Raymond’s highly scientific predictions for the 2013 NCAA men’s basketball tournament


Once again, it's time for Raymond to come up with an absurd, arbitrary criterion for filling out his NCAA bracket.

This year, I look at the number of fans of the basketball team's official Facebook page, or one tenth of the number of fans of the school's athletic department, whichever is greater. The fraction 1/10 is completely arbitrary, but that's what makes this algorithm highly scientific.

Notes on this year's algorithm:

  • Counting fans also includes people who hatewatch the team. I accept this, because if a lot of people hate a team, it's probably because they're "too good."
  • There are over a hundred thousand fans of the stub page for North Carolina Basketball. Too bad they don't count.
  • Others who don't count: Fans of unofficial pages. Now, you might think that this is an unscientific exclusion, becaus a fan of an unofficial page is still a fan. On the other hand, if an unofficial page out-draws the official page, then I blame the school for doing a bad job. Also: highly scientific.
  • Syracuse has not one but two spam pages for their basketball team. (You can tell it's a spam page because it has no meaningful content and was set up by a person selling SEO services.) Both of them have more fans than the official page, one of them by a factor of over 100×.
  • Some schools have a Facebook icon on their Athletics page, but it links to ESPN. I guess I know who their sponsor is.
  • Montana's Facebook icon links to the Facebook home page. Somebody wasn't paying attention in "Web design 101" class.

Once the field has been narrowed to four teams, the results are determined by a coin flip. (I should have done this when the field reduced to eight teams rather than four, but I messed up and am too lazy to fix it. That's what make this highly scientific.)

Update:

  • Correct predictions are in green.
  • Incorrect predictions are in red.
  • (!) marks upsets correctly predicted.
  • (*) marks upsets predicted but did not take place.
  • (x) marks actual upsets not predicted.

Opening Round Games

Middle Tennessee (2,509) Saint Mary's
(2,654)
Saint Mary's (2,654)
Liberty (3,281) Liberty
(3,281)
North Carolina A&T (4,324a)
Boise State (26,559a) Boise State
(2,655)
La Salle 1 (3,010a)
James Madison (247) James Madison
(247)
Long Island (755a)

Group 1

1 Louisville (45,835) Louisville
(45,835)
Louisville
(45,835)
Oregon
(50,525) (*)
Duke
(82,675)
16 Liberty (3,281)
8 Colorado State (1,915) Missouri
(35,712)
9 Missouri (357,124a)
5 Oklahoma State (3,428) Oregon
(50,525) (!)
Oregon
(50,525)
12 Oregon (505,253a)
4 Saint Louis (4,209a) New Mexico State
(690) (*)
13 New Mexico State (6,902a)
6 Memphis (64,967a) Memphis
(6,496)
Michigan State
(69,455)
Duke
(82,675)
11 Saint Mary's (2,654)
3 Michigan State (69,455) Michigan State
(69,455)
14 Valparaiso (3,664)
7 Creighton (8,175) Cincinnati
(12,660) (*)
Duke
(82,675)
10 Cincinnati (126,600a)
2 Duke (82,675) Duke
(82,675)
15 Albany (4,060a)

Group 2

1 Kansas (126,999) Kansas
(126,999)
Kansas
(126,999)
Kansas
(126,999) (x)
Kansas
(126,999)
16 Western Kentucky (7,513a)
8 North Carolina (686,843a) North Carolina
(68,684)
9 Villanova (33,889)
5 VCU (17,954a) Akron
(2,245) (*)
Michigan
(43,385)
12 Akron (2,245)
4 Michigan (433,854a) Michigan
(43,385)
13 South Dakota State (867)
6 UCLA (14,376) UCLA
(14,376) (x)
Florida
(108,604)
Florida
(108,604)
11 Minnesota (10,389)
3 Florida (1,086,046a) Florida
(108,604)
14 Northwestern State (2,299)
7 San Diego State (18,304a) Oklahoma
(39,465) (*)
Oklahoma
(39,465)
10 Oklahoma (394,654a)
2 Georgetown (28,997a) Georgetown
(2,899) (x)
15 Florida Gulf Coast (2,120)

