Well, this natural perception is the subject of my purpose. I'll try to demonstrate (statistically of course) the difference between the forecast trend line and the real line for each country in the last 5 editions in the votations to the other countries.
First of all, data has been extracted from the eurovision web page and any failure in the votations will be attributed to the web page creators because of the copy/paste from the same page to the spreadsheet.
Let's start!
The question in the study case is: Based in the last 3 historic years(2005-2007) , what could be the trend for the 2008 year?
We'll analyze first the trend lines for a country in the center of East-Europe like 'Latvia'.
The example is shown in the graphic.
In the graphic are relevant these two lines:The black line is the cubic average real line for the votations from the 2005 year.
The green line is based in the 2005-2007 range years and try to make a forecast for the 2008 years.
As you can see, the solution show a difference very small between the two lines in 2008, but this, in this way, not say nothing if we don't compare with the difference in other countries.
Choose now the example of Spain, in the sud-west of the europe and only surrounded by three countries: Portugal, France and Andorra.
The objective is the same:
The black line is the cubic average real line for the votations from the 2005 year.
The green line is based in the 2005-2007 range years and try to make a forecast for the 2008 years.
As you can see,the difference here is relevant, in other words, in most difficult to forecast the 2008 year for Spain than for Latvia (even for next years).
In fact, this is due to the votations system used. From 2005 to 2008 the votations were 100% choose by the people in each country. There are several factors which contribute to these votations but is not the purpose of this document.
