Re: Why does it matter if they’re biased? And

I hear a lot about ESPN's bias, and a lot of people argue that it is imagined. The fact that ESPN would hire Bill Connelly, last year, shows that it exist. He has done little to hide his bias to the Big Ten and SEC unless you count portraying the S&P+ that he engineered as an unbiased computer ranking.

His S&P+ rankings after the season last year still had Clemson at number 2 behind Alabama and had a 3 loss Georgia 3rd. I have heard a couple of people say that it was not a new ranking but the one that I looked at was dated for Jan 7th and had updated records. What might be worse is the S&P+ 5 year ranking for 2014-2018 (inclusive). Somehow despite 2 National Championships, 4 CFP appearances, and a 31-0 shut out of Ohio State in our last meeting, they still rank higher in those "computer" rankings.

Hiring Connelly in January, to me, proves an effort to push the SEC and B1G that is independent of saying something controversial for ratings. Of course, Connelly can still get a little of that in too. His playoff teams after this week: 1) Ohio State 2) LSU 3) Alabama 4) Penn State (2 B1G teams and 2 SEC teams)

quozzel

Hall of Famer [20540]

TigerPulse: 100%

Posts: 11686

Joined: 10/15/02

Re: Nothing shows ESPN bias more than hiring Bill Connelly

Oct 27, 2019, 5:47 AM

Those fancy maffs formulas are good, as they go. But they're missing a key metric and Clemson has borne that out numerous times over the past five years.

That metric is very simple - number of players played in key situations. Clemson leads the nation going away in not just players played per game, but the number of players who received quality snaps while the game was still in question.

Mind, this tends to happen at the detriment of the other stats, especially early. When you're playing so many guys it does disrupt continuity and because a significant chunk of those guys are young or inexperienced you're going to be living with a bunch of newb mistakes, and all at once, especially in September.

This is why Clemson historically has not looked tremendous under Dabo in September. But it's October now and Clemson's starting to take on its midseason form. By November Clemson tends to be utterly dominant and by the post-season Clemson is an absolute nightmare.

Why? Well, remember that old kid's game called Battleship? Clemson's basically playing that game with three full sets of pieces, that the Tigers rotate basically interchangeably from about Week 4-5 on...and that spreads the enemy hits out among those three full sets of pieces. When a guy gets a ding, Clemson can hold him and let somebody else take the abuse until that guy heals up. And because Clemson spreads those dings out among so many players they're taking far fewer hits per player anyway.

I don't know the exact per-player stats but I would be willing to bet you Clemson's starters only get about 60-65% of the snaps of an average D1 starter...and that means they take a lot less wear and tear over the course of a season. Which is why Clemson tends to look so incredibly frisky at the end of a season while it appears our opponents are held together with duct tape, bailing wire, and cortisone injections.

The fancy maffs don't take that into account at all, which is IMHO the biggest reason Clemson is almost always under-rated on the likes of S&P+...and why the fancy maffs tend to miss on Clemson every year as well. If they adjusted the true value of that particular metric I think Clemson would rate far, far higher.

Their model's not really biased since math by definition isn't; it's just incomplete.

flag link

military_tech

thumb_down

thumb_up

Nothing shows ESPN bias more than hiring Bill Connelly

Re: Nothing shows ESPN bias more than hiring Bill Connelly

Re: Nothing shows ESPN bias more than hiring Bill Connelly

Why does it matter if they’re biased? And

TigerBoard POTD