20
$\begingroup$

It has been mentioned before that a set of "standard loopholes" (similar to those used on CodeGolf.SE) could be useful for restricting the number of "silly" answers to questions which were funny the first time, and the second and maybe the third, but after being applied to half a dozen different questions they're starting to get annoying.

While I personally think that the question-asker should think carefully about the phrasing of their question so as to not allow easy get-out clauses, I also think that it's much easier for questions just to state "standard loopholes apply" and feel reasonably happy that they're not going to get any "lol derp" answers.

I'm not suggesting that we stop creative and/or tricky answers being posted (that's half the fun), I appreciate questions which are too open-ended and people come up with all sorts of fun stuff, but there's the flip side where someone takes the riddle too literally (eg. "Can you tell how many ..." being answered with "yes" [I'm not saying I'm not guilty of this myself]).

The question I'm asking is firstly "Do people think this is a good idea?" (feel free to up/downvote as you see fit), and if so, "How do we get a list put together?"

$\endgroup$
  • 12
    $\begingroup$ I don't think it's creative/tricky answers that are the problem; I think the problem has to do with answers that actively ignore the requirements of the question. (i.e. "The oracle is just lying!" or "Just stop imagining" or "Stab the robber with the knife you obviously took with you") $\endgroup$ – Aza Oct 19 '14 at 1:31
  • $\begingroup$ @Emrakul I'd consider all of those to be standard loopholes of a singluar type - assumptions and objects not included in the puzzle itself. $\endgroup$ – Zibbobz Oct 21 '14 at 14:55
  • 1
    $\begingroup$ You do have to be careful about this. Alot of riddle's answers are "find the loophole in this situation". This is why originally questions that invited speculative answers and riddles were always off topic. If you ask a riddle, answers that are valid but not what questioner intended still are valid answers. $\endgroup$ – kaine Oct 24 '14 at 14:23
  • $\begingroup$ Please see this previous discussion in which problems like the prophesized battle one were considered off topic and I provide a (downvoted) dissenting opinion: meta.puzzling.stackexchange.com/questions/227/… $\endgroup$ – kaine Oct 24 '14 at 14:43
8
$\begingroup$

"The oracle is lying"

If a question introduces facts through a party who is implied to be infallible, stating that they are not will almost never be an appropriate answer. This is especially true if OP is the oracle.

Possible exceptions:

  • Determining which statement/oracle is a lie/liar may be an appropriate solution to a problem if several statements by different parties are contradictory.
  • If there is evidence from the rest of the question to imply that this is the OP's (or original author's) intent, it may be appropriate if supported.
$\endgroup$
7
$\begingroup$

One thing that ruins the fun for me is when an answer adds a lot of detail that you wouldn't expect from the answer. For example this answer for Three logicians walk out of a bar.

If somebody asked me that riddle around the office, I gave an answer, but I'm wrong because THAT was the answer then I would absolutely stop playing into any riddles from that person. Answers like this imply that the question was a trick question when it absolutely wasn't meant to be. Trick questions (to me) are more jokes than riddles and my interpretation of this site is that it's not one for jokes.

Enforcement of this will be hard because you have to already know the answer in order to tell if it's a trick question or not, but I don't find these answers to be clever or witty. They come off as pretentious and immature. If you don't have an earnest answer, don't manipulate the puzzle with backstory or insanely specific details so that you can give an answer that wouldn't work otherwise.

</rant>
$\endgroup$
  • $\begingroup$ The good news is that normally, that type of answer is someone being smart for upvotes, not the OP saying that's correct :) $\endgroup$ – Joe Jan 11 '16 at 17:39
  • $\begingroup$ The bad news is that the upvotes do roll in. The answer rises above the rest as if it were legitimate. $\endgroup$ – Corey Ogburn Jan 11 '16 at 18:06
  • 1
    $\begingroup$ Also see "The oracle was lying!" - very similar problem of just adding details to hand-wave parts of a puzzle away. $\endgroup$ – Aza Jan 11 '16 at 18:23
  • $\begingroup$ Ahem... I am the OP of the post linked here, and after rereading my puzzle carefully I must recognize that my phrasing was not the most fortunate; the choice of some relevant details for the puzzle should have been more careful. Rather than that spoil-the-fun answers what I'd really appreciate are suggestions on how to improve my puzzles. After all, this site is about puzzle solving and puzzle creation, so any answer that points out weaknesses in a puzzle and suggestions on how to improve it should be welcome IMHO. Glad that you raised the concern. $\endgroup$ – GOTO 0 Jan 11 '16 at 21:27
  • 1
    $\begingroup$ @GOTO0 I don't blame the asker. I think it's unreasonable for every puzzle to be ironclad against non-sense answers. There might have been particular phrasings that could have prevented it from happening to your question, but ultimately it's on the users who submit these types of answers. $\endgroup$ – Corey Ogburn Jan 11 '16 at 21:31
  • $\begingroup$ The irony with the linked answer is that it got most of its upvotes in its original, pre-edit form, i.e., a much less "jokey" and still accurate answer. $\endgroup$ – question_asker Jan 12 '16 at 19:12
2
$\begingroup$

Perhaps a new tag is in order for this? While kaine's comment above is true (sometimes answers that weren't the intended answer are still valid), I think that's only true for a certain type of puzzle.

If you ask, "What am I?" and give a number of descriptive lines, it is fair and accurate that someone could come up with a non-intended solution that fits the puzzle fairly well. On the other hand, if you pose a question where the obvious answer is to find the solution to a puzzle that has been posed, it's annoying for the asker to have to add an entire paragraph explicitly excluding all the smart-ass answers people might come up with.

In the example linked above, the puzzle could just be presented without any story, as a question, "Find the path through this grid," but I feel like that would detract from it somewhat. Unfortunately, by spinning a story around the puzzle, the author is also required to specifically exclude every smart-ass answer that might come up.

When I read that question, it seems obvious to me that there is an intended answer. Anyone who answers differently is just trying to be "cute", and is not legitimately providing a different and potentially better answer to the puzzle (unless, of course, there is actually more than one way of solving the actual puzzle presented, outside of the story context).

Thus a tag (or perhaps something worded a little more kindly, like ) seems to me like it would fit very well. If it got applied to riddle-type questions that could legitimately have alternate answers, we could decided whether to remove it, or respect it, or have an alternate strategy.

If, for some reason, it is felt that a tag is not appropriate for this, then the CodeGolf.SE method of just stating in the question that no loopholes are allowed should be sufficient.

$\endgroup$
1
$\begingroup$

I am merely a casual observer of this site, and just created this account to post this, but perhaps a tag or disclaimer at the end of the question to define all words used based on common usage or conversational use would be helpful.

There seems to be an ongoing disagreement between some users about whether or not logicians are just logical or also literal (like the discussion here, for one example). I don't know which should be the default, but perhaps a tag or something of that sort could be employed by anyone not intentionally including language-tricks in their puzzles.

$\endgroup$

You must log in to answer this question.

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