Nerd Champions Seed Math

[jleves]

Based on the thread about total seeds beaten, I decided to do some extra math and analysis. First of all, using total seeds beaten to determine the hardest paths is worthless. A 1 or 2 seed accumulates more points just getting through the first weekend than the 8th seed 85 Nova team beat for the entire tournament. ie. Of course higher seeds have a tougher path on paper. So I looked at a couple other metrics.

All of these are based on the champions since 1985 when the tournament went to 64 teams.

The first is the difference between the lowest number you could face based on seed vs what you actually had to play. The champions for the 38 years averaged playing 10 points over the minimum they could face in a chalk bracket. ie if a 1 seed plays a 9 instead of an 8 in the second round, that's +1, a 3 faces an 11 instead of a 6 in the second round, that's plus 5, etc.

Next I took the 4 hardest games to see how things stack up as a 1 seed could face a 1, a 1, a 2 and a 4. Then just cause it was easy, how many top 4 seeds did they beat.

Based on the results, there is a pretty good argument that UConn has had some above average easy paths. Only the 2014 team faced the average of + 10 (also the only UConn team to face 4 top 4 seeds). The other four were all over the average: 1999 +13, 2004 +12, 2011 +14 and 2023 +15. 2023 is one of only two teams that only faced one top 4 seed en route to the trophy.

Nova and UNC have had the 4 hardest paths with a +1 and +2 for each.

I'll put the complete chart in the next post, but here is the lowest possible score for each seed.

1: 32
2: 28
3: 25
4: 23
5: 21
6: 19
7: 18
8: 18
9: 17
10: 15
11: 14
12: 14
13: 14
14: 14
15: 15
16: 17

 

[jleves]

Year	Winner	Seed	Total Beaten	Rank	Over Min	Best 4	Best 4 Avg	1-4 Wins
1985	Villanova	8	20	1	2	1,1,2,2	1.5	4
1986	Louisville	2	45	T31	17	1,3,7,8	4.75	2
1987	Indiana	1	43	T28	11	1,2,5,8	4	2
1988	Kansas	6	39	T20	20	1,2,4,7	3.5	3
1989	Michigan	3	36	T8	11	1,2,3,5	2.75	3
1990	UNLV	1	54	38	22	3,4,8,11	6.5	2
1991	Duke	2	41	26	13	1,3,4,7	3.75	3
1992	Duke	1	39	T20	7	2,2,4,6	3.5	3
1993	North Carolina	1	33	4	1	1,2,2,4	2.25	4
1994	Arkansas	1	44	30	12	2,2,3,9	4	3
1995	UCLA	1	37	T13	5	2,2,4,5	3.25	3
1996	Kentucky	1	36	T8	4	1,2,4,4	2.75	4
1997	Arizona	4	38	T16	15	1,1,1,10	3.25	3
1998	Kentucky	2	38	T16	10	1,3,3,6	3.25	3
1999	UConn	1	45	T31	13	1,4,5,9	4.75	2
2000	Michigan State	1	43	T28	11	2,4,5,8	4.75	2
2001	Duke	1	40	T24	8	2,3,4,6	3.75	3
2002	Maryland	1	36	T8	4	1,2,4,5	3	3
2003	Syracuse	3	34	T5	9	1,1,2,6	2.5	3
2004	UConn	2	40	T24	12	1,3,6,7	4.25	2
2005	North Carolina	1	42	27	10	1,5,5,6	4.25	1
2006	Florida	3	46	33	21	1,2,7,11	5.25	2
2007	Florida	1	36	T8	4	1,2,3,5	2.75	3
2008	Kansas	1	48	T34	16	1,1,8,10	5	2
2009	North Carolina	1	35	7	3	2,2,3,4	2.75	4
2010	Duke	1	38	T16	6	2,3,4,5	3.5	3
2011	UConn	3	39	T20	14	2,4,5,6	4.25	2
2012	Kentucky	1	37	T13	5	2,3,4,4	3.25	4
2013	Louisville*	1	51	37	19	2,4,8,9	5.75	2
2014	UConn	7	28	2	10	1,2,3,4	2.5	4
2015	Duke	1	39	T20	7	1,2,5,7	3.75	2
2016	Villanova	2	29	3	1	1,1,2,3	1.75	4
2017	North Carolina	1	34	T5	2	1,2,3,4	2.5	4
2018	Villanova	1	37	T13	5	1,3,3,5	3	3
2019	Virginia	1	48	T34	16	3,3,5,9	5	2
2020	COVID-19	0	598	39	0	1,1,1,1	1	16
2021	Baylor	1	36	T8	4	1,2,3,5	2.75	3
2022	Kansas	1	49	36	17	2,4,8,9	5.75	2
2023	UConn	4	38	T16	15	3,5,5,5	4.5	1
AVG					10		3.7	2.75

*Vacated but included for historical accuracy

 

[auror]

I thought this was about the boneyard nerd bracket. You were clearly underseeded.

