From how I'm reading it your method was correct in the first example, you're second method confused me. As the AV system only counts the second choice votes of the crisps with the least votes.
Well, for the second method I looked at the rankings of the candidates. So the flavour with the most "8"s was removed from the counting, then in the second round the flavours with the most "7"s was removed as to me that seemed a more scientific way of doing it.
But then I always see different patterns in data than I should.
When the public vote in the polls they don't generally place any votes for a party they don't want so you wouldn't get loads of negative votes for AW flavour.
It was however interesting to see how the result changed depending on the method used
I realised that not everyone would vote for all the flavours and I also realised that some people would vote incorrectly (which they did) but I made no assumptions and added up the votes with the data presented. Especially as if/when AV does become commonplace you can guarantee that there will be some people who won't fill in their voting slip right.
Comments 4
Reply
But then I always see different patterns in data than I should.
Reply
It was however interesting to see how the result changed depending on the method used
Reply
I realised that not everyone would vote for all the flavours and I also realised that some people would vote incorrectly (which they did) but I made no assumptions and added up the votes with the data presented. Especially as if/when AV does become commonplace you can guarantee that there will be some people who won't fill in their voting slip right.
Reply
Leave a comment