Ten fold - 2016-06-15

« first day (2129 days earlier) ← previous day next day → last day (2872 days later) »

11:03 AM

@Glen_b Somewhat in reply to your remark about most common regression misconceptions thread. I noticed a comment by somebody who is new to CV (rep just above 100) but has huge experience on SO, Meta.SE, and Programmers (over 100k on each of these three sites) -- they are genuinely surprised that we are using CW status for such threads: stats.stackexchange.com/questions/218156/#comment413514_218156. That has been precisely my point for a long time... CC to @gung.

11:24 AM

@amoeba We don't use CW quite the way SO does, for sure, but quite a few SE sites do things quite differently to SO (meta.SE and Programmers both tend to be closer to SO than we are, unsurprisingly).

There may be an argument for changing how we use CW, but "SO does it this way" doesn't necessarily mean it will suit us. They do lots of things there that I'd hate to see here.

I don't have any particular axe to grind with CW -- if there's some consensus that CW should be applied differently, that's okay by me as long as it's clear.

1 hour later…

12:42 PM

@amoeba Although it might not be clear, he was trying to suggest (I believe) that "marginal, broadly-asked big list questions" ought to be summarily closed rather than made CW.

1:11 PM

@whuber: Hmm, I did not think of such an interpretation of his comment. @RobertHarvey, can you please clarify what you meant in your comment: were you surprised that this thread was made CW rather than staying non-CW or rather than being closed? (I am not sure my ping will reach Robert as he has never used our chat room. I might later leave a comment to draw his attention, if the ping does not work)

1:27 PM

@amoeba You're correct, that ping won't work.

Stephan Kolassa

1:42 PM

@amoeba Considering that Robert spoke up in a comment on this particular question and not in a Meta post or here in chat, I'd interpret him as describing this particular question as a "marginal, broadly-asked big list question".

And here, I couldn't disagree more. Yes, it's broadly asked. Yes, it's a big list question. But it most emphatically is not marginal. These are not synonyms.

And right now, it seems like 49 upvoters and 27 people who favorited the question agree.

We can certainly debate about whether CW is useful here (I agree with @amoeba meta.stats.stackexchange.com/a/2957/1352 that it's not).

So we seem to have at least two separate discussions here.

I'd be interested in Robert clarifying his thoughts.

1:57 PM

@Stephan, I think Robert commented there because he noticed that Q in the hot network posts. He does not seem to be usually following CV or our Meta. In any case, I pinged him via a comment so he will hopefully clarify.

(And yes, I agree with you that this Q is on topic and should stay open.)

Stephan Kolassa

@amoeba Given his rep, I assume that he is familiar with meta and chat, so if he had been non-specifically surprised at our CW practices, I would have assumed that he'd use one of those venues. But that's just tea leaves reading on my part.

2:23 PM

@StephanKolassa If it's not marginal, it shouldn't need CW.

Stephan Kolassa

@RobertHarvey Hi Robert, nice to have you here!

Hello.

Stephan Kolassa

Could you elaborate on @amoeba's question above? chat.stackexchange.com/transcript/message/30374758#30374758

You can be on-topic, but still too broad. For reasons which are somewhat historical on Programmers, we don't accept Big List questions of any kind. But I do know of a few sites that have some loopholes; CS is one such site. I think they might even have a Big List tag. But I'm also pretty sure they don't use CW.

As to being "surprised," nothing really surprises me on SE anymore.

Stephan Kolassa

@RobertHarvey So for this particular question

49

Q: What are some of the most common misconceptions about linear regression?

I'm curious, for those of you who have extensive experience collaborating with other researchers, what are some of the most common misconceptions about linear regression that you encounter? I think can be a useful exercise to think about common misconceptions ahead of time in order to A...

regression multiple-regression

Would your reaction have been (a) to close it or (b) to leave it open non-CW?

2:36 PM

That question would have been "Too Broad" on programmers. We would have asked "Can you make your question more specific? Focus it on a specific, software design-related problem you are having."

Stephan Kolassa

Thanks. This seems like a case where usage simply differs between SE sites. I don't think this is too broad for CV (and apparently, nobody else does, either - no downvotes, no votes for closure).

@Stephan, I think what @Robert is saying is that each SE forum can decide what they are doing with big-list / opinion-based questions, but the more reasonable possible policies are either to leave them open non-CW, or to have them closed. As he wrote in the initial comment, CW status is sort of deprecated.

For the record, I have no problem with these kinds of questions. CV appears to have turned that question into an excellent resource. But for every one of these questions that produces a great post on Programmers, there are 99 others that produce nothing but bikeshedding.

Stephan Kolassa

@RobertHarvey: no argument from me there. We do close quite a few questions as "too broad".

Our crown jewel on Programmers... It's probably the only Big List question that ever grew up and became something good:

2189

Q: What technical details should a programmer of a web application consider before making the site public?

