Okay, in he past, YouTube views for a video have proved a fairly reliable indicator of popularity and results. Looking back at last week's videos, and equating the number of views of the BBC channel videos to public support, this is what was get:

Judges Public Total Views
Austin 11 11 22 16834
Rachel 10 10 20 7029
Tom 10 8 18 3218
Lisa 7 7 14 2861
Cherie 8 5 13 2539
Heather 6 6 12 2702
John 2 9 11 4428
Jodie 5 4 9 2139
Christine 3 3 6 1908
Andrew 4 2 6 1781
Mark 1 1 2 1762

Which, at the bottom at least is exactly what happened. Christine, in this case, would have escaped the dance off due to a higher public score.

Now this week's 'results'. I've put the predication in spoiler tags in case anyone doesn't want to see even this - please be considerate and don't mention the spoiler here!

I see where you're coming from on this Claire but which channel are you getting your YouTube viewing figures from? Surely the vids are posted on more than 1 channel. I watch them on Maltipom's but the BBC also does them as do others and some people may just watch on iPlayer. There may well be a link to these figures and the results but I don't think you've captured the full picture there.

Sorry, I'm really not trying to be mean just statistically correct.

Oh you're absolutely right and if I had time I would add up them all.

However, I used the BBC ones because they generally come up first when you put things into the search engine (therefore if people are looking for a dance, they're the one they find).

Also, the BBC ones tend to get the most views, hence should give the most accurate representation - for example, Austin's VW got 16,300 on the BBC video but only 700 on Malitpom's...

It's certainly not a 100% complete picture but it seems to turn into a very good indicator.

Another problem that I have about it Claire is that if you look at the viewing figures today for last week they surely may have been influenced by the result which everyone knew about last Sunday evening. In that case are they not a "reflection" of the result rather than an "indicator"?

That being said, viewing figures for this week will be a better test of your theory though and I will come back and look later tonight when the result has been broadcast.

That may be true for the bottom 2 but obviously people didn't know where the likes of Austin and Rachel came in the final combined leaderboard and their massive number of views does seem to suggest that people are enjoying and are interested in what those couples are doing above others.

I think we're taking the same argument Claire but coming up with different conclusions.

I agree we don't know the combined result in any order, only who was in the bottom 2. The order you have above could be right (and is very likely) but it could just as easily be all wrong except for the bottom 2 thereby not fully supporting your theory.

I think that last week there was a very predictable result in that the worst performer went out and it was no big surprise that John was saved above Andrew as he is proving very popular with the public. It might be better to look at a more surprising result such as the week Don went to get a picture of this but as so much time has passed it's not likely to be that accurate now.

The results must give some comfort to Rachel and Vincent (as well as obviously Austin and Erin) because it looks as if they are more popular with the public than the judges. Of course, not everyone who watches bothers to pick up the phone to vote.

I was very surprised at Rachel's seeming popularity since she didn't seem to attract much interest here.

Still, this forum never has been exactly representative of the views of the general public. We usually have the same winner but lower down the board, our results don't seem to match with what the public want.

Bumping to point out that YouTube seems to be a pretty accurate indicator - it would have had Heather and Andrew in the bottom 2 as well tonight.

Interesting...