forums

CFB Analytics Thread 2026-27

hurricaneman said:
A couple things:

Wake is 15th in yards per play against FBS opponents. They've also played the 3rd (Louisville) and 5th (Miami) teams in yards per play on offense. Wake is a really good defense.

I would argue Clemson is a top 50 defense and will likely finish the season there. The LSU game really messes with their stats. Their last 4 games would have them ranked around 37th, including playing us.
Click to expand...
His dataset is kinda weird.

Deleted my previous posts because I found out what the problem was.

He made the mistake of showing the chart that does not filter for FBS games only, but the claims he made based on the dataset are from a filtered FBS games only subset. If the chart matched the latter subset, it’d look like this:

IMG_1413.jpeg


This is why i kinda hate the FBS vs FBS only systems. Alabama has played zero FCS teams. Without filtering for FBS games only, they’ve played 5 top 50 defenses. By filtering it, they’ve only played 2.

Likewise, Miami has played two top 50 defenses without filtering for FBS games only, but Clemson and WF both slide out of the top 50 in this filter. The strange part of this is, is that neither team has played an FCS opponent yet! So why in the world is a FBS-only filter making them look so much worse?

My primary preferential filter, if any, will always be P4 vs P4 stats; but removing FCS teams from the input-equation distorts everything so much.
 
Last edited:
At the same LSU, who has played an FCS opponent, goes from the #3 defense to the #1 defense when filtering for FBS only games. How in the world does that make any sense?

Diving into the inputs, FCS opponent Mcneese state scored 14 points on 47 plays whereas FBS opponents Louisiana Tech scored 14 points on 55 plays, Clemson scored 10 points on 62 plays, Ole Miss scored 32 points on 73 plays, Kentucky scored 35 points on 68 plays, and Texas A&M scored 6 points on 60 plays.

So in the oddest way, removing an FCS opponent actually helps LSU who gave up a relatively high points/play average against McNeese State compared to the others:

14/47=0.298

FBS ONLY:

14/55=0.255
10/62=0.161
32/73=0.438
35/68=0.515
6/60=0.1

97/318=0.305

Slightly higher. Hmmm. The key is; Louisiana Tech scored on a pick 6 which isn’t the defense’s fault,

90/318=0.283

Boom. That’s why LSU’s defensive rating goes up when FCS opponents are left out. Since a lot of backups play against FCS teams, I’d imagine that’s how 25 teams are able to jump Clemson + WF (who haven’t played any FCS opponents) in defensive rank. It’s counterintuitive.
 
Al Davis----"Just win baby".

I mean the stats are interesting, but who did you play. Were you running up the score or playing young guys. How was the weather. was the other team healthy ( we weren't vs WF and). Does give us something to look at and talk about between games.
 
Back
Top