Stats Too Early Top 25 (week 6, after week 5)

AE 87

AE 87

Helluva Engineer
Messages
13,026
Over the off-season, I convinced myself that differential points/drive (a team's offensive pts/drive vs pwr 5 minus their defense's pts/drive vs pwr 5 allowed) was a pretty good raw estimate of team strength. Obviously, it becomes more accurate after more games are played and the quality of competition becomes more representative.

For a season, I restrict the ranking to teams that have played more than 2 (3 or more) pwr 5 opponents. For this too early estimate, I restrict it to teams who have played more than 1 (2 or more). Since ND is not actually a member of a pwr 5 conference, games against them aren't counted. So, CU is not included because they've only played ND Lulzville and 2 cupcakes.

So, here's what we have:
upload_2015-10-6_17-54-8.png


I think it's a pretty interesting list. We've played #7, #8, and Duke who played #11 pretty close. Duke is ranked #30, and we're ranked #48 (of 90 teams who have played more than 1 pwr 5 opponent).

Again, it's too early, but Utah is currently ranked #5, with 7 votes for #1. All 25 are in the AP 25 except UNC, WV, Iowa St and Wash St.

The AP includes #6 CU (not enough pwr 5 games), #14 Ole Miss (#40 DiffPPD), #23 Cal (#27 DiffPPD), and #3 Baylor (not enough pwr 5 games).

So, it will be interesting (to me) moving forward whether this ranking becomes more predictive/accurate.
 
AE 87

AE 87

Helluva Engineer
Messages
13,026
Boomergump said:
Oh Snap! Where is Clemson in the differential PPD metric? Doooooooh!
Click to expand...

I explained that in my op. They've only played 1 pwr 5 team (lulzville), so they didn't make the cut.

However, if I count ND as a pwr 5 team and include ND and Lulzville together, then CU's diff PPD is .89 which would've been enough for #13.
 
Top