Why are we still organizing teams by the RPI?
If we are going to evaluate teams based on the quality of their wins and losses, I sure wish the NCAA committee’s team sheets were organized differently. I wish the NCAA committee would rank quality wins based on one of the major systems (Kenpom, BPI, Sagarin) instead of based on the RPI. For example, as of Tuesday evening, the following teams are in the Kenpom Top 50, but not the RPI Top 50:
|
KenPom |
RPI |
|
|
Utah |
33 |
82 |
|
St. John's |
37 |
58 |
|
Florida St. |
39 |
57 |
|
Louisiana Tech |
44 |
71 |
|
Maryland |
46 |
65 |
|
Georgetown |
50 |
54 |
Utah is almost certainly a quality team this year. But because they played such a weak non-conference schedule, opponents do not get credit for beating them. Does that make any sense? Conversely, people are getting a lot of credit for beating these teams:
|
KenPom |
RPI |
|
|
Dayton |
51 |
39 |
|
North Dakota St. |
56 |
37 |
|
Colorado |
57 |
32 |
|
Saint Joseph's |
60 |
42 |
|
Southern Miss |
61 |
35 |
|
Toledo |
106 |
36 |
Kansas might be getting a little too much credit for that win against Toledo.
As I already said, the fact that these teams are ranked too low or high isn’t the issue. The issue is that other teams are not getting enough credit for playing and beating teams like Utah. Here is how many games each team is penalized because the committee organizes by the RPI instead of the Kenpom rankings. For example, if we used the Pomeroy rankings, Providence would get credit for having played five more Top 50 teams!
|
Diff. Top 50 Games |
|
|
Providence |
-5 |
|
Virginia |
-4 |
|
Xavier |
-4 |
|
Marquette |
-4 |
|
Villanova |
-3 |
|
Creighton |
-3 |
|
Syracuse |
-3 |
|
Pittsburgh |
-3 |
|
Clemson |
-3 |
|
Miami FL |
-3 |
|
…. |
|
|
Dayton |
2 |
|
UMass |
2 |
|
Utah |
2 |
|
Baylor |
2 |
|
St. Joe's |
2 |
|
St. Louis |
3 |
Want to know why there are so few ACC and Big East teams in people’s projected brackets? Maybe the committee has the wrong way to rank quality wins and losses.
Of course games against the Top 50 are one thing, but wins against the Top 50 are another thing. The following table shows the change in Top 50 wins. For example, Virginia has three less Top 50 wins because we use the RPI instead of the Kenpom.com rankings. Do you think that might have some bearing on whether Virginia gets a 1-seed or 2-seed in the NCAA tournament?
|
Diff. Top 50 Wins |
|
|
Virginia |
-3 |
|
Xavier |
-3 |
|
Villanova |
-3 |
|
Syracuse |
-3 |
|
Providence |
-2 |
|
Marquette |
-2 |
|
Pittsburgh |
-2 |
|
Clemson |
-2 |
|
Duke |
-2 |
|
North Carolina |
-2 |
|
Oregon |
-2 |
|
NC State |
-2 |
|
…. |
|
|
Baylor |
2 |
|
St. Joe's |
2 |
|
St. Louis |
2 |
Of course if you stare at this too long your brain starts to hurt. Why does a team get more credit for beating team #50 than team #51? What should we think when the Pomeroy and Sagarin ratings disagree? Truly, when you argue about NCAA tournament selection, you can go on forever. The good (or bad) news is that this is all constantly in flux. Georgetown may fall out of the Pomeroy Top 50 tomorrow, and then Big East fans will have less to complain about.
Speaking of wins against the RPI Top 50, let’s take a quick peak at where teams stand heading into the conference tournaments.
CR = Conference Record
RN = Road/Neutral Record
Top 50 = Record against the RPI Top 50
51-100 = Record against teams 51-100
BL = Bad Losses, Losses to Teams ranked 101+
American Conference Tournament Printable Bracket
|
AAC |
CR |
RN |
NCSOS |
Top 50 |
51-100 |
BL |
|
Cincinnati |
15-3 |
8-4 |
98 |
7-5 |
2-0 |
0 |
|
Louisville |
15-3 |
10-3 |
150 |
6-5 |
1-0 |
0 |
|
Connecticut |
12-6 |
9-4 |
72 |
5-6 |
4-0 |
1 |
|
Memphis |
12-6 |
8-6 |
55 |
5-7 |
1-0 |
1 |
|
SMU |
12-6 |
8-7 |
289 |
4-5 |
0-1 |
2 |
Everyone’s numbers look good here, but remember that the committee often likes to pick on a team with a poor non-conference schedule. SMU will be sweating on selection Sunday if they lose to Houston in the conference tournament opener.
A10 Tournament Printable Bracket
|
A10 |
CR |
RN |
NCSOS |
Top 50 |
51-100 |
BL |
|
Saint Louis |
13-3 |
12-2 |
105 |
5-4 |
3-0 |
1 |
|
VCU |
12-4 |
9-7 |
70 |
4-4 |
5-2 |
1 |
|
George Washington |
11-5 |
9-6 |
118 |
3-5 |
6-2 |
0 |
|
Saint Joseph's |
11-5 |
10-5 |
135 |
4-5 |
2-3 |
1 |
|
Massachusetts |
10-6 |
12-5 |
26 |
6-3 |
6-2 |
2 |
|
Dayton |
10-6 |
9-5 |
99 |
4-5 |
5-1 |
3 |
The consensus seems to be that all these teams are in good shape too, and looking at these numbers, I can see that.
ACC Tournament Printable Bracket
|
ACC |
CR |
RN |
NCSOS |
Top 50 |
51-100 |
BL |
|
Virginia |
16-2 |
10-4 |
34 |
4-4 |
6-2 |
0 |
|
Syracuse |
14-4 |
11-2 |
107 |
6-2 |
9-0 |
2 |
|
Duke |
13-5 |
7-7 |
17 |
6-4 |
4-1 |
2 |
|
North Carolina |
13-5 |
8-5 |
41 |
5-4 |
6-1 |
3 |
|
Pittsburgh |
11-7 |
10-3 |
222 |
1-6 |
5-2 |
0 |
|
Clemson |
10-8 |
6-8 |
267 |
1-6 |
3-2 |
3 |
|
Florida State |
9-9 |
9-7 |
90 |
3-7 |
2-4 |
1 |
|
Maryland |
9-9 |
5-10 |
18 |
1-9 |
3-5 |
0 |
|
NC State |
9-9 |
6-6 |
114 |
2-7 |
3-2 |
3 |
Given their lack of quality wins, Pittsburgh really cannot afford to lose to Wake Forest/Notre Dame in their ACC tournament opener.
I am looking forward to listening to some TV commentators make the argument for Florida St. or Clemson this week. But compare them to those teams I have listed in the A10 and American conference. The quality wins just are not there.
Big East Tournament Printable Bracket
|
Big East |
CR |
RN |
NCSOS |
Top 50 |
51-100 |
BL |
|
Villanova |
16-2 |
13-2 |
49 |
5-3 |
11-0 |
0 |
|
Creighton |
14-4 |
8-6 |
36 |
6-3 |
7-3 |
0 |
|
Xavier |
10-8 |
5-9 |
79 |
3-5 |
5-3 |
3 |
|
Providence |
10-8 |
7-8 |
177 |
2-6 |
4-4 |
1 |
|
Saint John's |
10-8 |
6-7 |
133 |
1-7 |
5-2 |
2 |
|
Marquette |
9-9 |
5-10 |
156 |
2-10 |
3-3 |
1 |
|
Georgetown |
8-10 |
5-10 |
24 |
5-6 |
2-4 |
3 |
Georgetown is going to be one of the most discussed teams in the country over the next five days. How do you weight those bad losses against the Top 50 wins? Is the poor record against teams 51-100 the tie-breaker?
Big Ten Tournament Printable Bracket
|
Big Ten |
CR |
RN |
NCSOS |
Top 50 |
51-100 |
BL |
|
Wisconsin |
12-6 |
11-3 |
9 |
7-3 |
9-2 |
1 |
|
Michigan |
15-3 |
10-5 |
95 |
8-5 |
5-1 |
1 |
|
Ohio State |
10-8 |
6-5 |
59 |
5-4 |
7-2 |
2 |
|
Michigan State |
12-6 |
11-4 |
81 |
6-6 |
4-2 |
0 |
|
Nebraska |
11-7 |
4-10 |
96 |
3-7 |
5-1 |
3 |
|
Iowa |
9-9 |
7-7 |
174 |
4-8 |
2-3 |
0 |
|
Minnesota |
8-10 |
3-9 |
44 |
3-8 |
3-2 |
2 |
With three bad losses and a 4-10 road/neutral court record, Nebraska might not have reached lock status yet. In that sense, I wonder if the “bye” in the Big Ten tournament works to their disadvantage. They will probably have to beat Ohio St. to get a neutral-court win.
