15
\$\begingroup\$

This was prompted by Mid-to-late 1970's Space Based RPG: what was it?, a question I've put on hold briefly so we can assess clean-up. The question needs some refinement (and this question will be updated later to provide guidance on what's useful to specify in an ID question), but the most alarming development was that every answer was basically just this:

I think it's {game name}.

{cover image}

(One answer mentions one feature in common; another answer mentions multiple.)

Another problem here is all of those answers are upvoted multiple times. Two of them have one downvote each. This isn't good. This is really bad. (To the community's credit, two answers were downvoted and deleted which were even worse than the quote above.)

If I sound alarmed, I am: answers not meeting basic quality expectations, and voters not enforcing and demanding those expectations, was what lead to the community voting that game system recommendations was no longer a viable topic here. I really want to see product identifications continue to have high quality expectations for answers and from voting and not head in the same direction.

What we need from answers to identification requests

When answering a product identification request, we need you to express why you think the identification you found is the appropriate one. You did the research effort, and you should be explaining to us what is with this product such that we can be confident you're right (or wrong). One of our post notices (called "insufficient explanation" in our mod tools) says this:

We're looking for long answers that provide some explanation and context. Don't just give a one-line answer; explain why your answer is right, ideally with citations. Answers that don't include explanations may be removed.

We can't go by the gut feeling or guess of some random person on the internet saying "I think it's this." We need you to give us confidence in what you're saying being the right thing.

Voters: Please downvote answers that have no explanation.

Dear voters: the entire community depends on you to enforce these expectations. If an answer doesn't include evidence and a convincing argument inside that answer that makes you think the answer might be right, downvote it. Request they improve, request they provide explanation and an argumentation case that can demonstrate this is the right product.

Please do this even if you think it's the right product from your personal knowledge. We need answers to meet a default quality bar, and "I think it's this" is not past that quality bar, nor anywhere near it. (In fact it's so not-above-our-quality-bar that if this were an Olympic event, those answers would be a misinformed athlete doing limbo at the high jump qualifiers.)

