- Joined
- Sep 14, 2011
- Messages
- 2,676
- Reaction Score
- 6,257
How well did the Selection Committee do in choosing and seeding teams for this year's tournament? Were the seedings fair? Are the regions well balanced? Which region, if any, is the toughest? Which the easiest? Which conferences were overvalued by the committee? Which undervalued?
Those were the questions I sought to answer by undertaking an analysis of the brackets as selected by the committee. In order to grade them, I needed a standard by which to evaluate the selections. As with other evaluations (primarily Creme's brackets) I've done here, I've chosen the AP Poll as the standard. I prefer a human based standard over a pure calculation (RPI, SOS) because it can accommodate more variables, e.g. injuries, albeit with less precision than more mechanical measures. The AP attempts to make up for the lack of precision inherent in individuals by sampling a group. I prefer the AP over the Coaches Poll because of the built in bias, sometimes intentional, a coach attaches to his/her choices.
The AP only ranks 25 teams so it is incomplete as a standard. It is what it is, however, and what I've chosen despite this limitation.
Each team in the poll receives a rank from one to twenty-five. I've applied a simple transformation of the AP rank so that larger numbers mean "better." I will assign 25 to the highest ranked team rather than 1 as the AP does. Thus the top four teams are ranked 25, 24, 23, and 22 rather than 1, 2, 3, and four. The values I used in my calculations are: UConn (25); Notre Dame (24); Tennessee (23); Louisville (22); Baylor (21); Stanford (20); West Virginia (19); South Carolina (18); Duke (17); Kentucky (16); Maryland (15); North Carolina (14); Nebraska (13); Penn State (12); Texas A&M (11); NC State (10); Purdue (9); Gonzaga (8); Iowa (7); Michigan State (6); Oklahoma State (5); Middle Tennessee (4); DePaul (3); California (2); and Southern Cal (1).
The ranked teams were distributed among the regional brackets as follows:
Lincoln -- UConn (25), Duke (17), Texas A&M (11), Nebraska (13); NC State (10), Gonzaga (8), and DePaul (3)
South Bend -- Notre Dame (24), Baylor (21), Kentucky (16), Purdue (9), Oklahoma State (5), and California (2)
Louisville -- Tennessee (23), West Virginia (19), Louisville (22), Maryland (15), Iowa (7) and Southern Cal (1)
Stanford -- South Carolina (18), Stanford (20), Penn State (12), North Carolina (14), Michigan State (6), and Middle Tennessee (4)
Reasonably equitable distribution of ranked teams across regions. Six teams in each region. Mathematics says that one region would necessarily have more teams than the other three because 4 doesn't divide evenly into 25. But why add the extra team to UConn's bracket? Isn't the #1 overall seed entitled to the easiest path to the final four?
The difficulty of each region (as measured by the strength of the teams in it is:
Lincoln -- 88 (25+17+11+13+10+8)
South Bend -- 83 (24+21+16+9+5+2)
Louisville -- 83 (23+19+22+15+7+1)
Stanford -- 63 (18+20+12+14+6+4)
Not particularly equitable considering the overall #1 team has the toughest region while the weakest #1 seed has by far the easiest placement. Note: I left DePaul's value out of the Lincoln region so that regional totals wouldn't be muddied by calculations based on unequal data points, otherwise UConn's region would have been even tougher.
How were the teams in each region valued by the committee? To determine this, I assigned each team to a category: overvalued; undervalued; or equitably valued. Teams with an AP ranking were assigned to seed lines based on their AP rank. Teams ranked 1 through 4 were given #1 seeds. Teams ranked 5 though 8 were assigned a #2 seed, and so forth.
Teams in the Lincoln Region were slightly undervalued with Duke and Texas A&M overvalued, NC State, Gonzaga, and DePaul undervalued, and UConn and Nebraska equitably valued.
Teams in the South Bend Region were slightly overvalued with Purdue and Oklahoma State overvalued, California undervalued, and Notre Dame, Baylor, and Kentucky equitably valued.
Teams in the Louisville Region were significantly undervalued with nobody overvalued, Louisville (grossly), Maryland, Southern Cal, and Iowa undervalued, and Tennessee and West Virginia equitably valued.
Teams in the Stanford Region were overvalued with South Carolina and Penn State overvalued, North Carolina and Middle Tennessee undervalued, and Stanford and Michigan State equitably valued. Note: even though the number of overvalued teams equaled the number of undervalued teams in this bracket, the placement of South Carolina as a #1 seed simply cannot be ignored.
Conference Valuations:
ACC -- Undervalued (one team overvalued, 3 undervalued, and 1 equitably valued
B1G -- Overvalued (2 over, 1 under, 2 equitable)
B12 -- Overvalued (1 over, none under, 2 equitable)
Pac12 -- Undervalued (none over, 2 under, 1 equitable)
SEC -- Overvalued (2 over, none under, 2 equitable)
Non-P5 -- Undervalued (none over, 4 under, 1 equitable)
Over all grade for the Selection Committee B-
The solidly inequitable Stanford bracket when compared to the other three simply cannot be ignored. If you want to tell me thee committee was hampered by the rules, well they (or someone in the next office) write the rules. They also deserve some lumps for the bizarre South Carolina placement. Not only were the Gamecocks overvalued but their overvalue was doubly rewarded by sending them to the easiest region.
