Hacker News new | past | comments | ask | show | jobs | submit login




The results of a straw poll after a debate are not evidence for anything beyond what the people who saw the debate thought of the debaters. People vote for all kinds of reasons - the underlying merits of the case, the impressiveness of the speakers that they've found to present it, even sympathy for someone who got obviously crushed by their opponent. When I was a debater I've seen some truly outrageous motions get passed simply because one side was funnier.


I understand. I was responding to comment that said its totally baseless caricature.

I think debaters did a good job of presenting evidence for their case. Why not give it a chance instead of saying its "staw poll", isn't any poll "straw poll" by that account?

Also I never said it evidence of anything. What is an example of "evidence" in this particular case for you?


Ah, sorry. You said "results of this debate disagree with you", so I took that to be the debate's results (the vote), not the debate itself.

(And as a side point, no to the straw poll question - there are certain statistical guarantees on the representativeness of polls as long as they're drawn from a random sample of the public, with debates about how far you can deviate from a random sample and still have results that are useful. Self-selected polls, on the other hand, pretty much mean nothing.)


> Self-selected polls, on the other hand, pretty much mean nothing.

Isn't our democracy built upon "self selected polling" , people who chose to go excesrice their right to vote. Its all meaningless?


If few enough people vote, arguably yes (though that's a fairly complex question and depends a lot on what your conception of democracy is). But in essence, an election is a full-population poll: like a random sample with probability 1 of being chosen. So it's statistically maximally valid. We (political scientists) tend to squint a bit and say the opportunity to vote is equivalent to actually voting, so non-voting doesn't harm the legitimacy of elections. I don't personally buy that, but it's a tricky problem to fix.


So if am I reading this correctly.

Previously you said "any self selected polling is meaningless"

Now you are saying "Self selected polling is meaningful when people who didn't self select had the opportunity to do so" .

Ok. So by that definition the poll I linked, everyone had the opportunity to participate in the debate, theoretically. What makes it meaningless ?

PS: Fixing democracy ect is not interesting to me in this context, I want to focus your claim self-selected = meaningless.


Because I, for one, didn't know it was happening and therefore could not participate. My chance of having my opinion count was zero. So pretty much by definition the poll is not representative of the universe of people-who-include-me. And by induction of all the other people who had no chance to participate.

If you want to know more, I recipient looking up the 1936 Literary Digest poll for US President, one of the most famous screwups caused by non-random sampling: https://www.math.upenn.edu/~deturck/m170/wk4/lecture/case1.h...


its the same with US polls, everyone having an opportunity is merely theoretical. Lots of people are unable to votes for all sorts of reasons.

Perhaps explain how you are defining * opportunity to participate* .


If you're arguing that elections with less than 100% turnout are flawed as models of public opinion, I agree with you.


There's a single democracy. There's no shortage of random polling websites. Telling me "You are ruled by this person, who was selected in a process you knew about since childhood" is one thing; telling me "This is likely true because people on this website you've never heard of say it's true" is another.

(And yes, governments where there are ad-hoc rationalizations for who gets to vote or where the procedures aren't publicly announced are in fact illegitmate and widely recognized as such.)


I didn't say it was the "truth" or anything like that. What are you even saying?


I claimed a thing was true. You said that the results of a debate disagree with me. Either you meant that my position was likely to be untrue as a result, and in turn the position advocated by the "winner" of the debate was likely true, or you didn't mean anything relevant at all. Which is it?


It merely meant lots of people were convinced that its true based on the arguments provided by side arguing for the motion.

"Undecided" went from 8% to 24% for the motion.


... pardon my assumption, but those four people don't look like college students. Are they?




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: