RE: [Epic] Final Flyer Formula

From: Kelley, Kevin J., JCS <kevin.kelley_at_...>
Date: Fri, 21 Nov 1997 08:23:37 -0500

Form 4 has a reasonbly small constant. I can drive the intercpt to 0
but remember that in theory you shouldnt use this to estimate anything
outside the range of the original data ie <15 or > 80. If we assume the
worst flyer possible is
r=0, fp=0, int= 1, gun= 0, arm =1, then form 4 gets a value of .62
which isnt bad.

> ----------
> From: C.Sinclair[SMTP:csi_at_...]
> Sent: Friday, November 21, 1997 4:16 AM
> To: space-marine_at_...
> Subject: Re: [Epic] Final Flyer Formula
>
> At the risk of turning the list into an "econometrics of GW points
> valuation_at_ workshop. Does the model work without a constant? Surely
> all
> the variation in CV should come from the variables? Of course, if the
> formulae don't work without the constant then keep it in! Just
> curious.
> >Formula (3) -
> >Combat Value(CV)= 2.1*(SQRT(Rng*FP)) + 1.1*Int + 1.9*Gun +
> > 5.8*Arm + 14.4*T- 13.2
> >T=1 if it can transport, 0 if not.
> >
> >
>
> >Formula (4) -
> >Combat Value(CV)= 2.56*(SQRT(Rng*FP)) + 2.7*Int + 1.2*Gun +
> > .42*(Arm)^2 + 14.4*T/2- 2.5
> >
> ******************************
> Colin Sinclair
> Department of Economics
> Glasgow Caledonian University
> Glasgow
> G4 0BA
> Scotland
>
> Tel: +44 (0141)331 3350
> Fax: +44 (0141)331 3293
> E-mail: csi_at_...
> ******************************
>
>
Received on Fri Nov 21 1997 - 13:23:37 UTC