Those were the questions I sought to answer by undertaking an analysis of the brackets as selected by the committee. In order to grade them, I needed a standard by which to evaluate the selections. As with other evaluations (primarily Creme's brackets) I've done here, I've chosen the AP Poll as the standard. I prefer a human based standard over a pure calculation (RPI, SOS) because it can accommodate more variables, e.g. injuries, albeit with less precision than more mechanical measures. The AP attempts to make up for the lack of precision inherent in individuals by sampling a group. I prefer the AP over the Coaches Poll because of the built in bias, sometimes intentional, a coach attaches to his/her choices.
The AP only ranks 25 teams so it is incomplete as a standard. It is what it is, however, and what I've chosen despite this limitation.
Each team in the poll receives a rank from one to twenty-five. I've applied a simple transformation of the AP rank so that larger numbers mean "better." I will assign 25 to the highest ranked team rather than 1 as the AP does. Thus the top four teams are ranked 25, 24, 23, and 22 rather than 1, 2, 3, and four. The values I used in my calculations are: UConn (25); Notre Dame (24); Tennessee (23); Louisville (22); Baylor (21); Stanford (20); West Virginia (19); South Carolina (18); Duke (17); Kentucky (16); Maryland (15); North Carolina (14); Nebraska (13); Penn State (12); Texas A&M (11); NC State (10); Purdue (9); Gonzaga (8); Iowa (7); Michigan State (6); Oklahoma State (5); Middle Tennessee (4); DePaul (3); California (2); and Southern Cal (1).
The ranked teams were distributed among the regional brackets as follows:
Lincoln -- UConn (25), Duke (17), Texas A&M (11), Nebraska (13); NC State (10), Gonzaga (8), and DePaul (3)
South Bend -- Notre Dame (24), Baylor (21), Kentucky (16), Purdue (9), Oklahoma State (5), and California (2)
Louisville -- Tennessee (23), West Virginia (19), Louisville (22), Maryland (15), Iowa (7) and Southern Cal (1)
Stanford -- South Carolina (18), Stanford (20), Penn State (12), North Carolina (14), Michigan State (6), and Middle Tennessee (4)
Reasonably equitable distribution of ranked teams across regions. Six teams in each region. Mathematics says that one region would necessarily have more teams than the other three because 4 doesn't divide evenly into 25. But why add the extra team to UConn's bracket? Isn't the #1 overall seed entitled to the easiest path to the final four?
The difficulty of each region (as measured by the strength of the teams in it is:
Lincoln -- 88 (25+17+11+13+10+8)
South Bend -- 83 (24+21+16+9+5+2)
Louisville -- 83 (23+19+22+15+7+1)
Stanford -- 63 (18+20+12+14+6+4)
Not particularly equitable considering the overall #1 team has the toughest region while the weakest #1 seed has by far the easiest placement. Note: I left DePaul's value out of the Lincoln region so that regional totals wouldn't be muddied by calculations based on unequal data points, otherwise UConn's region would have been even tougher.
How were the teams in each region valued by the committee? To determine this, I assigned each team to a category: overvalued; undervalued; or equitably valued. Teams with an AP ranking were assigned to seed lines based on their AP rank. Teams ranked 1 through 4 were given #1 seeds. Teams ranked 5 though 8 were assigned a #2 seed, and so forth.
Teams in the Lincoln Region were slightly undervalued with Duke and Texas A&M overvalued, NC State, Gonzaga, and DePaul undervalued, and UConn and Nebraska equitably valued.
Teams in the South Bend Region were slightly overvalued with Purdue and Oklahoma State overvalued, California undervalued, and Notre Dame, Baylor, and Kentucky equitably valued.
Teams in the Louisville Region were significantly undervalued with nobody overvalued, Louisville (grossly), Maryland, Southern Cal, and Iowa undervalued, and Tennessee and West Virginia equitably valued.
Teams in the Stanford Region were overvalued with South Carolina and Penn State overvalued, North Carolina and Middle Tennessee undervalued, and Stanford and Michigan State equitably valued. Note: even though the number of overvalued teams equaled the number of undervalued teams in this bracket, the placement of South Carolina as a #1 seed simply cannot be ignored.
Conference Valuations:
ACC -- Undervalued (one team overvalued, 3 undervalued, and 1 equitably valued
B1G -- Overvalued (2 over, 1 under, 2 equitable)
B12 -- Overvalued (1 over, none under, 2 equitable)
Pac12 -- Undervalued (none over, 2 under, 1 equitable)
SEC -- Overvalued (2 over, none under, 2 equitable)
Non-P5 -- Undervalued (none over, 4 under, 1 equitable)
Over all grade for the Selection Committee B-
The solidly inequitable Stanford bracket when compared to the other three simply cannot be ignored. If you want to tell me thee committee was hampered by the rules, well they (or someone in the next office) write the rules. They also deserve some lumps for the bizarre South Carolina placement. Not only were the Gamecocks overvalued but their overvalue was doubly rewarded by sending them to the easiest region.