UPDATED 4/1/2007 @ 17:00 (Scroll to Bottom)
Since this is the third time that I’ve documented the OOC strength of schedule for the ACC, I couldn’t come up with anything new and interesting to put into an introduction. But I can point new-comers to the first two entries on OOC SOS and especially to the comments on these entries.
1999-2005 OOC Summary
1999-2006 OOC Summary
If you don’t remember the standard complaints, please take a minute and review these previous entries. Let’s all endeavor to come up with something new to say and not just mindlessly rehash the same comments.
Thanks to kenpom.com we now have a nine-year summary on OOC SOS:
OOC Strength of Schedule |
||||||||||
|
’99 |
’00 |
’01 |
’02 |
’03 |
’04 |
’05 |
’06 |
’07 |
AVE |
Duke |
15 |
33 |
17 |
59 |
108 |
24 |
46 |
6 |
10 |
35 |
UNC |
30 |
17 |
42 |
34 |
46 |
29 |
54 |
70 |
7 |
37 |
UMd |
51 |
15 |
115 |
15 |
189 |
43 |
125 |
64 |
40 |
73 |
WF |
202 |
47 |
74 |
24 |
123 |
87 |
27 |
148 |
258 |
110 |
BC |
284 |
142 |
116 |
176 |
51 |
65 |
88 |
282 |
51 |
139 |
GT |
114 |
204 |
85 |
109 |
142 |
106 |
137 |
241 |
160 |
144 |
FSU |
21 |
98 |
178 |
227 |
163 |
179 |
98 |
316 |
111 |
155 |
UVa |
198 |
252 |
210 |
133 |
147 |
257 |
59 |
159 |
175 |
177 |
UM |
47 |
169 |
203 |
225 |
261 |
317 |
142 |
177 |
274 |
202 |
NCSU |
249 |
239 |
126 |
207 |
238 |
82 |
259 |
206 |
270 |
208 |
CU |
272 |
200 |
301 |
288 |
293 |
53 |
155 |
152 |
214 |
214 |
VT |
287 |
209 |
307 |
304 |
260 |
304 |
312 |
290 |
76 |
261 |
AVE |
148 |
135 |
148 |
150 |
168 |
129 |
125 |
176 |
137 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
DISTRIBUTION |
|
|
|
||||
1-99 |
5 |
5 |
4 |
4 |
2 |
7 |
6 |
3 |
5 |
|
100-199 |
2 |
3 |
4 |
3 |
6 |
2 |
4 |
4 |
3 |
|
200-299 |
5 |
4 |
2 |
4 |
4 |
1 |
1 |
4 |
4 |
|
300+ |
0 |
0 |
2 |
1 |
0 |
2 |
1 |
1 |
0 |
|
COMMENTS
– For those of you lucky enough to live close to a Wolfpack Caravan stop this summer, it would be interesting to ask Sid or Fowler if there are any plans to improve State’s OOC schedule. If memory serves, some lip-service has been paid to improving the schedule…but the results are a little slow in coming.
– For the first time since we’ve started looking at OOC SOS, no ACC team was adversely impacted by their choice of a weak schedule. VT (2005) and FSU (2006) certainly learned the consequences of weak OOC schedules in recent years.
Schedules of interest:
– WF and State had their weakest OOC schedules over the years archived at kenpom.com.
– VT had their best OOC schedule by far. We will have to see if this improvement was a fluke, a one-time choice with a senior-laden team, or if it represents a philosophy change in Blacksburg.
– If FSU had played the ’07 OOC schedule in ’06, that team would certainly have gone to the NCAAT instead of the NIT. I think that this goes to show that since you never know how a season is going to unfold, you don’t want an OOC schedule that turns into a big negative on Selection Sunday.
– UVa’s schedule was the weakest among the teams in the top-half of the conference this year….but they still tied for the regular season championship. I’ll just add this example to my list of reasons why I say that a tough schedule reveals good teams, it does not produce them.
We’ve been over the reasons why a weak SOS can come back to haunt a team, but we’ve never discussed in detail the mistakes made when setting up a given schedule. Since the Herbavores are no longer circling incessantly, I think that we should be able to take a closer look at State’s OOC schedule without a lot of emotion and posturing.
WARNING: For those that can’t handle ninth-grade math, this would be a good time to click on your next link of interest.
The SOS is calculated by the following formula:
2/3 x (Opponent’s Winning Percentage) +
1/3 x (Opponent’s Opponent’s Winning Percentage) =
Strength of Schedule
Notes:
– The game(s) against the given team are removed when calculating the Opponent’s Winning Percentage.
– Obviously, the OOC SOS only includes the OOC games.
– A team’s overall SOS is part of the RPI calculation.
– A more detailed explanation of SOS and RPI calcs is given here at kenpom.com
So here is State’s 2007 OOC schedule, each team’s adjusted winning percentage, and each team’s individual SOS.
|
Adjusted |
|
|
W/L |
SOS |
 |
0.1154 |
0.4960 |
 NC |
0.2500 |
0.5089 |
 Wofford |
0.2692 |
0.4610 |
 |
0.2857 |
0.4920 |
 |
0.3200 |
0.4663 |
 |
0.3448 |
0.5609 |
 Mount St. Mary’s |
0.3667 |
0.4636 |
 |
0.5172 |
0.4814 |
 NC |
0.5517 |
0.4477 |
 |
0.6333 |
0.5630 |
 Delaware St. |
0.6452 |
0.4668 |
 |
0.6563 |
0.5694 |
 |
0.6897 |
0.5307 |
|
|
|
Average: |
0.4342 |
|
Average w/o ECU: |
0.4608 |
|
|
|
|
Minimum: |
0.1154 |
0.4477 |
Maximum: |
0.6897 |
0.5694 |
If I had no life whatsoever, I could go through each opponent’s schedule and figure out their opponent’s winning percentage to get an exact look at each component of the SOS formula. In place of that boring, time-consuming process, I listed each team’s SOS. These SOS numbers are not directly used in calculating State’s SOS…but do give some indication of what type of competition State’s OOC opponents faced.
