Somehow, LSU finishes third in ESPN’s FPI rankings, behind the team it just waxed by seventeen. Even Sagarin’s computer managed to get things straightened out in its last list of the season.
Well done, Mickey.
Filed under ESPN Is The Devil, Stats Geek!
Some ESPN guy pregame last night was using FPI to predict that Clemson would win. I’m beginning to believe FPI is voodoo statistics.
That makes zero damn sense. In a season where you beat Alabama, Clemson, Georgia, Florida, TA&M, Texas, and Oklahoma (and dominating most of them), with an all-time season from your Heisman winning QB, finish 15-0, and win the national title game by 17 points, HOW are you the third best team?
What kind of Mickey Mouse operation is ESPN anyway?
LikeLiked by 5 people
Post of the week!
That rating can’t account for the weakness of OSU and Clemson’s schedule vs the strength of LSU’s schedule. The problem is the rankings fail to reflect the middle of the SEC’s relative strength. Teams 4-9 are undervalued while those teams in other conferences are overvalued.
I love the showing the SEC had for this bowl season. 8-2 looks good for the conference. Of course, we can always count on Auburn keeping us humble.
More evidence as to why those who want “computers” to pick the playoff teams should STFU.
Garbage in, garbage out. It’s operator error.
I noticed before the game started that ESPN gave Clemson a 56% chance of winning. This would explain what I thought was an odd prediction.
IIRC once Clemson got up 14-7 it was at %60.
Can you imagine if that was Georgia?
The NC trophy would have been surrendered before the sun came up this morning.
Have no idea why this post was here^^ It was meant to appear on the thread showing LSU players being paid.
(what I said earlier about cutting Stoop slack because there’s no “delete or edit post” feature…)
No, it is in the right place. McDoofus would have surrendered the trophy if the WWL had said we were undeserving and only ranked third. You wouldn’t expect him to challenge anything on UGA’s behalf, would you? He has never stood up for the program before. Surrender monkey, imo.
And now you know that when I am scouring the interwebs looking for insight on upcoming football match up analysis to challenge my own thinking before placing bets, I have honestly never looked once to the WWL. This huge company, which controls so much of the CFB inventory, and employs thousands of people, many with access to the coaches and practices, is no more reliable than Mr. Anonymous message board poster. Known for their bias, why would any bettor seek them out when looking for objectivity?
This ranking above done by their computer is just like Sagarin and the old computer polls, the garbage out portion of that old saying about garbage in. People mistakenly think computers are objective, forgetting there has to be an input process and programming…in this case, done by people who have biases.
Bill Connelly at least tries to challenge his results and tweak it each season in an attempt to make his projections line up closer with the results. But the formula is still based on his theory about what factors, and at what weight, will lea to the result. So not totally objective, but with a recognition it might be flawed, and a goal of minimizing the errors.
ESPN’s recruiting is just as bad. And thanks for ever having the NC game that doesn’t end until midnight on a Monday. What kind of Mickey Mouse horseshit is that?
Btw, what are we doing playing a game on Monday? It’s a holiday but ugh.
LikeLiked by 1 person
Holiday is next Monday. Yesterday was a regular day. It just means more.
I was referring to our game vs Virginia being on a holiday.
I just love scrolling all the way down to 110….bwahahahahaha
Statistics don”t factor in physicality. LSU assaulted them until they wilted. Excellent coaching by Clemson’s staff, made it closer than it really was.
Georgia #22 on offense and #50 on special teams. Cue Danny Kannell.
Why are they still pushing FPI after hiring Bill C.? Especially since he A. still computes SP+, and B. SP+ has been a better metric than FPI?
Lucy, tell your statistics to shut up.
The FPI ranking looks like Gator or NATS logic to me.
Never been super crazy about FPI. It’s game score metric seems to be HEAVILY weighted by the competition you play (so even if you play really well, if the opponent wasn’t good, you get dinged), and it has put out these odd rankings throughout the season. Under no circumstances should Clempson have been favored vs. LSU, which is was FPI was predicting pre-game.
SRS (simple rating system) has had a similar issue this year ranking anOSU as an all-time great team up until the Fiesta Bowl, and even after that loss they remain at #1 end of season. I think the schedule ranking component is simply off with these metrics.
Who started the rumor that Vegas’s spread dropped to -4.5 if LSU hadn’t been pinned in their end zone for 1-1/2 quarters…. it could have been 30+ on the final.
“It’s a lot more fun that way when you’re explosive,” Monken said. “It’s funny how it works.” — AB-H, 9/1/20
Subscribe in a reader