Content on site slowly becoming more and more like Y! Answers

Question

We're treading down a wrong path here, there needs to be some quality control in terms of answers. Just treating everything as "it's creative, don't hate" is such a bad way to regulate question and answer quality.

For example, this question: https://puzzling.stackexchange.com/questions/3911/a-man-pushes-his-car

An answer like this: https://puzzling.stackexchange.com/a/3925/4425 reads just like a Y! Answer. Completely random, made up. Sure it's a lateral thinking puzzle, but that's just a bad excuse for "anything goes in this question".

Is this just me?

Answer 1

I blame the questions.

If “solve this riddle” is a valid question, then those are decent answers.

These puzzles are officially off-topic, but pretty much nobody appears to be enforcing this.

Answer 2

One reason for this is the way the 'hotness' scores are calculated.

Questions with high 'hotness' can reach the SE home page and the multi-collider-thingy, which gives them a big boost in visitors and rep.

The hotness formula gives a big advantage to questions with plenty of answers, specifically MIN(AnswerCount, 10); see How do the "arbitrary hotness points" work on the new Stack Exchange home page?

Puzzles which are more likely to get lots of answers are those which (IMO) are:

• easy

• open-ended.

Puzzles which are less likely to get lots of answers and therefore less likely to reach the supercollider are those which (IMO) are:

• hard (which is a good thing)

• clearly defined (which is a good thing).

The hotness formula also has a time element - it favours new posts over old ones - so the net effect is to favour questions which get a lot of answers more quickly than other questions. Which it seems to me is exactly what we don't want. A puzzle that lots of people contribute different answers to without taking any thinking time at all is perhaps not a puzzle with much depth to it.

It seems (if I understand the formula correctly) that questions which don't yet have any answers have a hotness score of 0. That's a real shame because the unsolved questions are some of our most interesting ones.

I had that wrong, was looking at an old version of the formula. The current formula is here: What formula should be used to determine "hot" questions? "This algorithm will heavily favor questions with LOTS of answers"

Just an observation.

---

I'd also like to point out that there's two different ways in which one can improve the average quality of content:

• One can criticise other people's contributions which one perceives as poor quality, in the hope that they will either stop contributing or improve the standard of their content,

or

• One can 'raise the batting average' by contributing high-quality content oneself.

thinly, I notice that (so far) you have contributed 0 answers and 0 questions to puzzling.SE.

I'm curious as to what it is about the poor quality you perceive in other people's contributions which prevents you from making any contribution at all?

Are you worried that your quality content will be overlooked, lost amid everyone else's content? Or that it won't be rewarded with the upvotes it deserves? If so then some adjustments to the site-specific hotness formula might help with that.

Answer 3

It's not just you. The problems you've mentioned here are the main reason I stopped participating in this site several months ago. I have no intent to return in the future unless they are worked out (which I'm not hopeful of). I decided to describe what I expected this site would be and compare to what it actually became; while this isn't entirely relevant to the question I think it may provide some useful context.

Early on, based on the definition questions, I thought the questions this site would entertain would not be so much individual "puzzles" as trying to understand the theory and practice behind them. By puzzles, I mean specifically things like the following (not exhaustive):

• Rubik's cubes and other twisty/mechanical puzzles
• Sudoku and other similar puzzles
• Crossword puzzles, word searches, cryptograms, and other linguistic puzzles

And questions on the following aspects of these:

• The strategy behind solving a particular puzzle
• Mathematical theory of the puzzle
• "Practical" questions like e.g. how best to modify a Rubik's cube for speedcubing
• The history of various puzzles
• Very few (if any) questions of the nature "Solve this puzzle: [description]"

To be honest, I was almost exclusively interested in twisty puzzles, mostly for theoretically oriented questions. I have quite a bit of knowledge about higher dimensional analogues of the Rubik's cube, for example. I would not have minded sharing the site with the other topics above though, which are similar in nature. During private beta there were some signs that we were going in a good direction. Sure, the overall volume would have been low if we had gone in this way, but it would have been a high quality site.

Instead, a few new types of questions emerged which have basically completely took over the site. I'm specifically referring to logic-puzzle, brainteaser, riddle, word-problem, lateral-thinking, and other similar types of questions. These are very much "Solve this puzzle" type questions, and they use a very broad definition of the term "puzzle" which was not apparent to me at all when we were in definition and commitment phase. It seems that "puzzle" here means essentially any question (on any topic) which is cryptic enough to not be obvious, but not so cryptic that there are multiple answers, none of which is obviously correct (at which point it becomes "speculative"). So "what did I have for lunch today?" might very well be a puzzle if I chose to include a specific set of facts to make it into one. I find this simultaneously incredibly broad and very confusing. But at any rate, it's pretty clearly not the vision that I had for this site or where I expected we'd go from our Area 51 definition, and most of the questions I want to ask don't seem to fit in at all any more.

Given that this community accepts, and even encourages (by measure of voting), questions like this, it's not surprising how the site has turned out. There's little to no formal expertise or training involved in solving riddles that I know of. What sets Stack Exchange sites apart from Yahoo Answers and other Q&A sites is that the communities here are supposed to be experts on the topic. Without that distinction, there's really no reason to expect the quality here to stay high; it'll naturally go to the same level as most everything else on the internet of people talking about things they don't really know or care about very much. That's what you're seeing happen here.

Rather than trying to stick around and fix the site, I decided just to leave. I'm not interested in the direction this site is heading, and I'm not invested enough in this site's success to care if it fails. For the questions I do care about, many of them can go on Math SE or MathOverflow. Speedcubing questions still have no good home in the network, which is a shame, but there are other sites dedicated to it outside the network and I don't see any indication that such a question would be any better here than e.g. on Sports SE. For anyone who is still around and agrees that this needs to improve, I wish you the best of luck, but I think it may be too little too late to turn this site around and I have no idea how to do it.

Answer 4

That is actually a good answer by StackExchange standards. The answerer provided a link to a reputable source (Wikipedia) as well as summarizing the condition in his own words. Then he explained how it answered the question.

There's some debate here as to whether we want to include well-known puzzles like these that are easily found on the web. That answer is a great reason for why we should. Most people may know (or can easily find) the classic answer, but that doesn't mean that new answers can't be discovered. Finding new answers to classic puzzles is a great way of adding value to the site and also growing and improving the art of puzzle solving.