What things should a programmer implementing the technical details of a web application consider before making the site public? If Jeff Atwood can forget about HttpOnly cookies, sitemaps, and cross-site request forgeries all in the same site, what important thing could I be forgetting as well? ...

web-development

And now it's used as evidence by new users that they can ask any Big List question they want.

It has 88 answers posted to it. Only one is now visible. It was heavily curated, and took a lot of work to get it into its present form.

Joel Coohorn made it CW the moment he posted it, back when you could still do that.

It was closed and reopened twice, locked a couple of times, and even deleted once.

Stephan Kolassa

2:49 PM

SO has a number of highly-voted questions that remain but are closed and carry a big sticker "this stays for historical reasons, but questions like this are NOT welcome any more" or similar. Wouldn't something like that solve your problem of people using this question as an excuse?

It has a "collaborative lock" on it. But yes, we probably should put something prominently at the top of the question.

The real problem with Programmers is not CW Big List questions; it is that first time users never see the Help Center, and even if they did, the wrong information is there.

91

Q: Usability issues for first-time Stack Exchange users - a micro-study

I come from lands afar, also known as Hardware Recommendations Stack Exchange. Over there, we have a serious problem with first-time Stack Exchange users coming along and asking off-topic questions - to the extent that 50% of incoming questions are closed. We've had a few attempts at solving this...

discussion asking-questions new-users

Programmers has a terrible problem with off-topic questions. We close about 40 percent of all new questions asked there. But the small contingent of close voters left there is greatly outnumbered by new users.

SE has finally recognized that there is a problem, and has asked us to propose site name and Help Center changes. This is my proposal:

86

A: New Site Name and Scope Proposals

Software Engineering. I think it's as good a site title as you're going to get. Here's why: It's not "Programmers," which smacks of people who write teh codez. Electrical Engineering's name change (from Electronics) was very good for them. It raised the quality bar on that site significantly...

Anyway, that's probably more information than you wanted.

Stephan Kolassa

Well, it does clarify where you come from.

I don't think we here have major problems with big list questions.

Our problems are more with "unclear what you're asking", plus driveby users that never reply to requests for clarification.

@StephanKolassa Closely related to that, we have an exponentially growing problem with awful questions--homework dumps, gibberish, inability to articulate a clear question. These have reached about 15% of the total (up from next to zero a few years ago). The acceleration shows no sign of letting up.

Stephan Kolassa

3:04 PM

@whuber Agreed. I have wondered whether we need a new closure reason. "This question is simply awful."

2

SE really needs a fast-track way to dispose of such questions.

@StephanKolassa :-) If you could find a tactful way to restate that, along with guidance for making the question un-awful (or making it go away), that would indeed be useful.

You don't need more close reasons. You need a hammer. I wonder what would happen if we put a diamond next to some people's names, and gave them a binding vote... Oh, wait.

Stephan Kolassa

@whuber Ah, but now you are talking about our "Unclear what you are asking" closure reason, which comes with explicit recommendations to edit the question to clarify.

@RobertHarvey: I have noticed a few of our mods commenting on unclear questions and requesting clarification. It seems like less than 10% of such questions are ever clarified. The rest just scrolls off the landing page and remains unclarified, unclosed, unvoted upon and generally zombiefied.

What we may need is a time-delayed binding closure vote by mods. "Please clarify within 24 hours, otherwise the question will be auto-closed." (Or a bigger hammer: auto-deleted.)

@StephanKolassa I confess I'm guilty of some of that commenting-but-not-closing behavior. I do try to follow up later, but I haven't been doing it systematically.

@StephanKolassa The time-delayed closure idea is a really good one!

Stephan Kolassa

3:16 PM

@whuber You shouldn't need to remember to check back.

If you have two unbiased coins, what's the probability of getting both heads on the nth round, given that all previous rounds failed? How do you work it out? I thought it was a binomial distribution, but that allows for success on previous rounds.

Stephan Kolassa

Right now, you need to rely on lowly non-mods to VTC, so the offending Q stays in the review queue. The problem is that when I see such a request for clarification from a mod, I am very reluctant to VTC, because action has, in a way, already been taken. I assume others think the same. Result: nobody votes to close, and unless the OP does clarify, the question zombiefies.

@noahnu: if your coins are fair, then they don't have a memory, so the chance on the nth round (no matter what happened before) should be the same as on the first round: 0.25. (Am I misunderstanding you?)

@StephanKolassa Okay, that's what I thought. Thanks.

Btw, the "fair" refers to the 0.5 probability of heads/tails.

@StephanKolassa I would suggest interpreting that situation differently: if a mod has made comments and you still think the question should be closed, probably the mod did too, but was a little unsure and was looking for some reinforcing behavior from the community. So send us a message and vote your opinion.

