HANN4H, I like your graph. So much more informative than mine. Whoever said the SCD board isn't educational! I never knew what a scatter plot was before tonight. Now I am going to be sticking everything into one.
Lies, damn lies and statistics
2 of 2
manipulate the statistics a bit and you can pretty much give yourself the result you want. it's very useful to be able to do when an experiment hasn't gone quite the way you want it to. 
But surely, discounting data that doesn't fit your argument invalidates it?
If you are going to attempt a scientific analysis, you have to accept all the evidence, whether you lkie it or not, otherwise you enter the realms of pseudoscience.
If you are going to attempt a scientific analysis, you have to accept all the evidence, whether you lkie it or not, otherwise you enter the realms of pseudoscience.
Originally Posted by GlenP:
“But surely, discounting data that doesn't fit your argument invalidates it?
If you are going to attempt a scientific analysis, you have to accept all the evidence, whether you lkie it or not, otherwise you enter the realms of pseudoscience.”
“But surely, discounting data that doesn't fit your argument invalidates it?
If you are going to attempt a scientific analysis, you have to accept all the evidence, whether you lkie it or not, otherwise you enter the realms of pseudoscience.”
No, it's called discounting the outliers in the trade.
Originally Posted by GlenP:
“If you are going to attempt a scientific analysis, you have to accept all the evidence, whether you lkie it or not, otherwise you enter the realms of pseudoscience.”
“If you are going to attempt a scientific analysis, you have to accept all the evidence, whether you lkie it or not, otherwise you enter the realms of pseudoscience.”
unless your professor has told you what it is supposed to be like and you can't find a good reason for why it didnt turn out that way
i do agree though. it's just interesting to see how a little bit of manipulation can change the result entirely. zoe and colin's lines dont change hardly at all without the week one results whereas darren's does drastically which could mean that he either had a massive leap of improvement over one week which then plateaus or that the judges were unfair on him that week.
Seems the same as disallowing the other 75 pyramids, in Egypt to prove that those on the Giza Plateau mirror Orion's belt to me; "Abu Roash & Dashur don't fit the pattern so we don't tell anyone about them".
If it helps your argument and you can get away with it, then a touch of pseudoscience is the order of the day ......lol
If the politicians can dine out on it I know I can on a SCD forum !!
If the politicians can dine out on it I know I can on a SCD forum !!
Getting back to being serious .... You would have to say that week 1 for Darren was the exception to the rule and it would make more sense (and help my argument) if you ignored it.
Originally Posted by Muggsy:
“No, it's called discounting the outliers in the trade.”
“No, it's called discounting the outliers in the trade.
”
that's the word i was looking for.
ok so SCD isn't the same as a scientific experiment but i find that quite often i have to discard the first result i get in a set of readings as the reaction may not have properly started yet; which in SCD terms would mean that 1) the dancers haven't settled in yet 2) the judges haven't got into the swing of marking yet. it does give a good case for having a week or two without anyone being booted off, though i haven't looked at the stats for the other couples so i don't know if darren is the only one with an anomoly in week one.
This is all great stuff, but does it prove "Lupes Theory" ?
" the general concensus of most people on these forums is that, of all the three finalists, Darren has shown the most improvement, strictly on the judges marking over the entire series I do not agree"
" the general concensus of most people on these forums is that, of all the three finalists, Darren has shown the most improvement, strictly on the judges marking over the entire series I do not agree"
Remember the title of this thread.
I did not dispute the original graph only the possible analysis/arguments that would ensue.
I have posted on other threads showing detailed figures (can't raise my game to posting graphs). My gut feeling is that
a. Zoe has consistently scored well
b. Colin started well, dipped for a few weeks and has now improved.
c. Darren started badly in week one and then has been consistent.
I did not dispute the original graph only the possible analysis/arguments that would ensue.
I have posted on other threads showing detailed figures (can't raise my game to posting graphs). My gut feeling is that
a. Zoe has consistently scored well
b. Colin started well, dipped for a few weeks and has now improved.
c. Darren started badly in week one and then has been consistent.
ok i'm not saying that this is fact but from looking at the graph according to the judges marks:
-darren has been the most variable (real up and down of highs and lows)
-colin had been consistantly improving (look at the red dots not the line-6 of them show a good constant increase from wk4-10)
-zoe has been consistantly high at about the same level (her lows have got higher which is why her graph still has a good increase)
-darren has been the most variable (real up and down of highs and lows)
-colin had been consistantly improving (look at the red dots not the line-6 of them show a good constant increase from wk4-10)
-zoe has been consistantly high at about the same level (her lows have got higher which is why her graph still has a good increase)
Thanks HANN4H, you are clearly an exceptional student........lol
the "Colin has been consistently improving" was what I was looking for.
Case proved. The court will rise....
the "Colin has been consistently improving" was what I was looking for.
Case proved. The court will rise....
What a great thread! Thanks for making me smile tonight 
It is actually a great help in trying to figure out who should win the week... statistically anyway.. ..there just is no runaway winner this year.. I still think it is really going to depend on the dances on the night and who does one of those "Brilliant moments" in a dance. At least I think that is what will clinch it for me this year.
It is actually a great help in trying to figure out who should win the week... statistically anyway.. ..there just is no runaway winner this year.. I still think it is really going to depend on the dances on the night and who does one of those "Brilliant moments" in a dance. At least I think that is what will clinch it for me this year.
Originally Posted by bendymixer:
“I would love to know the week on week viewer votes”
“I would love to know the week on week viewer votes”
I agree. Even if they didn't want to put the number of votes, they could always give a percentage for each person. I would imagine in the first weeks the votes are fairly low and only really take off when people get more attached to various partnerships. Having said that I surprised myself by voting quite a lot for Fiona in the early weeks just 'cos I loved her on GMTV - and I felt sorry for the way Brendan was treating her.
2 of 2
