Gilliam Ratings Updated - New #1 team might surprise you.
- Thread starter GilliamRatings
- Start date
I just looked over and I am shocked we are not on the bottom!!! I was a little surprised that Brookville was number 1 overall for your rating system. Good for that coaching staff especially Coach Meeks he took a lot of criticism at his old school that was unfair according to people in that area.
It is early in the year, I do not think Brookville will stay number one by the end of the regular season. Gilliams ratings tend to pan out by the end of the season.
Phoebus is easy to explain. My computer thought they would win 48.71 points. The way the ratings work they'd have to win by 49 or more for the computer to improve their rating (makes sense, that would mean they played better than expected), due to blow out correction formulas it would leave their rating alone if they won by 45-48 points (figuring they underperformed, but we could attribute that to running clock, calling off the dogs etc.), but if they won by 44 or less they were going to drop. Their 34 point win meant they underperformed by nearly 15 points so their rating was going to drop.
Staying in the stratosphere in these ratings is hard. You've got to absolutely destroy bad teams or the computer gets suspicious. People might not like that, but it's proved to make the predictions much more accurate over the years to require big blowouts when predicted.
So Phoebus drops enough in the rankings to fall from 2 to 5. This week the Phantoms are favored by 27, if their rating is going to go up they need to win by more than 27.
They don't have as hard of a job keeping their rating as some schools do. Later in the year L.C. Bird is favored to defeat George Wythe by 107-0. Even with my blowout adjustments, Bird has to win that game by at least 74 points in order to hold their rating. They actually won by that much the last couple of years, so we'll see what happens.
It's all explained here.
They also bounce more than polls. In polls the #1 team tends to own the spot until they lose, but not in predictive ratings systems. That's why Vegas uses systems like mine and not polls to set initial point spreads (there's a little more to it, but systems like mine are generally the first step).
You can also see exactly how well the ratings did predicting game each week if you click on the predictions link. I leave up all the predictions I made and put the real score right next to them and then give you a statistical summary. Nothing to hide here. It is what it is.
Which explains my point clearly in the either topic. Early season rankings and ratings are pointless until the teams lineup on the field.
Yep, it's just for fun and to get some conversation going. As dead as the boards have been I'll be glad to take all of everyone's buckshot right now if it'll get people posting.
Gotcha, Mr Gilliam. I've been following your work for years. You have been and still is great at what you do.
