Official Sticky Polls for dot upgrade issues?

[mkrishnan]

Hi everyone,

I hope this hasn't been discussed in the too recent past, and if it's a bad idea, fine.... but, do you think that it would be good if we had a simple sticky poll of the form:

- 10.4.9 installed without problems
- The 10.4.9 upgrade caused some kind of system problem

Just so we can address the question broadly of what user experiences are? This way, once we do this for a couple of dot upgrades, say, 10.5.1 and 10.5.2, then we will have an idea of the expected base rate of botched installs. If that rate turns out to be, say, 10%, and then 10.5.3 has a botch report of 40%, then we have a real objective indicator that one upgrade is more problematic than others.

I think, if we make it an official, stickied poll, and we put it in the MacRumors announcement, we can easily get hundreds or even thousands of votes, and the statistic we provide will be really useful to the community. But it has to be a single, simple poll with only a couple of options... like the kind of polls we do with MacPolls.

 

[Teh Don Ditty]

I like it. It would seem to cut down on the numeous threads about problems with the current dot upgrade and put them in a centralized thread.

 

[fivetoadsloth]

I like it. It would seem to cut down on the numeous threads about problems with the current dot upgrade and put them in a centralized thread.

I agree, iThink this would be nice.

 

[Doctor Q]

That's a thoughtful suggestion, mkrishnan.

Thinking ahead...

Suppose we had such a poll and said something like this after an update was released:

Once you have installed this update, please vote in our Mac OS X 10.n.n Update Poll to let others know whether or not you had problems during or after the installation.​

How could we make it best serve forum readers?

* Would the poll have its own discussion thread, for details about the issues people noted? Or would all discussion be in the news thread?

* Would we permit "no problems" comments in the discussion thread, or would we ask the moderators to remove such posts?

* How likely is it that a poll would be weighted toward the negative simply because people with problems are more likely to vote? That's already a consideration in a discussion thread, but would a poll, with its bar graph, make it even more likely that people will get a misleading negative impression? Would removing "no problems" posts have the same problem?

 

[nbs2]

* Would the poll have its own discussion thread, for details about the issues people noted? Or would all discussion be in the news thread?

From my understanding of the idea, it would be ideal for discussion to be contained in the thread. There are normally so many threads, it is tough to figure out what issues in particular people are facing. If the mods would be so kind as to toss any new threads into the grandfather, that would reduce clutter. As a counterpoint, the thread would explode with the sheer quantity of posts - making that thread tougher to read.

* Would we permit "no problems" comments in the discussion thread, or would we ask the moderators to remove such posts?

Delete the no problem posts, but keep the posts that point out some benefit felt from the upgrade would be useful. And any post containing the word "snapier" would self-destruct.

* How likely is it that a poll would be weighted toward the negative simply because people with problems are more likely to vote? That's already a consideration in a discussion thread, but would a poll, with its bar graph, make it even more likely that people will get a misleading negative impression? Would removing "no problems" posts have the same problem?

That is a risk, and unfortunately we have no carrot (nor stick) that would spur a true sample. I would think that voting being easier than posting would have some effect, but the real world has shown otherwise. I don't think removing the "no problem" posts would imply negativity, so long as we could be reasonably sure that people without problems were posting. In the end, though, as long as there isn't a random spike (either up or down) in the percentage of "no problem"s that vote, a relative comparison by savvy users would compensate for the overall negative perception.

 

[mkrishnan]

I think those are all valid points, but I think that, actually, with the bar graph, there's also the chance for the masses who are having good results, especially if they are encouraged to do so, to vote, so that actually, the number may end up being 7% when the thread makes it look like the number is 50%. (my suspicion, probably with all of you, is that the rate of botched upgrades is very low).

 

[thejadedmonkey]

but after a few times, we'd see a pattern. Say 35% of users who vote having issues is "normal", but if it spikes to 50 or 60%, we have a problem, and 15% would be low. Measure the results against themselves, and it'll be more accurate then what happens after every point upgrade.

 

[TheAnswer]

The best implementation that I can think of would be some kind of simple poll on the front page as a separate story, with a range of options from No Problems to Severe Problems (maybe with examples of each below the choices)...that increases the voting % of people with no problems.

Under the poll should be the link to the thread similar to "Those having problems with their upgrade to 10.X.X please search the following thread for links to existing threads about problems similar to yours. If your problem is not addressed in any existing threads, feel free to create a new thread and please post a link to your new thread in the main 10.x.x problem thread."

This would minimize the number of similar threads created dealing with the same issue...also, having it as a story would make it easier for users to find later if someone is upgrading an older machine...they could find the story and then have the links to the problem threads at hand.

 

[mkrishnan]

Reviving this idea and also adding to it...

It would be really interesting if we came up with a few standardized polls for major Apple products that addressed our community's product quality experience. There's always the concern of multiple voting or voters who don't really meet the requirements (e.g. don't own the product), but it would be very helpful for situations such as... a new Apple product or revision is rolling out and there is a high defect rate scare. We could provide data on whether this appears to be true or not.

I would be willing to volunteer some effort to this.