New to the thread and board, wanted to add some points to the earlier discussion about the matrix.
On my iPad, I have a hard time with that site, so I can't really link anything and I have a tough time finding certain pages, but it may be behind a paywall on that site anyway.
As I recall, from either reading it or listening to him on 1620, Bartoo uses an attrition-adjusted recruiting ranking average for the talent portion of the matrix. This is done in March. His goal was to make prediction as simple as possible, and that's 'show he gets there. Not perfect, but over time it has shown to be pretty damn accurate. Talent is #1, then if that's close, location home/away, and #3 is coach effect. Win a game you shouldnt against a more talented team and it's a plus one, lose one against a schlub and you get a -1. Favored and win or 'dog and lose is neutral. Obviously, if you are rarely favored, it's tough not to have a positive CE if you win a few games, conversely if you are favored most of the time, it's easier to end up with a negative.
Bo was a +1 after his first few years, ended up -0.5 ish. Riley, IIRC, used to be in the +2, +3 range, which was really really good even though he was rarely favored to win based on talent at OSU, but slipped to just +1 at the end of his time there. Now that he is favored nearly every game, he's the worst in the country at - 4 this season.
The matrix predicted NU at either 10-2 or 11-1 this season, sorry can't find the page to link, and that's using a -1 CE for a new Hc. The only game we've lost that we should have this season is Miami. Pretty sure the matrix has us beating MSU, and Bartoo didn't back off of that last week (because he doesn't change anything in-season). We have had more talent than every team except Miami, adjusted for attrition. He has said in the past that NU has a pretty low attrition rate compared to most, which usually has our talent slightly better than the avg recruiting rankings. Northwestern actually is a team that routinely has much better talent than recruiting ranks because of super low attrition in the matrix.
One thing the matrix cannot do ...It does not adjust for injuries, because it's done in March. No doubt those have had an impact, just as they did the last several years going back to that Steinkuhler knee that led directly to the CCG debacle, ameer's dings, etc.
Riley is failing HUGE this year based on the matrix. That much is clear. Bartoo's formula is simple and it's pretty darn accurate over time. NU has had the talent to win every game, and was favored in all but one. This is just an epic disaster.