More importantly, these SOS numbers illustrate that there is relatively little variation in the 33% of the SOS formula that State has no control over. State’s SOS rank is not due to bad luck, but is a reflection of choices made by the Athletic Department. Hopefully it is obvious that playing a top team from a weak conference (Delaware St) is much better than a weak opponent from a stronger conference (ECU).
It should be obvious to all State fans 15 years old and up, that the best way to improve State’s SOS is to eliminate the games against bad teams. Playing a team with an overall record of 0.500 is much better than playing ECU with a 3-23 record. You have to schedule a lot of teams with winning records to make up for just one team that essentially has no wins.
This is the main point that a lot of people don’t seem to understand. Picking out the best four or five OOC games and drawing conclusions about the SOS can often be misleading. Don’t forget that about 75% of NCAAT teams are seeded within one place of what you would calculate based solely on RPI. Thus chances are high that weak OOC schedules will ultimately hurt a team’s NCAAT seed.
How can you know which teams to schedule before the season starts? Well, it seems quite simple to me. Start by distinguishing between “bad teams†and “bad programs†in your current schedule…and never schedule another game against a bad program. Here are a few examples to illustrate what I mean:
– ECU has not had a winning record in the nine years of data available at kenpom.com. Since joining C-USA in 2002, they are averaging about three conference wins per season…with five wins being their best. In short, ECU is a bad program and should not be scheduled. This has absolutely nothing to do with hating on Pie-Rats, it is simple mathematics.
– Savannah State is another bad program that should not be on State’s schedule. How bad are they? Their eight Div 1 wins in 2007 matched their total Div 1 wins since moving up to Div 1 in 2003. Why is State playing a small school from Georgia that has just recently moved to Div 1?
– The second worst team on State’s schedule was UNC-W at 7-22. However, this was UNC-W’s first losing record since 1999. UNC-W plays in a good mid-major conference (Colonial Athletic) and has won the conference several times in recent years. So, why was this year so bad for the Seahawks? I have absolutely no idea, but someone needs to figure out if UNC-W just had a bad year or if their program is headed for the dumper before they are scheduled in the future.
UNC-W represents another reason to never schedule a game against a bad program. Even good programs have off-years. So quite often you will end up with a few clunkers on your OOC schedule through no fault of your own. Thus you don’t want to intentionally add more “RPI anchors†to the OOC schedule.
In contrast to ECU and Savannah St, Delaware St is consistently near the top of the MEAC and thus represents the perfect cupcake game….a nearly sure win that doesn’t destroy a team’s RPI ranking. Filling your schedule with these types of games won’t impress season-ticket holders or the NCAAT Selection Committee, but it is far better than playing pickup squads from the Sisters of Mercy School for Orphans.
Here’s my idea of what State’s OOC schedule should look like:
– SOS ranking of somewhere between 75-150 ie, neither overly challenging nor too big a drag on State’s overall SOS and thus its RPI ranking. On the opposite end of where State currently schedules, you don’t want an OOC schedule so difficult that you end up on the NCAAT bubble. This is what happened to Indiana in 2005, going 5-6 against the #13 ranked OOC schedule. 10 conference wins and a first-round loss in the conference tournament left them with a 15-13 record and a trip to the NIT.
– Minimum of 3-4 teams with good enough programs that they would be expected to make the NCAAT more years than not. (The ACC/Big 10 Challenge would contribute one game to this “requirementâ€.)
– A preseason tournament that actually includes real, Division 1 teams every couple of years. (Note that the BCA Classic from last season does not qualify.)
– Enough home games that the Athletic Department’s budget isn’t compromised. (This is the one “requirement†that I feel sure will be met.)
– Experience says that many people will scream that they want at least one game against a real power program (Kansas, UConn, UCLA, etc). I agree that this would be nice, but it doesn’t really rank as one of my “must-havesâ€.
State’s OOC schedule in 2004 comes pretty close to what I would like to see most years. The 2004 season didn’t really have any 5-star OOC programs, but was reasonably respectable, had a good number of NCAAT teams, and minimized the number of truly horrible teams. This type of schedule would maximize benefit and minimize risk to any team ranging from the NCAAT bubble through a Final Four contender.
UPDATED INFO
Noah had an interesting question in the first comment to this entry that made me curious what type of OOC schedule the top two-thirds of the NCAAT field played. Stanford was the last at-large bid this year with an RPI ranking of #65. So here is the distribution of OOC SOS played by the conference champions ranked higher than #65 and all 34 of the teams to receive an at-large bid.
OOC |
Conf |
At-Large |
|
SOS |
Champs |
Bids |
Total |
1-50 |
5 |
14 |
19 |
51-100 |
3 |
8 |
11 |
101-150 |
3 |
6 |
9 |
151-200 |
1 |
4 |
5 |
201-250 |
1 |
0 |
1 |
251-300 |
0 |
0 |
0 |
301+ |
0 |
2 |
2 |
What this table tells me is that it shouldn’t be that hard to come up with an OOC schedule that is considerably better than we’ve seen for most of the past nine years.