Stephan Kolassa

@whuber That makes sense. Thanks, then I'll do so. I foresee a lot of closure votes in my near future.

3:22 PM

@noahnu As a practical matter (as opposed to a purely mathematical one), if $n$ is sufficiently large I would question the assumption that the coins are unbiased (as well as other implicit assumptions, such as that the flips are independent).

Stephan Kolassa

3:37 PM

Related meta discussion:

23

Q: Are we closing questions too fast?

It seems to me that we have been closing more and more questions, and that we are moving them into the close votes review queue very quickly. This seems to have evolved naturally, but I wonder if we have moved past the optimum point without anyone stopping to think about whether we are moving in...

discussion closing-questions

Where I fully agree with @NickCox's answer: meta.stats.stackexchange.com/a/3127/1352

4:11 PM

Sheldon Cooper(<-- a character) in a TV show: "The Big Bang Theory" says:

"In the game of rock paper scissors, players familiar with each other will tie 75-80% of the time due to the limited number of outcomes"

Is this in any way correct or close to correct, or at least provable?

Sounds like a decent question for the main site. But I'm not a regular here, and so I may not be fully cognizant of the site scope.

Stephan Kolassa

4:41 PM

@Dawny33 I don't see how this could be correct, unless players are too incompetent to notice patterns in the opponent's choices. The best strategy is to pick at random. (Any patterns can be detected and exploited in repeated play.) Then you will get 1/3 wins, 1/3 draws, 1/3 losses.

(I'd be careful of statistical advice given by physicists. Just as I'd be careful of physical advice given by statisticians. Particularly statisticians that are in fact characters on TV shows.)

2

5:04 PM

@Stephan Haha about picking at random! You can try out how random you can be here people.ischool.berkeley.edu/~nick/aaronson-oracle -- it's very neat, you should press F and D randomly and the program predicts what you will press. After you play it for 30-60 s, it's very difficult to bring the accuracy below 60%.

Stephan Kolassa

@amoeba When I play rock-paper-scissors, I bring along a Mersenne Twister, or a one-time pad of cosmic radiation-generated random numbers. What do you do, sir?

2

:-)

Stephan Kolassa

Current accuracy is 49%.

And all it took was:

while ( TRUE ) cat(ifelse(runif(1)<0.5,"F","D"),"\n")

5:52 PM

@StephanKolassa I play this card came called Netrunner. I actually do bring a one time pad of random numbers to games.

Android: Netrunner

Android: Netrunner is a Living Card Game (LCG) produced by Fantasy Flight Games. It is a two-player game set in the dystopian future of the Android universe. Each game is played as a battle between a megacorporation and a Hacker (computer security) ("runner") in a duel to take control of data. It is based on Richard Garfield's Netrunner collectible card game, produced by Wizards of the Coast in 1996. == Gameplay == Like the original, the game is asymmetric and involves two players, one playing the Runner, and the other playing a Corporation. The runner wins by hacking into the corporation...

Using "free will" I was able to keep it below 55% consistently. It required pressing a lot of runs of three or four. I was able to do better--45% to 50%--by following a simple pattern in which I rotated through runs of "d" and "f" from lengths of 1 through 5. Fibonacci numbers (modulo 5) worked very well, too.

Stephan Kolassa

My little loop above gave me a run of 6 F, then 6 D. No human would do this "at random".

Not unless they had studied probability :-)

1 hour later…

7:09 PM

After many years of helping our site, especially through extensive edits and brilliant responses, @mbq and @cardinal have been promoted to moderators emeriti.

5

Although the diamonds no longer appear next to their names, I hope they will continue to participate when they can. They will always be welcome and their voices will for ever carry great weight for me and, I am sure, for many others.

4

Stephan Kolassa

@whuber: thank you. Will you be posting this on meta? This will likely get more attention there. (As well as the upvotes @mbq and @cardinal deserve.)

Is there any way to see which users once were mods but are not anymore?

8:05 PM

@amoeba Probably at a nearby pub

8:39 PM

@amoeba One way is to look at the sheriff badge recipients.

8:51 PM

@StephanKolassa You're right. The challenge is doing so in the requisite Q&A format. With @amoeba's help, I think I found a suitable way.

0

Q: How Can I Discover Former Moderators?

After many years of helping our site, especially through extensive edits and brilliant responses, @mbq and @cardinal have been promoted to moderators emeriti. Although the diamonds no longer appear next to their names, I hope they will continue to participate when they can. They will always be w...

discussion moderation election

« first day (2129 days earlier) ← previous day next day → last day (2872 days later) »

Transcript for

Jun '1615

Ten fold

CrossValidated's general room for gossip, grumbles, and idle c...

join this room
about this room

00:00

06:00

12:00

18:00

all times are UTC

site design / logo © 2024 Stack Exchange Inc; legal

mobile