[Data-modeling] 2008 California partisan elections

[Robert Cook]

I was thinking of something simple:  Who won.  All the other stuff  
could be modeled as well, and although that data (when eventually  
added) would make the simple property redundant, I think that it has  
such a different audience that denormalization is OK.

R

On Apr 8, 2008, at 4:21 PM, Christopher R. Maden wrote:

> Robert Cook wrote:
>> Would it make sense to have a property on Election for who the winner
>> was?
>
> That’s possible.  I believe Tadhg is actually looking at modeling
> election results, but it’s non-trivial.  One way to do it is for
> contesting an election to be modeled as a CVT, with a Boolean for  
> being
> elected and other properties for vote totals and/or percentages.   
> But it
> is more complicated for multi-seat elections, parliamentary elections,
> ranked choice voting, etc.  Right now, the important part — who holds
> the office — is just modeled completely separately.
>
> Jeff Prucher wrote:
>> I wonder if we shouldn't change "election year" to "date of  
>> election"; I
>> think that would make things a bit clearer.
>
> Makes sense to me.  The existing dates could be changed easily enough.
>
> ~Chris
> -- 
> Christopher R. Maden
> Data Architect
> Metaweb Technologies, Inc.
> <URL: http://www.metaweb.com/ >
> _______________________________________________
> Data-modeling mailing list
> Data-modeling at freebase.com
> http://lists.freebase.com/mailman/listinfo/data-modeling