 

[UConNick]

COVID really mauled everyone in 2020. Shame they fell victim to the defending champions’ curse and couldn’t get out of the first weekend in 2021, only winning one game (poor VCU).

On a serious note, it will be interesting to see if we’re able to buck this trend next year - really though KU was good enough.

 

[Chin Diesel]

Year	Winner	Seed	Total Beaten	Rank	Over Min	Best 4	Best 4 Avg	1-4 Wins
1985	Villanova	8	20	1	2	1,1,2,2	1.5	4
1986	Louisville	2	45	T31	17	1,3,7,8	4.75	2
1987	Indiana	1	43	T28	11	1,2,5,8	4	2
1988	Kansas	6	39	T20	20	1,2,4,7	3.5	3
1989	Michigan	3	36	T8	11	1,2,3,5	2.75	3
1990	UNLV	1	54	38	22	3,4,8,11	6.5	2
1991	Duke	2	41	26	13	1,3,4,7	3.75	3
1992	Duke	1	39	T20	7	2,2,4,6	3.5	3
1993	North Carolina	1	33	4	1	1,2,2,4	2.25	4
1994	Arkansas	1	44	30	12	2,2,3,9	4	3
1995	UCLA	1	37	T13	5	2,2,4,5	3.25	3
1996	Kentucky	1	36	T8	4	1,2,4,4	2.75	4
1997	Arizona	4	38	T16	15	1,1,1,10	3.25	3
1998	Kentucky	2	38	T16	10	1,3,3,6	3.25	3
1999	UConn	1	45	T31	13	1,4,5,9	4.75	2
2000	Michigan State	1	43	T28	11	2,4,5,8	4.75	2
2001	Duke	1	40	T24	8	2,3,4,6	3.75	3
2002	Maryland	1	36	T8	4	1,2,4,5	3	3
2003	Syracuse	3	34	T5	9	1,1,2,6	2.5	3
2004	UConn	2	40	T24	12	1,3,6,7	4.25	2
2005	North Carolina	1	42	27	10	1,5,5,6	4.25	1
2006	Florida	3	46	33	21	1,2,7,11	5.25	2
2007	Florida	1	36	T8	4	1,2,3,5	2.75	3
2008	Kansas	1	48	T34	16	1,1,8,10	5	2
2009	North Carolina	1	35	7	3	2,2,3,4	2.75	4
2010	Duke	1	38	T16	6	2,3,4,5	3.5	3
2011	UConn	3	39	T20	14	2,4,5,6	4.25	2
2012	Kentucky	1	37	T13	5	2,3,4,4	3.25	4
2013	Louisville*	1	51	37	19	2,4,8,9	5.75	2
2014	UConn	7	28	2	10	1,2,3,4	2.5	4
2015	Duke	1	39	T20	7	1,2,5,7	3.75	2
2016	Villanova	2	29	3	1	1,1,2,3	1.75	4
2017	North Carolina	1	34	T5	2	1,2,3,4	2.5	4
2018	Villanova	1	37	T13	5	1,3,3,5	3	3
2019	Virginia	1	48	T34	16	3,3,5,9	5	2
2020	COVID-19	0	598	39	0	1,1,1,1	1	16
2021	Baylor	1	36	T8	4	1,2,3,5	2.75	3
2022	Kansas	1	49	36	17	2,4,8,9	5.75	2
2023	UConn	4	38	T16	15	3,5,5,5	4.5	1
AVG					10		3.7	2.75

*Vacated but included for historical accuracy

This is one of the best starting points you could come up with to determine difficulty of path to the championship. Perfect baseline of year tourney expanded to 64, includes total beaten as compared to expected beaten for overall difficulty and then breaks out performance against the highest seeds.

Well done.

 

[SubbaBub]

Only one team each year wins 6 games. Start there.

 

[jleves]

I thought this was about the boneyard nerd bracket. You were clearly underseeded.

I didn't even make the selection. For the record, that started as listing all the seed beaten totals in response to the other thread. Then it morphed....