Just think of its use in OMNI/PR!

That's why I want to see a feasible path to forum-specific appearance of reader feedback before recommending even a trial. I want to see a low-risk approach in which we also don't raise expectations that IB cannot meet later.

I'm not sure fact that some people will want the counters make them valuable. I don't see the harm in them (any more than post count) but I don't really see the gain.

I'm not sure the format of FT is such that the reputation scoring aspect would work, nor curation to help "quality content rise to the top" -- we are unlikely to ever have a personalization engine (where one's own "likes" can help surface content) and ordering content by likes would open this up to exactly the sort of gaming that people have complained might be possible.

Doing this the way SlickDeals and/or FatWallet do it is exactly what we DON'T want to do.

Well said. I have become convinced of that, especially after this discussion. I think we will want to defer ANY publicly viewable compilation of votes for a given user's posts.

Not defending the math, but it is possible to view it like this:
In a group of 100 people, 50 voted against and 50 for in the first vote. In the second vote, 9 people "changed sides" and thus the second vote resulted 59-41. That is the way I read the "9% change" comment.





Can you give some examples on how voting systems helped/improved discussion boards? I don't see slickdeals as a site very similar to flyertalk.

And again, how is a poll with 700 votes out of a population of 500 000+ members, where you have no information of the statistical selection of voters, a clear indicator of the whole population?

If it's so good as to be used in the miles and points and/or other travel sections of FT, it should be just as good for use in OMNI/PR. Apparently IB created this feature for use across all forums on FT and other sites using its VBulletin board software and that is what support for this would get if this poll were to be implemented as voted now.

About the poll's representative/unrepresentative nature of FT membership base, I'm not sure that matters. Most members are probably apathetic about such matters, even if the poll margin is off by +/- 9 percent or more.

I think the reluctance to try it in OMNI/PR is perhaps because it would, most likely, not be used in the spirit it was intended there, and would cause problems, thus making it less likely to be implemented elsewhere on FT.

It's a nice effort :) but the post was completely incorrect.

If a referendum in real-life (as opposed to here) is held and it passes 51-49-- it "won" by 2%.

If the result was 59-39-- it passed by 20%. Which would then be described as a "landslide."

We can most definitely disagree about if this whole this means much of anything; % of FTers that participated, do they even understand what they're voting on, are people opposed more likely to just not vote-- all valid issues and none I'd really argue. Maybe it should be like a totalitarian republic where everybody is forced to cast a vote.

But, I would argue, that at least be honest and accurate about the results as they stand-- meaningless or meaningful as they may be. If the vote is to be ignored or dismissed-- fine, there may be excellent reasons to do exactly that, or to just use it as one of many factors, or to simply decide "this feature is foolish, better for FT if we -don't- implement it at this time-- but accuracy about the results shouldn't be optional.

I'd also add-- but most certainly cannot support/prove/know, obviously-- that if the results of the poll were the exact opposite (20% margin against) that those who oppose this feature would be bandying that result about as "a clear rejection of turning FT into Facebook!" and "a decisive defeat for this dumb idea!" And they wouldn't be talking about "well, only a fraction of FTers voted, so...." and/or "well, if you look at it, it's still basically not far from 50/50, right??"

Be honest-- you know it's true. :D

Well, that is not entirely mathematically true either. I'm lacking the correct word in english, but a 59% to 39% result won by 20 percentage units, not by 20%. In fact, it won by 51%, ie the winning side is 51% larger than the losing side (59/39=1,51).

(See, you can have a 51% lead for free! I'm not only honest but also generous :D

As I said, I wasn't out to defend the math, just to say that I understand the reasoning behind it. IMO no dishonesty was meant by the commenter.


About the landslide, I guess that term is somewhat dependent on ones cultural environment. For me personally a 60-40 is nowhere near a landslide. For me, a 60-40 contains one small majority and one large minority.


I wouldn't, but you are probably right this would happen either way a poll would go. So how about - we agree on setting ourselves above that? :)

Exactly. It's 3:2, which is suggestive but not dispositive. We have a lot of work left to do before proceeding with anything.

Fact vs. Conjecture
 

Per your own logic, you would know it's true when the yes camp could place the same argument if the no votes were taking the lead in this poll, wouldn't you? ;)

I personally don't think this line of what-ifs conjecture would add any value to the discussion especially when you can't provide evidence to prove your "theory."

Exactly!

It's a fact FT now has 580,319 members. With 712 voters so far in this poll (which is only 0.123% of FTers), no one can claim the majority of FTers have spoken, no matter what the result is.

Sorry to disappoint you. Not until you get 290,160 FTers (aka 50% of current 580,319 registered members) to show their support, nobody knows what the majority of FTers want for anything. Yes, statistics is a b*tch. :) See Post#645 by halls120 for a more accurate description of the voting trend in this poll:

I think the bottom line is how a new addition, such as this LIKE/Helpful button, can really improve a non social media like FT. That's the million-dollar question for our TB and CD to ponder upon.

Sure, it's going to work like this:

OMNICon: Obama is bad!
--> OmniCon2, OmniCon3, OmniCon4 like this post.
OMNILib: Bush is bad!
--> OmniLib2, OmniLib3, OmniLib4 like this post.
OMNILibertarian: You're all statists!
--> OmniLibertariane likes this post.

...but beyond being pointless, so what? Seriously, what problems would that cause?

(In the best case, it might actually reduce some of the less-substantive echo chamber posts, and just lead to likes on the other guy who said it first.)

How many of those users are active monthly, or daily? How many of those users have no posts, or 10 or fewer posts? Does IB ever purge old users?

From experience with other similar-but-smaller systems, the real "active user" population is going to be vastly smaller than the historical total number of signups.

The fact that we've now got nearly twice as many votes as the talkboard election got suggests that there's a great deal of interest in this issue, beyond simply a core of FT insiders.

IIRC, the active population hovers around 10k people. I do not believe IB has ever purged the user rolls.

So that means that 500k people includes spammers of all kinds, people joining for contests, and those who registered 12 years ago, asked a single question, and haven't been seen since :)

IHMO, the user rolls should be purged after a certain amount of years of no activity - certainly someone who joined in 1998 and hasn't posted since 1999 shouldn't be counted as a user any longer.

I agree.

As a real life example, the result of last year's Scottish Independence referendum went 55/45. The 55 was broadly viewed as a commanding majority