NCAA College Men's Basketball:
Prediction History

Logit Success Record

Games Played on	Number of Games	Successful Predictions	Success % all games	Wins at home %	Predicted from games thru
03-29	1	0	0%	0%	03-25
03-27	2	2	100%	0%	03-25
03-25	2	1	50%	0%	03-23
03-23	2	0	0%	0%	03-21
03-21	2	2	100%	0%	03-20
03-20	2	2	100%	0%	03-18
03-19	4	3	75%	0%	03-18
03-18	6	3	50%	100%	03-17
03-17	2	1	50%	50%	03-16
03-16	3	1	33%	33%	03-15
03-15	5	4	80%	80%	03-14
03-14	8	3	38%	100%	03-13
03-13	8	7	88%	0%	03-11
03-12	16	11	69%	100%	03-11
03-11	16	10	62%	0%	03-10
03-10	16	13	81%	81%	03-07
03-07	4	4	100%	0%	03-05
03-06	23	13	57%	0%	03-05
03-05	34	26	76%	83%	03-04
03-04	41	32	78%	58%	03-03
03-03	22	15	68%	0%	03-02
03-02	11	8	73%	0%	03-01
03-01	7	6	86%	0%	02-28
02-28	32	21	66%	81%	02-27
02-27	110	78	71%	71%	02-25
02-26	25	18	72%	68%	02-25
02-25	49	33	67%	63%	02-24
02-24	31	20	65%	68%	02-23
02-23	14	10	71%	71%	02-22
02-22	19	15	79%	63%	02-21
02-21	19	16	84%	74%	02-20
02-20	126	85	67%	65%	02-18
02-19	10	8	80%	50%	02-18
02-18	60	42	70%	55%	02-16
02-17	44	34	77%	70%	02-16
02-16	19	18	95%	63%	02-15
02-15	33	24	73%	48%	02-14
02-14	17	11	65%	53%	02-13
02-13	128	91	71%	69%	02-11
02-12	10	8	80%	80%	02-11
02-11	65	41	63%	69%	02-10
02-10	49	38	78%	61%	02-09
02-09	10	8	80%	50%	02-08
02-08	37	26	70%	68%	02-07
02-07	16	12	75%	69%	02-06
02-06	130	77	59%	52%	02-04
02-05	7	6	86%	43%	02-04
02-04	60	41	68%	68%	02-03
02-03	45	33	73%	62%	02-02
02-02	16	11	69%	62%	02-01
02-01	39	25	64%	56%	01-30
01-31	17	13	76%	88%	01-30
01-30	132	106	80%	63%	01-28
01-29	7	7	100%	43%	01-28
01-28	62	44	71%	66%	01-27
01-27	49	36	73%	70%	01-26
01-26	7	5	71%	43%	01-25
01-25	37	32	86%	72%	01-24
01-24	17	12	71%	71%	01-23
01-23	131	90	69%	63%	01-21
01-22	7	3	43%	29%	01-21
01-21	60	41	68%	58%	01-20
01-20	41	29	71%	56%	01-19
01-19	17	13	76%	59%	01-18
01-18	42	33	79%	69%	01-17
01-17	13	10	77%	62%	01-16
01-16	129	82	64%	61%	01-13
01-15	8	7	88%	88%	01-13
01-14	55	38	69%	64%	01-13
01-13	53	45	85%	62%	01-12
01-12	21	13	62%	43%	01-10
01-11	27	19	70%	59%	01-10
01-10	19	14	74%	68%	01-09
01-09	123	88	72%	67%	01-07
01-08	12	7	58%	42%	01-07
01-07	59	44	75%	66%	01-05
01-06	41	31	76%	59%	01-05
01-05	18	9	50%	82%	01-03
01-04	67	45	67%	64%	01-03
01-03	16	12	75%	73%	01-01
01-02	113	82	73%	65%	01-01
12-31	7	6	86%	86%	12-30
12-30	74	55	74%	75%	12-27
12-29	61	51	84%	79%	12-27
12-28	47	35	74%	70%	12-27
12-27	28	19	68%	76%	12-20
12-26	5	2	40%	60%	12-20
12-24	1	1	100%	100%	12-20
12-23	31	24	77%	66%	12-20
12-22	68	53	78%	82%	12-20
12-21	47	37	79%	71%	12-20
12-20	18	14	78%	78%	12-13
12-19	88	67	76%	68%	12-13
12-18	9	7	78%	62%	12-13
12-17	25	23	92%	92%	12-13
12-16	14	9	64%	57%	12-13
12-15	20	11	55%	75%	12-13
12-14	10	10	100%	90%	12-13
12-13	6	5	83%	80%	12-07
12-12	102	73	72%	74%	12-07
12-11	12	9	75%	100%	12-07
12-10	18	15	83%	78%	12-07
12-09	31	21	68%	81%	12-07
12-08	44	31	70%	72%	12-07

1. For NCAA College Men's Basketball: Top 25 Measures and Rankings
2. For Men's College Basketball: Top 300 teams rankings
3. For Men's College Basketball: Logit prediction success
4. For College Football predictions & Top 25 rankings
5. For College Football all 200+ rankings

How to predict

Logit measures predict as follows:
(a) Note down the BCUs of the guest team.
(b) Note down the BCUs of the home team.
(c) Add the home court advantage BCUs to the home team.
(d) The team with more BCUs will win (guest or home+advantage)
(e) Teams in the complete ranking will win over unranked teams.

Notes:

• Home field advantage is worth extra BCUs (Basketball Competency Units). These are added to the home team's BCUs.
• An advantage of 70 BCU's gives a team "2-to-1 on" odds of winning.
• Based on Rasch analysis of published results of major games available at time of computation.
• Intended solely as the demonstration of a measurement method.
• The game scores were provided by a variety of Media and Internet sources, including ABC, NBC, CBS, ESPN, AP, Yahoo and the websites of the schools themselves. AP and ESPN rankings are provided by those organizations.

This is an illustration of measurement using the method of paired comparisons. Leon L. Thurstone, of the University of Chicago, perceived the necessity that measures be independent of sampling and testing specifics. Measures based on paired comparisons meet this criterion. Bradley and Terry proposed this model but from a descriptive, rather than measurement, perspective. Consequently, they noticed, but did not capitalize on, the linearity of the measures produced. Georg Rasch perceived the linearity of measures produced from ordinal data by logit-linear models. Since mathematical expression of such models obscures their immediate utility and ease of application, they are illustrated here by a simple paired comparison analysis of NCAA game outcomes. Current applications of this method include the pricing of medical procedures and determining the relative severity of crimes.

Thurstone, L.L. The method of paired comparisons for social values. Journal of Abnormal and Social Psychology, 1927b, 21, 384-400.

Bradley, R.A. and Terry, M E. Rank analysis of incomplete block designs I: The method of paired comparisons. Biometrika, 1952, 39, 324-345.

Go to Top of Page
Go to Winsteps & Facets home Page