World of Chig   

26.9.02

Let's have a heated debate!


Before finalising the chart positions later today, I seek your advice, especially if you’re a statistician (or a moral philosopher, he said, grandly). Not surprisingly, there are a few tied positions in the top 50, where two or more tracks have received the same number of points. It would be boring if we had lots of tied records in the same positions in the chart. It would also mean that the countdown would jump positions, say from ‘equal 49’ to ‘equal 47’, and we don’t really want that. So the question is, how do we separate them, fairly?

The problem is that the method used can determine or change the final result. Two possible options are the ‘countback’ method, (as used at Eurovision) or the ‘most voters’ option. The countback method means that when songs receive equal points, we then look to see which one received the most 10s. If they are still tied, which one received the most 9s, then 8s, 7s, etc. The other possibility is to count how many different people voted for each song, regardless of how many points they gave. The latter method would imply broader appeal and therefore arguably the most popular song, but the winner by the countback method has higher points and therefore STRONGER support. Which do you think is most fair?

Let me show a hypothetical example;
Song A and song B are tied on 33 points.
Song A has received votes of 9+8+7+7+2 = 33
Song B has received votes of 9+7+6+5+3+2+1 = 33.

Using the countback system, both songs have received one 9, but song A wins because it has received an 8, whereas song B’s next highest vote is a 7.
Using the ‘most voters’ system however, Song B wins, because seven people voted for it in their top ten, against five people who voted for Song A.
It’s a dilemma, isn’t it? Your opinion is valued. Please leave a comment below. Or if you have another, better method of separating tied scores, please let me know that too. Thanks.

· link

Home