Started By
Message

re: Graphic: FPI Probability of wins by week

Posted on 8/16/18 at 10:32 pm to
Posted by moneyg
Member since Jun 2006
56532 posts
Posted on 8/16/18 at 10:32 pm to
LSU fans who don't think LSU is facing a tough schedule this year need to really study that chart.
Posted by Greace
Member since May 2009
4696 posts
Posted on 8/16/18 at 10:50 pm to
Schedule has always been tough. This is nothing new.
Posted by Bucks2TigerFan
Member since Jun 2018
825 posts
Posted on 8/16/18 at 10:57 pm to
FPI is ESPN's pet statistical projection system and it has some built in biases.

You ever heard the expression GIGO, Garbage In, Garbage Out. That's my take on ESPN's FPI.

They need several weeks of results from the current season for it to actually produce somewhat realistic results. It's still just a fancy statistical guessing game right now.

ESPN uses a weird system where each team has a different home field advantage, so it gets really complicated to try and figure out what the actual point spreads that go with those numbers are.

But they do publish the raw numbers that indicate what the neutral field point spread for each matchup would be.

Using their numbers, if all these games were at a neutral field, then LSU would be favorites in only these games:
LaTech by 13, Ole Miss by 4, Arkansas by 6, Rice by 33, and SE La by 33.
They would be neutral field underdogs in all these games:
Miami by 6, Auburn by 10, Florida by 2, Georgia by 13, Miss. State by 5, Bama by 13, and Texas A&M by 2.5.

Sorry, I'm not buying it, but if you take pleasure in hearing bad things about LSU, then you should read these BS numbers and be very happy.

Posted by emanresu
Member since Dec 2009
9368 posts
Posted on 8/16/18 at 11:10 pm to
quote:

LSUHobNailBoot


Posted by Palm Beach Tiger
Orlando, Florida
Member since Jan 2007
29863 posts
Posted on 8/16/18 at 11:29 pm to
Well one thing is for sure.....if we go 5 and 7. 2 and 6 vs the sec like this says we will ...... a lot of people are getting fired. I honestly do not even think the floor is 5 and 7. But I have been wrong before.
Posted by L5UT1ger
Member since Feb 2004
2600 posts
Posted on 8/17/18 at 12:29 am to
Maybe I’m looking at it wrong, but the highest percentage chance of wins on the last row is under 6 wins. With a higher chance we end up with 7 wins than 5 wins.

Or maybe I’m blind drunk
Posted by beauxroux
Baton Rouge, LA
Member since Nov 2010
2144 posts
Posted on 8/17/18 at 6:04 am to
This chart goes to show there are 3 kinds of lies:
1. Lies;
2. Damn lies; and
3. Statistics
Posted by Lebowski
Dallas
Member since Oct 2013
3544 posts
Posted on 8/17/18 at 6:13 am to
Alaways next year. Keep moving the bar. Lol
Posted by moneyg
Member since Jun 2006
56532 posts
Posted on 8/17/18 at 6:51 am to
quote:

Schedule has always been tough. This is nothing new.



Do we always have two home games in which we are a 4-1 dog? And another home game which we are a small dog...and, a road SEC game where we are about a 6-1 dog...and another neutral site game in which we are a 2-1 dog?

Posted by drdoct
Atlanta, GA
Member since Oct 2015
1609 posts
Posted on 8/17/18 at 7:31 am to
quote:

Do we always have two home games in which we are a 4-1 dog? And another home game which we are a small dog...and, a road SEC game where we are about a 6-1 dog...and another neutral site game in which we are a 2-1 dog?



No, normally we're better and so less of a dog. Last year a lot of the SEC was down. The only difference between them and us is they got good coaches and we went cheap. Last year we had a chance to take advantage of this down SEC year and ended up sh!tting the bed against teams we had no business losing to while beating the down SEC teams.

Us being so far of underdogs is LSU's fault, not the schedule.
Posted by JayDeerTay84
Texas
Member since May 2013
9847 posts
Posted on 8/17/18 at 8:11 am to
So when we pack Miami's shite in game one, what does that do to this model?
Posted by Bucks2TigerFan
Member since Jun 2018
825 posts
Posted on 8/17/18 at 8:15 am to
quote:

So when we pack Miami's shite in game one, what does that do to this model?

It changes and gives a more realistic idea of LSU's chances, but even then, it takes at least a month of games for it even to be half good.
Posted by Cashman
Palm coast, FL
Member since Apr 2008
130 posts
Posted on 8/17/18 at 8:57 am to
So, we have a 98.6 probability to beat SlU, but only a 65.3 probability to have one win after 2 games? If we have a 98.6% chance to beat SLU, our chance of having one win after 2 games should be at least 98.6. Something doesn't quite add up. This is total crap
Posted by TigerFan55555
Tomball, TX
Member since Nov 2008
9583 posts
Posted on 8/17/18 at 11:00 am to
quote:

Man a lot of you are in for a rude awakening this season if you think he's that far off.

That UF to Bama stretch is brutal.




hell, we havent won at AR in years!, dont count them out of that stretch either! very tough!

And TAMU is going to be better than alot of people think... I think they are going to be the surprise this year... very quiet with whats going on over there...

Posted by jrodLSUke
Premium
Member since Jan 2011
22178 posts
Posted on 8/17/18 at 11:02 am to
That's pretty cool info. It would be interesting to see how another program compares to LSU; maybe Auburn's?
Posted by sabanisarustedspoke
Member since Jan 2007
4947 posts
Posted on 8/17/18 at 11:10 am to
quote:


So we have a 0.0% chance to run the board... so you’re saying there’s..not a chance?

And a 6% chance we start off 3-0, this is dumb





I don't know man, he did say a guy on Reddit put it together...
Posted by Greace
Member since May 2009
4696 posts
Posted on 8/17/18 at 11:12 am to
Click the link in the OP its got every CFB team
Posted by NAsh-vegas Tigah
Franklin, TN
Member since Jan 2004
2329 posts
Posted on 8/17/18 at 11:59 am to
That looks about right. I think this team with Zereaux leading the charge is more 4-8 than 8-4. He can’t be gone quick enough for me.
Posted by emanresu
Member since Dec 2009
9368 posts
Posted on 8/17/18 at 12:14 pm to
I get the sneaking suspicion that some are missing where it says that this is based on FBI. The guy on reddit didn't subjectively pull the numbers out of the nether.
Posted by tmjones2
TX
Member since Feb 2013
1511 posts
Posted on 8/17/18 at 12:44 pm to
If you want to get the probability of at least 1 win, you have to add up 1 win and 2 wins, so it's like 99.2% or something.
first pageprev pagePage 2 of 3Next pagelast page

Back to top
logoFollow TigerDroppings for LSU Football News
Follow us on Twitter, Facebook and Instagram to get the latest updates on LSU Football and Recruiting.

FacebookTwitterInstagram