Everyone has their favorite state, and we tend to emphasize the features of that state that put it in the best light. This is "putting the best foot forward."

But the spreadsheets are our connection to reality. Those features we emphasize for our states are often spreadsheet variables. So we should be able to rank variables in a way that is 1) plausible, and 2) puts our state high in the rankings. Other things not quantified, the "intangibles", also have an effect on our choices, but if you can't first get your state pretty high in the rankings (if not in first place), you are not being realistic, and are not making FSP success pre-eminent.

Just as we ranked states in other threads,

**let's try ranking spreadsheet variables** in importance. This should correspond to the weights you assign when you run the spreadsheet.

I'd also like to use this thread to investigate the notion of dependent variables. These are pairs of variables that are related, or "correlated". One prime example would be the Vot, Pop, and VotingAgePop variables. I have run a correlation between Vot and VotingAgePop variables on the big spreadsheet (it is a spreadsheet function, easy to do), and find the correlation coefficient is .97 (1.0 would mean "perfectly correlated"). This means these two variables are essentially the same variable. When weights are assigned to Vot and VotingAgePop, putting a 10 in each of them is pretty much the same as putting a 20 in one and 0 in the other. The same is true for any other pair of highly correlated variables, so be careful you don't overweigh these by assigning high weights to both.

BTW, I am using variable names that appear either on the small spreadsheet or the big one. I will send my big spreadsheet (that has many more rows than the small one) to anyone who asks.

Here are correllation coefficients for pairs of interesting variables:

Pop, VotingAgePop: 1.00 (perfect!)

Vot, VotingAgePop: .97

Vot, Pop: .95

GovEmp, NEATeach: .81

Pop, Job: .79

Dep, FarmSubsidy: .66

Tax, Ideology: .60

EFI, SBSI: .56

Area, PrivateLand: .52

Tax, Revenue: -.69 (now

*that's* interesting!)

Bottom line here, is that I would use just one of Pop, Vot, or VotingAgePop variables, and zero out the weight on the other two. And I'd down-rate the Job weight somewhat since it correlates pretty well with Pop, and NEATeach since it correlates pretty well with GovEmp. If anyone wants any other pair of variables checked, let me know and I will add them to this list.

Now to the ranking of variables. Here's how I rank them, along with the percentage of total weight I assign to them (note that both spreadsheets totals all the weights, so if you make sure the total equals 100, then each weight becomes the percentage of total weight). First for my big spreadsheet:

Vot: 20%

Ideology: 8%

Gov2, Dep, Pres, GovEmp, UrbArea%, PrivateLand, EFI, SBSI: 5%

Job, Tax, NEATeach, Revenue: 4%

Area, Geo, Blm, Gun, Homeschool: 2%

Native%, Fin, Inc: 1%

Land, HunterOrange, HunterTraining, MCHelmet, BikeHelmet, SeatBelt: 0.5%

This yields 728 for WY, a 67-point lead over 2nd-place SD.

Now for the small spreadsheet:

Voters: 25%

EFI: 10%

Spending, Dependence, Taxes: 8%

Geography, Prez, UrbanAreas, GovEmp: 5%

Jobs, NEA, PrivLand: 4%

GunControl, Homeschooling, Natives: 2%

Finance, Income, LandPlanning: 1%

This yields 652 for WY, a 40-point lead over 2nd-place SD.

I'm interested in any other rankings of variables folks want to come up with. Then I'd like to haggle over these; I certainly think my numbers can be adjusted to be better. Tell me if you think I've weighed things too high or too low.

Finally, I'd suggest if you can't come up with weights that put your favorite state in the top 3, you might as well forget about intangibles, and choose another state. Or find more spreadsheet data that will help your state out!