Group 3

1 Indiana (132,015) Indiana
(132,015)
Indiana
(132,015)
Indiana
(132,015)
Indiana
(132,015)
16 James Madison (247)
8 NC State (10,409) NC State
(10,409)
9 Temple (538)
5 UNLV (5,822) UNLV
(5,822) (x)
Montana
(7,702)
12 California (1,001)
4 Syracuse (2,528) Montana
(7,702) (*)
13 Montana (77,020a)
6 Butler (87,481a) Butler
(8,748)
Marquette
(18,379)
Illinois
(77,222)
11 Bucknell (1,596)
3 Marquette (18,379) Marquette
(18,379)
14 Davidson (523)
7 Illinois (77,222) Illinois
(77,222)
Illinois
(77,222) (x)
10 Colorado (7,961)
2 Miami (11,061) Miami
(11,061)
15 Pacific (4,850a)

Group 4

1 Gonzaga (3,698) Gonzaga
(3,698)
Wichita State
(10,446) (!)
Wisconsin
(26,509)
Ohio State
(33,536)
16 Southern (3,884a)
8 Pittsburgh (86,187a) Wichita State
(10,446)
9 Wichita State (10,446)
5 Wisconsin (26,509) Wisconsin
(26,509) (x)
Wisconsin
(26,509)
12 Mississippi (3,198)
4 Kansas State (46,355a) Kansas State
(4,635) (x)
13 Boise State (26,559a)
6 Arizona (36,578a) Arizona
(3,657)
Arizona
(3,657)
Ohio State
(33,536)
11 Belmont (3,956a)
3 New Mexico (97) Harvard
(778) (!)
14 Harvard (7,781a)
7 Notre Dame (254,683a) Notre Dame
(25,468) (x)
Ohio State
(33,536)
10 Iowa State (13,968)
2 Ohio State (33,536) Ohio State
(33,536)
15 Iona (2,549a)

Finals

Duke Indiana
Indiana
Indiana
Kansas Ohio State
Ohio State
Comments (12)
  1. Anonymous says:

    This bracket seems more likely to happen than most of the previous brackets you've come up with.

  2. What does the "a" mean?

  3. Anonymous says:

    @Maurits:  I assume it means fans of the entire athletic department, which he then divides by 10 before making the comparison according to his formula.

  4. Anonymous says:

    This not fair, for the big University such as Ohio State, they have more than 100,000 student. The St. Mary can never compete with that!

  5. Anonymous says:

    @Jim: And Division I basketball is fair? Can St. Mary's recruit the same people as Ohio State, or Duke? When taken to the extreme, if sports was "fair", we may as well just flip a coin. It's the unfairness, and the stories that come out of how individuals and teams react to that unfairness, that makes sports interesting.

    Raymond, any thoughts that instead of using the absolute value of the "scientific" process and picking a winner, you could use the numbers to weigh the outcome of a percentile die roll? That way Gonzaga at #1 would at least have a chance to win over #9 Witchita State. Or would that be too "unscientific" (that is to say, less fun.)

    [I already used that highly scientific technique in 2008. It took a huge amount of time to format the results of manual dice-rolling, so I'm not inclined to try it again. -Raymond]
  6. Anonymous says:

    Right, they have more teams this year, even Harvard??

  7. Anonymous says:

    Unless I'm missing something, it looks like you have Ohio St as an upset over Kansas in the final four since they're at 33,536 and Kansas is at 126,999.

    While Indiana would still win it all, Kansas would at least make it to back-to-back title games.

    [The final games are decided by a coin flip. -Raymond]
  8. Anonymous says:

    Not a bad choice. Too bad there's no way to separate basketball fans from football fans. Oregon for instance has tons more fans since their football team has been good lately.

  9. Anonymous says:

    I don't know, having a Facebook icon actually go to Facebook's home page isn't the worst thing ever. Imagine if, instead, it went to some other service that started with an f or even some service that didn't even start with f (like MySpace)!

  10. Programmerman says:

    I like that you called the groups by number instead of their geographic regional names to avoid the inevitable question about why, for instance, a west coast team plays in the east region.

  11. Anonymous says:

    It looks like there is a slight problem with your final four pairings.

    You have Group 1 paired with Group 3, and Group 2 paired with Group 4.

    It should be Group 1 paired with Group 4, and Group 2 paired with Group 3.

    Time to re-flip!

  12. Anonymous says:

    @Ian: If it's possible for Indiana to play Ohio State in the final, does it really matter who the coin says they'll beat? It doesn't meaningfully affect the outcome when the only goal is to select winners. Seems like an edit is better than a reflip.

Comments are closed.