnovers can be tricky because it's also a little tempo-based and games based. It's kinda a clunky number. 

That said, it's remarkable how low LSU was considering that spans 3.5 OCs with four or so changes, including one mid-season.

very true on turnovers. spreading it over several seasons should normalize it some, though.

as for lsu, 3-4 new oc's or not, they ran the same offense for basically the last decade... until 2019.

No, no, no. 

They were in the Cam Cameron pro-style till they fired him. Then tinkered with what they could for the last 8 games that year, then Matt Canada, then back to Ensminger. Then Ensminger/Brady with the 2009 stuff. 

Some different stuff in there. 

Risk is more than turn over potential obviously.

Yeah, I wonder if you can ferret out average plays per drive from this the available data.

Or ratio of 3 and out drives to drives which get at least one 1st down.

PSU?


Do you mean Purdue?

Yep. That’s the one.