| |
\$\endgroup\$
  • \$\begingroup\$ Thanks for this meta. I edited the question this morning after Mxy called out problem answers. Am pleased to edit the question again so that it improves. My problem: after an internet search tripped over a bunch of false leads, I had limited detailed memory. When I asked our old GM he came back with "Empire of the Petal Throne" which was a game I ran, not him, for our group for about three years. (arrghh, getting old seems to have struck us both). All I had was what I put into the question. Thankfully, someone provided a link to an RPG timeline that I hope helped with scope. \$\endgroup\$ – KorvinStarmast Aug 10 '17 at 17:40
  • \$\begingroup\$ I'll put this here instead of on main site. Your call out to me to provide more detail is unfortunately part of the problem: if I had more detail, I'd have it in the question, and I'd have had more search ammo for my own original search and maybe would never have asked the question. The depth of RPG experience, through time, is IMO one of our core expertise features here, which is what I was hoping to tap into. And I think it worked. This is not system rec, this is product identification. I don't think conflating the two is a reasonable approach to take. \$\endgroup\$ – KorvinStarmast Aug 10 '17 at 17:55
  • 2
    \$\begingroup\$ Arqade used to do ID questions, and their higher volume of traffic accelerated their experience with them. They deemed them off topic when based on only memory, precisely because relying on memory can lead to questions that can only be answered with guessing. They struggled with what to do, and eventually banned them, because of the exact problems we're having here. We're farther behind their path, and I hope we can find a better solution, but “can't be helped, memory is fallible” is more of an indictment of the question type than a defence. \$\endgroup\$ – SevenSidedDie Aug 10 '17 at 18:12
  • \$\begingroup\$ @SevenSidedDie I'll accept that it's an indictment of my question, and my memory, which through workshopping was improved thanks to the community(like a lot of questions are) but not of the question type. If you look at the edit history I think you'll see where better scoping was made after the initial question. I agree that it's potentially difficult if the one asking the question Does Not Stay Engaged. For querents who do, I don't see a problem. Prod ID taps into a unique resource of Experts on this site: exactly what I was hoping to tap into. \$\endgroup\$ – KorvinStarmast Aug 10 '17 at 18:54
  • 1
    \$\begingroup\$ I'm hoping we might be able to identify some things that ID questions work well by having, so that we can give askers pointers of "include some/all of these things." Currently not able to work on that though. \$\endgroup\$ – doppelgreener Aug 10 '17 at 18:56
  • \$\begingroup\$ @SevenSidedDie Do you have a clear outline, or even a few examples, of what you think makes a good Product ID question, versus a bad one? That might be helpful, rather than advocating a throwing out of the baby with the bathwater. \$\endgroup\$ – KorvinStarmast Aug 10 '17 at 18:58
  • 1
    \$\begingroup\$ This definitely isn't a bathwater tossing situation, since this is I think the first time I've seen an ID question go awry like this (which just makes it extra alarming that it did). I don't think SSD was advocating that! Just conveying or cautioning what another site did. It's useful to learn what other sites decided they needed to do, though. \$\endgroup\$ – doppelgreener Aug 10 '17 at 19:01
  • \$\begingroup\$ @KorvinStarmast I don't — I found myself recently observing that there's nothing in questions that tends to distinguish ones that get answered with guesses and ones that don't, only external accident of how well reality lines up with the details in the memory. That means there's no objective pre-answers criteria (that I see) for separating answerable ID questions from only-guessable ones. That leaves us divining the difference by observing the answering patterns. \$\endgroup\$ – SevenSidedDie Aug 10 '17 at 19:06
  • \$\begingroup\$ But, I also observed recently that we have a pretty good success rate (I think?), and a low-enough proportion of ID questions among our overall questions, that I think they're not creating a difficult absolute number of guessable-only questions. Arqade had a flood of memory ID questions, so they had to deal with it; if our ID questions have some issues, there are few enough ID questions here that we can likely let them coast along doing so-so without considering drastic responses (to our non-drastic situation). \$\endgroup\$ – SevenSidedDie Aug 10 '17 at 19:08
  • \$\begingroup\$ @SevenSidedDie OK, and I am always open to improving the few questions I do ask. I did provide some criteria based on when the game was published, which some of the answerers did not seem to take note of. We can't control that with the topic of a question, but we may be able to influence it by writing a better question with better criteria. (So that's mea culpa on me for doing a less than stellar job at that). \$\endgroup\$ – KorvinStarmast Aug 10 '17 at 19:11
  • 4
    \$\begingroup\$ I actually think the criteria you have in your question are OK. You even go so far as to specifically name a weapon you remember from the game. For me the issue is with the quality of answers, with people basically randomly suggesting any sci-fi game they remember that might or might not be from the time. \$\endgroup\$ – Wibbs Aug 10 '17 at 19:13
  • \$\begingroup\$ @KorvinStarmast I don't think there's any blame there — if we accept memory-based ID questions (and we do), then it will simply be the case that some are harder to answer than others, some perhaps unanswerable. I don't think there's any way we can develop a clear (or even fuzzy?) line that would let us say “must be this well-remembered to ride”, because how can we tell how well-matched and uniquely-defined the memory is to the field of RPGs, except by waiting for the answers to try? \$\endgroup\$ – SevenSidedDie Aug 10 '17 at 19:14
  • \$\begingroup\$ @SevenSidedDie I am pretty sure, based on one more comment that mentions a Trilax race(he triggered a memory), that Space Quest is it. As I said to doppel I want to award a bounty, so I am waiting for the time to pass so that I can. \$\endgroup\$ – KorvinStarmast Aug 10 '17 at 19:15
  • \$\begingroup\$ @Wibbs Yeah, this meta is about the answer quality. \$\endgroup\$ – doppelgreener Aug 10 '17 at 19:54
5
\$\begingroup\$

There is nothing alarming about bad answers; they happen. We have tools to deal with them.

