Scheduling again (this time with less typos!)

[CBB_Fan]

Starting a new thread just because I was annoyed with "&H" at the end of the last thread's title (somehow jumped up to the top when typing H&Hs). Anyway, I did some more work trying to figure out what makes a good schedule. Why is KU #1 in SOS, but Texas #56? Why do Wichita State, SDSU, Louisville, and Gonzaga have low SoS numbers?

First, I'll show the current top 25 and their SoS numbers (numbers from ESPN):

1. Syracuse - 69
2. Florida - 37
3. Wichita St - 103
4. Zona - 9
5. Duke - 11
6. San Diego State - 129
7. Cincinnati - 82
8. Kansas - 1
9. Villanova - 22
10. St Louis - 61
11. Creighton - 24
12. Louisville - 112
13. Michigan St - 34
14. Virginia - 32
15. Iowa - 71
16. Wisconsin - 2
17. Iowa St - 25
18. Kentucky - 5
19. Texas - 56
20. Michigan - 7
21. Connecticut - 67
22. Memphis - 52
23. UCLA - 16
24. Ohio St - 6
25. Gonzaga - 106

I've bolded the "weak" schedules, the ones above above 50. Three of those were power conference teams, one each from the ACC, 1 B1G, and 1 Big 12. We also have an MVC team, an A10 team, a WCC team, and 3 teams from the AAC. The power teams seem to be the exception, but teams from the other conference all have low SOSs. Are these discrepancies the result of playing more bad teams, or less good teams? For this breakdown though I'll ignore mediocre teams, and look at top 100 teams and bottom 150 teams.

Syracuse: 14 < 100, 8 > 150
Wichita State: 5 < 100, 10 > 150
SDSU: 9 < 100, 8 > 150
Cincinnati: 11 < 100, 14 > 150
SLU: 10 < 100, 13 > 150
Louisville: 7 < 100, 15 > 150
Iowa: 13 < 100, 8 > 150
Texas: 16 < 100, 7 > 150
UConn: 12 < 100, 12 > 150
Memphis: 10 < 100, 14 > 150
Gonzaga: 10 < 100, 14 > 150

This data set alone is not enough to get a conclusion. Now I'll look at the top SoS teams: Kansas, Wisconsin, Alabama, Minnesota, Kentucky, Ohio State, Michigan, Baylor, Arizona.

Kansas: 21 < 100, 3 > 150
Wisconsin: 17 < 100, 4 > 150
Alabama: 14 < 100, 6 > 150
Minnesota: 12 < 100, 7 > 150
Kentucky: 16 < 100, 6 > 150
Ohio State: 16 < 100, 5 > 150
Michigan: 15 < 100, 4 > 150
Baylor: 15 < 100, 8 > 150
Arizona: 15 < 100, 8 > 150

While the games against the Top 50 is impressive for some teams, it doesn't explain the why Syracuse or Texas have worse SoS numbers, despite playing a similar number of top teams. The biggest difference is that most of the top teams have played very few bad teams. So, let's look shift to looking at conferences. I'll divide these up with good teams < 50 RPI, mediocre teams between 50 and 150, and bad teams below 150. First, the conferences that seem to generate bad SoS numbers:

A10: 5 good teams (RPI < 50), 3 mediocre teams (50 > RPI > 150), 4 bad teams (RPI > 150)
AAC: 5 good teams, 0 mediocre teams, 5 bad teams
MVC: 1 good team, 4 mediocre teams, 5 bad teams
WCC: 2 good teams, 5 mediocre teams, 3 bad teams

And the power conferences:

Big 12: 6 good teams, 3 mediocre teams, 1 bad team
B1G: 6 good teams, 6 mediocre teams, 0 bad teams
Pac 12: 7 good teams, 3 mediocre teams, 2 bad teams
SEC: 4 good teams, 6 mediocre teams, 4 bad teams

The A10 and AAC are competition at the top with any of these teams, but they have more baggage at the bottom. This further highlights the flaw with SoS: it depends more on NOT playing bad teams than actually playing top teams (though both are important). And that is a flaw because it attempts to gain predictive value out of games that by definition are not supposed to be competitive. In other words, using SoS as a measure of team strength is similar to saying that because Wichita State played Evansville, they would lose to Michigan. Not because they won or lost, but that they played the game.

What is worse is that low SoS games have MORE strength than high SoS games. In other words, playing the #1 team and the #350 team is considerably worse than playing the #101 and #102 teams, even though the prior example is the only example of a game that would be considered a competitive game for a top team. IMHO, the farther away a team is from its opponent's RPI, the less the game should matter for either RPI or SoS, simply because those games are less predictive than games that are roughly close in value.

The point is that the biggest factor for SoS is avoiding bad teams, and the best way to do that is to play in a good conference; if you play in a non-power conference you need not apply. It is somewhat valid for teams like Syracuse, Texas, and Iowa; all of those teams have the opportunity to avoid playing bad teams by scheduling well in the OOC. For teams in conferences with 4-5 bad teams, they are guaranteed 8-10 bad RPI games, enough to knock them from a good RPI regardless of the number of good teams they've played.

We have been hurt by this more than any other team, because our conference has given us 0 chances for a good win (we can't play ourselves!) and automatically adds 10 bad RPI games. So we started way behind the ~60 teams in power conferences. Our OOC SoS was in the 40s, but I'm not certain we'd have a top 50 schedule even with the #1 OOC SoS. Essentially, the idea that we should have "scheduled up" to make up for the MVC is ridiculous because it is impossible. Not because Marshall isn't liked, not because we refused to accept buy games, but because the premise of scheduling up is flawed itself.

TL;DR: SoS = Conference Affiliation. Bad games hurt more than good games help, and thus SoS depends most on avoiding H&H matchups with bad teams in the conference, not scheduling up outside of it. OOC scheduling can separate power conference teams from power conference teams, but it cannot make up the gap between conferences like the MVC and the B1G.

[Shock Top]

As ******* and Gotllieb will tell you though, computers can't tell you anything.. I wonder if they have told Billy Beane that??

[Aargh]

Imagine you are an upper-tier D1 BB coach. You are making $1.5+ mill a year- probably $2 mill+.Your job is dependent on making the NCAA tournament, winning some games there, and keeping the fans happy.

You get a call from Gregg Marshall. He's got a highly rated team that doesn't have a lot of name respect among your fan base and he wants a home-and-home series with you. You know that Koch is considered mong the hardest places to win a game. Do you take the challenge or do you try to find a reasonably respectable road game that might be an easier win? Or do you just schedule an esier road game against a team with more name recognition, where you fan base won't turn on you if you lose?

Let's assume you didn't want to take a road trip into either Koch or IBA. So you're offered a home game and actual neutral game in KC as the "return" game. Do you take the home game against the team with the winningest road record in D1 ball over the last 3 years?

It's my opinion that not many upper-level coaches want even that series. so, how many would even give the Shox a reasonable guarantee game at maybe $120K. SWAC teams are getting $80K. Putting the SHox on your home court has to be worth more than that. Do you risk alienating your athletic department and fan base by giving a huge guarantee to the #1 road team in D1 ball when you can get games against lesser teams with less risk of losing the money and the game?

Until fans of "name" teams can accept a loss to WSU, it's going to be difficult for coaches to schedule the game.

[DUShock]

Unfortunately Aargh is correct about the whole perception thing. Another deep run in the Dance will go far to broaden our brand.

Go Shocks!

[im4wsu]

Unfortunately Aargh is correct about the whole perception thing. Another deep run in the Dance will go far to broaden our brand.

Go Shocks!

Why can't all the "name" schools just do what the "little" schools do and make this their Super Bowl -- let me be the team to knock off Wichita State?