Big 12 Tournament Printable Bracket
|
Big 12 |
CR |
RN |
NCSOS |
Top 50 |
51-100 |
BL |
|
Kansas |
14-4 |
8-7 |
1 |
12-7 |
4-1 |
0 |
|
Oklahoma |
12-6 |
9-5 |
77 |
8-5 |
3-2 |
1 |
|
Iowa State |
11-7 |
8-6 |
76 |
9-6 |
2-1 |
0 |
|
Texas |
11-7 |
6-7 |
164 |
7-8 |
4-0 |
1 |
|
Kansas State |
10-8 |
5-9 |
143 |
7-7 |
2-2 |
2 |
|
Baylor |
9-9 |
7-6 |
37 |
8-8 |
1-1 |
1 |
|
West Virginia |
9-9 |
6-9 |
215 |
5-11 |
0-1 |
2 |
|
Oklahoma State |
8-10 |
7-8 |
102 |
5-10 |
3-0 |
1 |
Oklahoma St.’s potential seed remains a huge mystery to me. They have some great wins, but that profile isn’t elite at this point. Sure, they might not get dinged as much for the three losses without Marcus Smart, but why not? He was suspended for conduct on the court. Shouldn’t that count against them? And isn’t this a surprise? They could face Kansas in the Big 12 quarterfinals.
I’m looking forward to someone unexpected (Texas, Kansas St.) winning the whole thing and then suddenly showing up as a 2 or 3 seed on selection Sunday while everyone stammers that the committee must have really liked what they did in the Big 12 tournament.
Pac-12 Tournament (Printable Bracket Link in Corner, must make Landscape)
|
Pac-12 |
CR |
RN |
NCSOS |
Top 50 |
51-100 |
BL |
|
Arizona |
15-3 |
10-3 |
38 |
10-2 |
7-1 |
0 |
|
Oregon |
10-8 |
7-5 |
51 |
4-5 |
7-3 |
0 |
|
UCLA |
12-6 |
7-6 |
113 |
5-4 |
6-3 |
1 |
|
Colorado |
10-8 |
5-8 |
46 |
5-7 |
5-3 |
0 |
|
Arizona State |
10-8 |
5-9 |
226 |
4-6 |
5-3 |
1 |
|
Stanford |
10-8 |
7-7 |
65 |
4-8 |
3-3 |
0 |
|
California |
10-8 |
6-8 |
89 |
4-9 |
4-2 |
1 |
|
Utah |
9-9 |
2-8 |
341 |
4-7 |
3-2 |
1 |
I mentioned Utah in the opener. If the margin-of-victory numbers like them, you may wonder why they are not realistically on the bubble. Well, that terrible NCSOS is reason #1. And that terrible road record is reason #2.
SEC Tournament Printable Bracket
|
SEC |
CR |
RN |
NCSOS |
Top 50 |
51-100 |
BL |
|
Florida |
18-0 |
12-2 |
16 |
6-2 |
8-0 |
0 |
|
Kentucky |
12-6 |
6-7 |
10 |
2-5 |
10-3 |
1 |
|
Georgia |
12-6 |
4-10 |
130 |
0-6 |
6-1 |
5 |
|
Tennessee |
11-7 |
6-8 |
43 |
2-5 |
5-2 |
4 |
|
Arkansas |
10-8 |
4-8 |
198 |
3-3 |
5-5 |
2 |
|
Missouri |
9-9 |
6-8 |
129 |
2-3 |
5-5 |
2 |
|
LSU |
9-9 |
5-9 |
152 |
2-5 |
2-4 |
3 |
|
Mississippi |
9-9 |
6-8 |
103 |
0-7 |
3-3 |
3 |
It is sort of amazing that a team can have 0 top 50 wins and 5 bad losses and still finish tied for second in the SEC, but that’s exactly what Georgia did this year.
Kentucky cannot be happy that they might face LSU in the quarterfinals. LSU has the size and strength to really match up with the Cats and has played them tough twice this year.
