I identified seven different selection methods, however all seven were characterized by two things, whether or not they allowed the audience to vote and/or used an internal jury. I then found the average number points countries got in the semi-final and final categorized by selection method.
Semi-Final Points | Final Points | |
Average points regardless of selection method | 66 | 100 |
Used both audience vote and internal jury | 82 | 139 |
Internal jury involved | 70 | 131 |
Audience vote involved | 71 | 100 |
Internal jury only | 41 | 98 |
Audience vote only | 58 | 53 |
Seems the UK is right in mixing up their selection process as their previous method of just having the audience vote has a poor record. What I find interesting is that, alone, neither audience voting nor internal jury have good results, but when they're somehow used together, results improve. There seems to be a kind of synergy between them. One thing to note is that all gimmick acts were chosen by audience voting only, so possibly juries keep obviously bad acts from getting through. On the other hand, having audience participation may lessen the chances of getting an unpopular act as the jury's selection has a lot more to do with the kind of people in the jury than the popularity of the acts. Juries may tend to pick the best quality act, but Eurovision is all about popularity.
Specifically, the most successful method of selection in 2008 was when the act was picked by an internal jury, but the final song was chosen by the audience. This method was used by Armenia, Israel, and Ukraine. This method netted an average Semi-Final score of 132 and Final score of 184. Why this method was so successful (even though none of those countries won), I can't say.
The least successful method was actually the exact opposite of the most successful method, having the act picked by the audience and the song picked later by an internal jury. Of course, only one country used this method, Montenegro. It got 23 points in the Semi-Final and did not participate in the Final. Unfortunately for the UK, this is similar to the method it plans to use this year.
Just some general information, of the 43 countries that participated last year, 17 used audience voting only in choosing their entry, 8 used only an internal jury, and 18 used a mix of both. Belgium probably has the oddest method in that on even-numbered years (2008, 2006, etc) it uses only audience voting, but on odd-numbered years (2009, 2007, etc) it uses only an internal jury. This is partially because they alternate between letting Flanders and Walloon pick their entry.