The conclusive headline (use of the word will) is the Guardian's editorial decision, not used by the scientists, if that's the main objection.
The article has a more measured statement: "Extrapolating the best fit model into the future suggests that Facebook will undergo a rapid decline in the coming years, losing 80% of its peak user base between 2015 and 2017."
Observing common network effects and using it to make predictions seems like valid scienfitic research to me. The hypothesis may or may not be correct, but it's a (very) testable and falsifiable hypothesis. The science is good.
Edited to add - jarrett: I accept your point that the research could be more thorough; it's a shame only Myspace data appears to have been used. However, I think this is a valid thread of research and the word 'bullshit' was unwarranted.
Yes, it is a testable and falsifiable hypothesis. So is "George Washington will be elected President in 2016." Nonetheless, we can fairly argue that both hypotheses lack a firm grounding in the evidence.
> Observing common network effects and using it to make predictions seems like valid scienfitic research to me.
It's valid I suppose, but purely speculative at this point. It would be much more meaningful if a study demonstrated that social networks almost always behave like viruses in both their growth and decline, than to simply observe a similarity and growth and extrapolate the decline.
Which of the above posts is "just saying 'I don't like it' over and over?" This thread is a discussion of the structure of a certain argument. To recap:
swombat3 explains why he thinks Princeton's method of reasoning by analogy is flawed. ronaldx disagrees, asking why the analogy wouldn't be applicable. I reply that, in my view, ronaldx has disagreed with a different point than the one swombat3 made. ronaldx replies in defense of Princeton's claims, citing the testability and falsifiability of the hypothesis. I reply that there is little evidence for the hypothesis, despite its testability and falsifiability.
So, all participants have advanced specific, well-articulated arguments throughout. I don't see anything similar to "I don't like it."
Humorously, this discussion has become rather meta by now. We're now arguing about the form of argumentation used in a discussion that was itself about forms of argumentation. It's like we're living out an LSAT question. I find that sort of conversation hard to resist, so I suppose I'm just rising to the bait.
>swombat3 explains why he thinks Princeton's method of reasoning by analogy is flawed. ronaldx disagrees, asking why the analogy wouldn't be applicable. I reply that, in my view, ronaldx has disagreed with a different point than the one swombat3 made. ronaldx replies in defense of Princeton's claims, citing the testability and falsifiability of the hypothesis. I reply that there is little evidence for the hypothesis, despite its testability and falsifiability.
There's not an argument in this entire paragraph. Is it meant to be a list of pointers to arguments?
Neither "X is(n't) flawed," "X is(n't) applicable," nor "X has (very little/lots of) evidence" is an argument.
An argument for falsifiability is an argument that the statement of a particular belief has content and is not just a trick of language, but can't be said to be an argument for or against that particular belief.
> There's not an argument in this entire paragraph. Is it meant to be a list of pointers to arguments?
As stated, it's a recap, or summary, of arguments previously made. So no, the arguments themselves are not contained or fully expressed in the summary. In this sense you're correct that it's a list of pointers to arguments--arguments which can be readily located earlier in the same thread.
> Neither "X is(n't) flawed," "X is(n't) applicable," nor "X has (very little/lots of) evidence" is an argument.
Correct. Those are conclusions. The arguments presented in support of those conclusions are available earlier in the thread.
> An argument for falsifiability is an argument that the statement of a particular belief has content and is not just a trick of language, but can't be said to be an argument for or against that particular belief.
Agreed. That's why I said the falsifiability of Princeton's hypothesis was not an argument for the likelihood of that hypothesis. I think we're on the same page about this.
Follow-up reply to ronaldx's edit: I personally would not have (and didnt't) use the word "bullshit" in this context, so I agree with you there. I still agree with swombat3's overall argument, though.
I believe "3" comes from "3 hours ago", which was displayed next to your username at the time when user jarrett mentioned you. So I presume it's just a slight copy-paste typo.
The article has a more measured statement: "Extrapolating the best fit model into the future suggests that Facebook will undergo a rapid decline in the coming years, losing 80% of its peak user base between 2015 and 2017."
Observing common network effects and using it to make predictions seems like valid scienfitic research to me. The hypothesis may or may not be correct, but it's a (very) testable and falsifiable hypothesis. The science is good.
http://arxiv.org/pdf/1401.4208v1.pdf
Edited to add - jarrett: I accept your point that the research could be more thorough; it's a shame only Myspace data appears to have been used. However, I think this is a valid thread of research and the word 'bullshit' was unwarranted.