NCAA Women's Soccer			Somis Sports
	Most improved teams
	2023 Season
	Explanation:
	In determining the most improved teams, the primary Somis Sports objective is to favor neither stronger nor weaker teams.  Specifically, the general guideline is that the mean CY22 ranking percentile of the top 50 CY23 most improved teams should be around 50%.  For example, in CY22, John Carroll was ranked 164 out of 431 D3 teams (ranking percentile 62.1%) and Whitman was ranked 283 out of 431 D3 teams (ranking percentile 34.5%).  Using this calculation methodology, the overall CY22 ranking percentage of the CY23 MIT top 50 is 53.5% and the MIT top 100 is 50.5%.  To achieve this objective, we lessened the CY23 improvement of those teams that had a negative CY22 rating (Section 2B).  Otherwise, the MIT listing would tend to favor the weaker CY22 teams.
	A secondary objective is to somewhat balance the MIT teams among the three divisions (D1, D2 and D3).  Sections 2A and 2C below show the two adjustments which are designed to offset the differing levels of variability among the three divisions.
	Below is an example.  In the example, Team A is in D1 and Team B is in D2 and Team C is in Division 3.  The calculation involves the following steps:
	1. Calculate the amount by which each team improved.
		Division 1	Division 2	Division 3
	Rating increase (points)	Team A	Team B	Team C
	2022 rating	(0.97)	(2.05)	(5.08)
	2023 rating	0.77	(0.33)	(0.92)
	Net increase (points)	1.74	1.72	4.16
	2. Apply three adjustments to the points increase as follows:
	A. Divide each team's points increase by V.  Note: V is a divisionally calculated value based on the variability of the teams within each division (D1, D2 and D3).
	B. For those teams that had a negative 2022 rating, lessen the team's net increase (see Section 4, for details of this calculation).
	C. Standardize each team's points increase based on the variability of its division's interyear team rating change (CY22 vs. CY3).
		Division 1	Division 2	Division 3
	Rating increase (points)	Team A	Team B	Team C
	Net increase (points)	1.74	1.72	4.16
	2A. Restate in terms of V	1.052	1.077	1.185
	Net increase (V points)	1.654	1.597	3.511
	2B. Negative adjustment	(0.132)	(0.486)	(2.462)
	Net increase (V points)	1.522	1.111	1.049
	2C. InterYear Variability	1.088	1.003	0.937
	Net increase (V points)	1.656	1.114	0.983
	3. Switch the V value back to points.
		Division 1	Division 2	Division 3
	Rating increase (points)	Team A	Team B	Team C
	In terms of V	1.656	1.114	0.983
	Points per V (average)	1.114	1.114	1.114
	Adjusted increase	1.84	1.24	1.10
	Comments:
	4. Details of Section 2B adjustment (applies to teams with a negative CY22 rating).
	First, calculate the improvement in the team's rating.
	Team ratings	Team A	Team B	Team C
	2022 Rating	(0.97)	(2.05)	(5.08)
	2023 Rating	0.77	(0.33)	(0.92)
	Net improvement	1.74	1.72	4.16
	Next, restate the improvement in terms of V.
	Team ratings (in terms of V)	Team A	Team B	Team C
	2022 Rating (V)	(0.922)	(1.903)	(4.287)
	2023 Rating (V)	0.732	(0.306)	(0.776)
	Net improvement  (V)	1.654	1.597	3.511
	V rating used by division	1.052	1.077	1.185
	Then, allocate the improvement by range.
	Allocation within V ranges	Team A	Team B	Team C
	Above 0V	0.732	-	-
	From -1V to 0V	0.922	0.694	0.224
	From -2V to -1V	-	0.903	1.000
	From -3V to -2V	-	-	1.000
	Less than -3V	-	-	1.287
		1.654	1.597	3.511
	Finally, calculate the adjustment by range based on the percentages listed parenthetically below:
	Apply adjustment	Team A	Team B	Team C
	Above 0V (0%)	0.000	0.000	0.000		0.000
	From -1V to 0 (14.3%)	(0.132)	(0.099)	(0.032)		0.143
	From -2V to -1V (42.9%)	0.000	(0.387)	(0.429)		0.429
	From -2V to -3V (71.4%)	0.000	0.000	(0.714)		0.714
	Less than -3V (100%)	0.000	0.000	(1.287)		1.000
		(0.132)	(0.486)	(2.462)