Pred

Predicted Conference Finish, 1995-96

The seven statistics that go into the prediction score are now added such that the sum should reflect each team�s league points in the coming season. This requires a little playful mathematics, as described below. For each team, first we compute AvgMod, the average of the five percentage modifiers. The mean of AvgMod is itself used in calculating an adjustment, AdjMod, which, when added to RS and PS produces the predictor of conference points, Pred.

For the hardcore types: 

AvgMod = (5Pct + 1/2 + RetG + RetD + RetM) / 5
AdjMod = 44 * (AvgMod - MEAN[AvgMod])
Pred = AdjMod + RS + PS
Team		Pred96	RS	PS	5Pct	 1/2	RetG	RetD	RetM	AvgMod	AdjMod

Clarkson		30	31	-1	.682	-.136	.554	 .676	 .995	.554	  0
Brown			28	28	-1	.608	-.182	.736	 .688	1.000	.570	  1
Colgate		27	25	-1	.494	 .136	.835	 .646	1.000	.622	  3
Vermont		27	24	 0	.534	 .273	.713	 .619	 .954	.619	  3
St. Lawrence	26	20	-1	.528	 .182	.914	1.000	 .984	.722	  7
Cornell		25	20	 1	.420	 .182	.790	 .842	 .967	.640	  4
Harvard		24	25	-1	.727	-.045	.551	 .709	 .865	.561	  0
RPI			24	23	 2	.557	-.227	.537	 .854	 .968	.538	- 1
Princeton		23	21	 2	.409	 .045	.699	 .618	1.000	.554	  0
Dartmouth		14	16	 0	.290	 .182	.479	 .710	 .931	.518	- 2
Union			12	16	 0	.284	 .000	.670	 .830	 .572	.471	- 4
Yale			 4	15	 0	.466	-.227	.462	 .714	 .083	.300	-11
										
Sum		   264	264	0	5.999	.183	7.940	8.906	10.319	6.669	0
Mean			22	 22	0	 .500	.015	 .662	 .742	  .860	 .556	0