There is apparently a need for better guidance on how to answer a product identification question, and quite possibly better guidance on how to ask a product identification question. Having a meta with guidance on both (or one each) seems a good idea.

The proposed guidance proffered in this meta question, as regards answering the product identification answers, is so-far-so-good. I move that we adopt the following elements of the proposed guidance as something formal, or perhaps something added to the tag, or a bit of both. This can be workshopped.

When answering a product identification request, express why you think the identification you found is the appropriate one.

  • You did the research
  • you explain why this product is what the question is looking for

"Insufficient explanation" (as @doppelgreener cites from the mod tools) says this:

We're looking for long answers that provide some explanation and context. Don't just give a one-line answer; explain why your answer is right, ideally with citations. Answers that don't include explanations may be removed.

That, if it appears on your answer, is a flag to you. Give the answer more explanation or better support.

Guessing can be expected to attract down votes, and question deletion, because this site aspires to a high signal to noise ratio. A bunch of guesswork isn't high signal.

Anyway, that's my recommended way to provide some guidance.

| |
\$\endgroup\$
  • 2
    \$\begingroup\$ I guess, it's not bad answers being alarming, it is the voting on bad answers that is alarming: they are not downvoted into Oblivion, but rather have a bunch of upvotes. \$\endgroup\$ – Baskakov_Dmitriy Aug 12 '17 at 22:28
  • \$\begingroup\$ Baskakov_Dmitriy Alarming? What if some people found those answers helpful? (Yeah, I understand what you are driving at, but you can't control how people vote). \$\endgroup\$ – KorvinStarmast Aug 13 '17 at 2:08
  • 2
    \$\begingroup\$ Well, that's exactly the problem. People found helpful answers that are not helpful, or at least aren't didn't show why are they helpful (if I understand @doppelgreener right, they mean "Did not show it at all, were one-line answers". And that's a very bad signal. \$\endgroup\$ – Baskakov_Dmitriy Aug 13 '17 at 7:51
  • \$\begingroup\$ @Baskakov_Dmitriy You got the second part right. "didn't show why they were helpful." Whomever voted for them may have still found them helpful. Voters are not required to say why an answer is helpful. The answerers were given that reminder about doing a better job at explaining, at making the answer fit the SE standard better. Why you are taking voters to task for sending a bad signal I don't understand. The signal doppel sent to the answerers was (I thought) productive: it didn't just correct them, it sent/reaffirmed the message to all on what is preferred in answers. \$\endgroup\$ – KorvinStarmast Aug 13 '17 at 11:21
  • 2
    \$\begingroup\$ The point @Baskakov_Dmitriy makes is fair actually. When voting is behaving oddly and way beyond just “people can vote how they like”, that is a signal that there is something wrong happening. Voting has a well-defined purpose: to sort the stack of things (hence the name Stack Exchange) into a quality order. When we can see that is not happening and very unlikely to self-correct, we have an objectively-visible failure of the voting system and it's time for humans to conduct meta oversight and investigation. (We can't control how people vote, but we are mandated to control what they vote on.) \$\endgroup\$ – SevenSidedDie Aug 15 '17 at 17:11
1
\$\begingroup\$

This problem could be solved by ok-ing the deletion/flagging for deletion of ID answers that don't sufficiently support themselves. Upvotes shouldn't prevent deletion-- we know upvotes don't track utility on answers like these.

| |
\$\endgroup\$
  • 1
    \$\begingroup\$ I'm concerned about considering this a solution since it is more of a bandaid: moderators having to do this as correction in spite of community behaviour (as opposed to alongside it) was a source of trouble for game recs, and I'd rather find a sustainable way to make them work with community cooperation. We could do it, but we ought not call that a day. \$\endgroup\$ – doppelgreener Aug 20 '17 at 17:03
  • \$\begingroup\$ @doppelgreener that's fair \$\endgroup\$ – Please stop being evil Aug 20 '17 at 17:53

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .