Quote from: AO on February 27, 2018, 10:11:38 PMQuote from: Ryan Scott (Hoops Fan) on February 27, 2018, 07:29:32 PMYes you can still have a "good" SoS with a 20 game schedule but it will require your conference to win a greater share of their 5 non-conference games since you have more losses built into your SoS. A NESCAC team like Middlebury can replace one of their sub .500 conference foes like Colby with a 19-8 Morrisville State. Colby will also have the opportunity to not hurt the NESCAC SoS as much since they can rack up more wins against non-conference foes. The maximum possible SoS is lower when you have more conference games.Quote from: AO on February 27, 2018, 03:07:24 PMQuote from: Ryan Scott (Hoops Fan) on February 27, 2018, 12:17:33 PMI think the easiest way to say it is "the more conference games you play, the closer you'll be to .500" Since the NESCAC only has 10 games, they've got more opportunities to improve their SOS. If the NESCAC was terrible they might play bad teams and get beat and have the worst SOS in the nation. But since they're all pretty good they can beat winning teams from relatively poor conferences and inflate their SOS.
Maybe one of your math guys can explain it for me, but we often talk about how the conference schedule pulls an SOS closer to .500 - and I get the rationale for that, since every conference game ends up being a wash (.500) in SOS terms. However, I'm wondering if a power conference has, say, an average SOS of .580 going into conference play and all or almost all the teams have winning records, wouldn't the pull be less, or even actually help the SOS in some cases, since it's both OWP and OOWP?
The NESCAC's the best example - they get a boost in conference play for their SOS, right? Because they play so many non-con games and their records and SOS numbers are so high to begin with.
Does that translate to other conferences, like the UAA or the WIAC where often all the teams are pretty good?
It would also work in reverse for really bad conferences, right?
The committee knows the NESCAC numbers are inflated and might have taken that into consideration when they put Amherst 10th in the regional rankings.
But that's what I'm saying, I'm wondering if that "more conference games you play, the better your SOS" is actually true for really good conferences. I mean, at some point, if you play enough conferences games, it would eventually get to .500, but what if that number is 50? I think it's conceivable you could play even 18 or 20 conference games and still have a decent SOS if the entire conference did really well against good teams in those 5-7 non-con.
I was just hoping someone better at math could explain it a little better. I can sort of get the general concept in my head, but I don't have the terminology to explain it in detail.
For example, if your nine team conference entered conference play with everyone having between 7-2 and 9-0 records with an SOS between .550 and .600. I feel like you'd get pretty far into a 16 game regular season before the SOS started regressing to the mean. Maybe I'm wrong, but I was hoping one of our resident mathematicians would weigh in.
To your point regarding the NESCAC, rumor has it the league is considering doing "home & home" in conference in coming seasons (probably not next b/c of contracts already signed but 19'-20'). Going from 10 to 20 conference games and limiting their non-conference games to 4 (I believe NESCAC decided to play 24 , they might change if this happens). Again to your point I hear coaches within the league are not excited about the possibility.
LF