Projected 2018 S&P+ rankings

Georgia is up to sixth in Bill Connelly’s preliminary rankings for the coming season.  The analysis is what you’d expect:  stellar showings in recruiting (2nd) and returning production (5th), dragged down by the weighted five-year average (28th).

Going forward, this should be about as low as it gets, then.

7 Comments

Filed under Georgia Football, Stats Geek!

7 responses to “Projected 2018 S&P+ rankings

  1. UGA'97

    Connelly knows his stuff, but not sure how much of each teams full schedule is taken into consideration here. According to this S&P+ Missouri is the 2nd place contender this fall ranked at #30? If that’s the case, man, doesn’t the SEC East looks like a dump? Florida came in ranked 5th in the state behind even Florida A&M and the Dogs out of conference schedule is abysmal and Tech still has the “I wont change my ways” playchartless, player pusher Genious on the sideline. Only concerns are Auburn (although in Athens) and both SC and LSU games on the road. Still, LSU is #17 and had changes at OC, QB, QB Coach and lost Guice etc…not to mention Orgeron just flat out admitted publicly that his recruiting year stunk. This should be the perfect schedule to get plenty of the the young guys on the field so they can get early season reps for a post season push.

    Like

  2. Gaskilldawg

    I am ignorant about his methodology so I have an ignorant question. Does the weighted five year average mean that our 2013 season affects our projection for 2018?

    Like

    • Start here, and then clink on the link in my post.

      And note what Bill says in what I linked to today:

      For recent history, I get a little weird. I found last year that the previous year’s S&P+ ratings — which make up the starting point for the returning production figures — were carrying a little too much weight. So what you see below is a projection based solely off of seasons two to five years ago. Recent history now carries less weight in the overall formulas, under 20 percent.

      Like

      • Macallanlover

        Reducing the impact of the prior year seems counter intuitive to me, especially since it increases the weight of years where fewer players were even involved. When you add staff changes at so many schools, going back 5 years really makes the older numbers less reliable to me. I appreciate the effort but this task is pretty much Mission Impossible anyway.

        On the production lost data, it seems over weighted because it isn’t really a total loss. As great as Sony and Nick were, the combo of Swift/Herrien/Holyfield would have picked up 85%+ of that yardage, in my opinion. Not minimizing how significant those lost yards would have been if we didn’t have them in certain games, but I am pretty confident UGA will rush for similar yards in 2018. Maybe that is factored in at a lesser rate and I just don’t see it.

        Like

        • Macallanlover

          OK, I see there is a factor involved in how lost production is accounted for, and rushing yardage is treated differently from passing yardage, but it still seems too high to me. Doesn’t matter, I still feel we will be better offensively in 2018, by a decent amount. I just cannot ‘splain the methodology I used (SWAG) to a guy like Bill. 🙂

          Like

        • Charles

          I think Bill reduced the weighting of the most recent year based on empirical data. He found the model fit better with a different balance. You can come up with reasons for why or you could say that he’s overfitting.

          Like