Would a vote counter appearing by default be a good thing?

• I wouldn't like to see the voting mechanism replace discussions. Even if things seem to be said already, new points of view can be brought up. -- ArthurClemens - 29 Sep 2004
  • I strongly agree with Arthur on this one. This is why I removed the vote from the WebTopicEditTemplate -- PeterThoeny - 30 Sep 2004

If people vote strongly (especially against consensus) then they have a duty to provide reasoning to support their view. If however the conversation captures what people want to say then voting is a handy way to get nods from the non-speakers. The majority is too often too silent. They might agree or disagree but not have the time to write.

There is nothing stopping us from asking the individuals who vote against the consensus to explain themselves, nor from their being multiple votes to chose among options on a page. All of this provides food for thought which in turn encourages discourse and which in turn encourages further differentiation in the form of new pages on about topics.

Before the enforcement of preview was abolished I often could not bare the though of waiting for the pages on SF to load to express myself, especially if what they said was close to my thoughts. I'd just think "I agree".

I thus posit that voting is a good thing and should be used liberally.

-- MartinCleaver - 30 Sep 2004

Boy, here's a discussion topic wrought with confusion! I hardly know where to start sorting it all out but I'll make a couple of jabs.

Point 1: The issue is not voting versus discussion. No one is proposing to replace discussion with voting. If fact, voting is the wrong term here. Voting implies a decision-making process and none of this has anything to do with decision-making (more on this in a minute). What is being proposed is simply increasing the use of polls to guage the broad sentiment of the TWiki community on a particular point.

Point 2: Polls merely offer a another and different channel of information from discussion. Discussion is great for unwrapping an issue and developing alternative solutions. However, discussion is structurally limited in terms of how many voices can be included. Polls offer little in terms of developing potential solutions - however they are an efficient way to get feedback from a larger sampling of the community about which solution is preferred by the greatest number of people.

This much seems to me should be obvious. What is not obvious and has puzzled me for some time is the apparent resistance to more effective feedback mechanisms in the TWiki community. The only thing I can figure is the perception that more efficient feedback mechanisms might have the effect of highlighting, how should I put it, certain anolomolies in the decision-making process. What would happen if, say, a poll shows broad support for a particular feature or change and that sentiment is ignored by the decision makers?

Point 3: Polling requires well-articulated proposals. I am tempted to digress into a broader discussion of the decision-making process in this community but will resist this and return to the specific question at hand: "Would a vote counter appearing by default be a good thing?" This question is somewhat too simplistic and as such makes very little sense. What polling question would be relevant to every topic? I can't imagine one. That being said, I fully concur with Martin's suggestion that "voting is a good thing and should be used liberally." However, what this implies is that we also develop specific, well formulated proposals that are suitable for submitting to a poll - and that's something we could use a lot more of also.

Point 4: Some polls should be standard in certain templates. For example, I would propose that all feature requests or proposals should have a poll to guage the broad community's support for that feature. Later note: I just remembered that this is about to be implemented with PluginAppraisal. Great trend! - LB

Conclusion: No harm in trying more polls, possible harm in denying this. Bottom line: I'd like to see a lot more liberal use of polls. Why not try it and see how it goes? To suggest that it should not be tried because it would somehow subvert substantive discussions seems to me unfounded and, to be honest, slightly disingenuous.

-- LynnwoodBrown - 30 Sep 2004.

As Anton seems to have experimented: there is currently no way of someone voting twice - both votes get counted. Perhaps someone would like to demonstrate their SpreadsheetPlugin skills with a nifty count "only the last vote by any given person" equation?

-- MartinCleaver - 30 Sep 2004

To comment on the apparent resistance to more effective feedback mechanisms in the TWiki community as that could seem to be related to my top comment; my comment seen in isolation can be interpreted in many ways, especially right under this topic's title.

In the original context, I was reacting on Martin's spur: I thought the poll was clear for the issue of users in Main... if you have an opinion on this add your vote. (RenameTheMainWeb) This seemed to replace discussion with voting. I am not against voting in general, but I think the issue should be well stated, so people know what they are voting for. Also it would be good if the "assigned to" person sets up the question and vote, and sets up a deadline when voting ends. Why put it by default in a template if you don't (that early) hardly know what you are going to vote on? Else there should be a mechanism to review your vote.

Another method we've seen in use is the name - argument table. Easy to add your name and a short opinion comment. Harder to count but better to read.

-- ArthurClemens - 30 Sep 2004

Arthur, I assure you my comment was not specifically related to your (or Peter's) comment here but a more general sense of discussions over past months and even year. Perhaps I should store away bold statements written in the middle of the night to reflect on later. Today I not feeling so strident. I was speaking from months of frustration about not using polls - but the fact is that we are now, so why should I be complaining? I've very excited about experimentation with different forms of user input and basically think it's all good. I'd be real surprised if the quality of discussions suffers but hey, let's keep an eye on it. We can always roll it back or try something else. It's a wiki, right?

Regarding the name-argument tables, I don't think they've been very effective. To my memory, very few people added their input on these. We've already had as good or better rates of participation on the few polls that have been used in just a week or two. This should tell us something.

I think your other points are well taken: initial discussion should take place to develop a well-formed question poll and it would be good to have someone actively manage the use of the poll. If fact, I'm reconsidering my initial position on including polls by default for the same reasons you state.

-- LynnwoodBrown - 01 Oct 2004

Appearing by default? I'm not so sure. Polls can be useful when breaking apparent deadlock, but in the final analysis this isn't a democracy, and polls are likely just to raise frustration levels when a heavily voted-for feature is determined to be anti-mission.

I like people to be able to express their wishes, but a poll actually implies some sort of selection mechanism, where your vote has some real meaning (unless you live in Florida, of course).

I'd rather use DesireMeter. But in the final analysis I don't care as long as the chosen mechanism is used consistently and we can use it to draw some real conclusions.

-- CrawfordCurrie - 01 Oct 2004

Let me start by repeating that I'm thrilled about the new possiblities of polling in TWiki.org. However, now I also am seeing the need for constraint in how we use polls or at least how quickly we use them. Martin - this is specifically a recommendation to you to not be quite so quick in inserting polls into topics. There should first be a process for developing good questions for the poll. This morning, it seems ContributorSurveyOct2004 and PageLayout are examples of the problems of pre-mature jumping to polls. It's analogous to a meeting in which participants are ask to vote on something cold, without any prior discussion. In my experience, this is pretty much always a formula for creating confusion and un-necessary conflict. Consequently, this has prompted me to reverse my vote on the question as formulated in VotingVsDiscussion. This was also an example of poorly formulated question since it is definitely not the same thing as the question presented in the topic summary.

-- LynnwoodBrown - 02 Oct 2004

Hi there. Am glad to be stirring conversation, though I hope this is taken the right way. Let's focus now on what the process should be for developing questions.

-- MartinCleaver - 02 Oct 2004

No worries - I hope you know I'm very aligned with your general intentions with the polls and greatly appreciate you contribution to moving this capability forward. It's helping TWiki make a great step forward. Thanks! Regarding developing a process for use of polls, I've been thinking about how we could develop something along the lines of PollsGoodStyle with suggestions on how to develop and use them. So the question shifts from VotingVsPolls to PollsSupportingDialog. I've actually worked with this a fair amount in real-world group process and am trying to see how it could translate to on-line process.

-- LynnwoodBrown - 02 Oct 2004

Great, Lynnwood! I look forward to your leadership in this area.

-- MartinCleaver - 02 Oct 2004

Very good discussions here. I can resonate with Crawford and Lynnwood.

-- PeterThoeny - 03 Oct 2004