"highly accurate system for predicting records" | Syracusefan.com

"highly accurate system for predicting records"

moqui

generational talent
Joined
Aug 25, 2011
Messages
12,094
Like
25,580
When formulating any statistical rating system, the most important step is removing all human bias from the equation. I use Strength of Schedule, Home Advantage, and Margin of Victory data that is straight from the actual game results. I run 50 iterations of my formula to make sure that all levels of the Strength of Schedule and Margin of Victory are balanced, and what comes out is a highly accurate system for predicting future results based on past performance.

So claims the creator of the Logan Sports Rating System.

Logan currently has Syracuse ranked 32nd. And predicts them to finish at 24-7. That's right - it's a 10-7 finish from here on out.

 
Last edited:
Margin of victory will get you every time. LOL
 
Duke lost to ND last weekend and moved up 8 positions :rolling:
 
Not only does everyone in the media hate us, now inanimate objects do as well.
 
So claims the creator of the Logan Sports Rating System.

Logan has Syracuse currently ranked 32nd. And predicts them to finish at 24-7. That's right - it's a 10-7 finish from here on out.


Presumably what's driving these numbers here and in some of the other computer systems is the relatively poor performance early in the season. Most of us have mostly forgotten the pre-Maui games, but the computers haven't. I mostly think we're right to forget about them, but I'll admit to a little worry that those games remain relevant data points.

Not to say these rankings make sense - it has Syracuse behind 9-7 BYU most blatantly. My total wild guess at the reasons for the weirdness is that it's giving too much of an adjustment for road games (to the point of almost ignoring road losses).
 
So how does it compute injuries/people returning from injury? Or our 6th man sitting out opening night?
 
I was going to check his accuracy, but he makes all his picks using team nicknames, and for most of the small schools, I have no idea which team he is selecting.
 
So how does it compute injuries/people returning from injury? Or our 6th man sitting out opening night?

In most cases they account for that the future results, more recent games get a larger weight.
 
Presumably what's driving these numbers here and in some of the other computer systems is the relatively poor performance early in the season. Most of us have mostly forgotten the pre-Maui games, but the computers haven't. I mostly think we're right to forget about them, but I'll admit to a little worry that those games remain relevant data points.

Not to say these rankings make sense - it has Syracuse behind 9-7 BYU most blatantly. My total wild guess at the reasons for the weirdness is that it's giving too much of an adjustment for road games (to the point of almost ignoring road losses).
Logan's system is an extreme outlier. Many of the Sagarin-like systems have SU ranked in the teens, but only 3 of the 50 ranking systems tracked by Ken Massey have SU out of the top 20, and of those three, only Logan has SU out of the top 25.
 
Logan's system is an extreme outlier. Many of the Sagarin-like systems have SU ranked in the teens, but only 3 of the 50 ranking systems tracked by Ken Massey have SU out of the top 20, and of those three, only Logan has SU out of the top 25.

Yea, sorry, I just meant the reason the computer rankings don't like Cuse quite as much as the people rankings.

The Massey site is really cool. I hadn't seen it before. I love that they're somehow even tracking the Canadian teams - not too surprisingly, Carleton is actually better than many of the U.S. teams SU has faced so far (and Boston College!).
 
Baylor is also not a numbers darling due to the fact that they only beat South Carolina by 2 at home, Charleston Southern by 5, and Northwestern St by 7. That is not good at home.

I look to use ranking systems, but I also combine in with my view of results. For Baylor, its performance against top teams outweighs its performance against lesser likes.
 

Similar threads

Replies
6
Views
408
Replies
9
Views
602
Replies
5
Views
515
Replies
5
Views
665
    • Like
Orangeyes Daily Articles for Wednesday for Basketball
Replies
6
Views
435

Forum statistics

Threads
167,507
Messages
4,707,673
Members
5,908
Latest member
Cuseman17

Online statistics

Members online
249
Guests online
2,104
Total visitors
2,353


Top Bottom