Sunday, March 7, 2010

Comic 710: Unsolved Problem

Collatzeral damage

This comic is fine. It reminds me a little of the Drake Equation comic, in that it begins by telling you about a scientific principle but then just descends into a kind of meta-commentary on the principle itself. It's the kind of thing you could imagine him doing to death, but luckily he hasn't so far. So I'm not going to say much about the comic and instead use this as a way to talk about the intended audience of xkcd.

This is an interesting example of the hard math/easy math stuff we've been talking about a bit recently. It occurs to me that there are really two ways of looking at whether the science content of a comic is obscure or not: Some people might say this comic proves I'm wrong, that xkcd is not trying to broaden its audience to the point of stupidity. After all, they might say, who but math students has heard of the Collatz Conjecture? But there's a difference between not having heard of it before and not understanding it. The comic makes clear that the actual conjecture is that halving or multiplying by 3 and adding one will eventually lead you to 1. Even though it's not stated, it's clear. And once it's clear, someone with only a passing knowledge of simple math can understand the point. The math is simple. Someone could try it out themselves with any number they choose. So it's a highly accessible topic.

Contrast that with two old xkcd comics: #12 and #62. In these comics, you really have to know what they are talking about to get the joke, and it's not obvious at all from the comic what a poisson distribution or karnaugh map is. It's a joke made for people who already know these things. That's the kind of thing we don't see anymore - the requirements, so to speak, of the comic have been lowered substantially.

The question, then, is whether that is a problem. I don't think it's a problem per se but it is indicative of the selling out of xkcd - the shift from actually wanting to make actual nerd jokes (and not "nerd as a subculture like any other" jokes) to wanting to have as broad an appeal (and as broad an audience of potential shirt buyers) as possible. What do you guys think?

Lastly, while randall's graphical depiction of the conjecture is nice, I find myself fascinated by the patterns in this one.


  1. Oh my. Carl said something positive about XKCD. The end is near!

  2. Oh, Carl! You are so /wrong/ on this one! Coming right after Rob's essay on 'in jokes' makes it all the more obvious; the fact that is is accessible is what makes it great. You don't have to be on the inside. I don't have to feel like an uber-4chanismothersmilktome-nerd.

    As an outsider I can immediately grasp the beauty of the Collatz conjecture: that is fiendishly simple to understand and thus far devilishly impossible to prove.

    But, I hear you clamour, Is It* Funny? I would say, Yes! It draws you in, regardless of whether you have a degree in NASA or not. And just as you start to appreciate it, the punchline makes you realise your friends are starting to make plans without you. It's classic feint-right left-hook.

    I particularly object to your implication that this is dumbing down. I know you don't commit yourself, but you claim dumbing-down-as-selling-out. You can't have it both ways: either Randall is pandering to nerd subculture or he is making intelligent comics for intelligent (but not necessarily educated in a narrow branch) people.


    * xkcd, not the conjecture, obviously.

  3. No Carl, the comic is not fine. Another difference between comics 12 and 62 and this is: the maths is necessary to the joke in those earlier comics. Remove the maths reference your emove the whole raison d'etre of the comic.

    But 710? You can swap out Collatz Conjecture for any maths problem you care to name. Hell you could swap it out with any conjecture from any discipline, sociology right through to astrophysics.

    So the whole maths aspect of this comic is actually a total name-drop. It's got no bearing, no necessity, no significance on the actual joke.

    This comic is another nail in the "xkcd is about maths" coffin.

  4. I always hated that Poisson distribution one. It's actually the comic I think about most when people say, "XKCD used to be sooo good!"

    Being valid only for positive numbers is probably the least unique or characteristic thing about poisson distributions. There are millions of functions that don't take negatives and this joke would be just as valid for all of them. "I am a natural logarithm!"

    This was particularly annoying when I first read it because I didn't know what a poisson distribution was. I had to look it up to find out that I didn't have to look it up because the joke wasn't really about poisson distributions.

    It's not a high level math joke, its a low level math joke with a high-level tarp thrown over it.

  5. It "draws you in" and makes a totally cliche joke about nerds being loners. Amazing. I'm pretty sure my friends won't be calling me because I just shit myself from laughing and I lack proper hygienic habits because I love math.

    So no, its not funny. Despite being a "classic feint," the feint is followed up by nothing surprising, nothing new and nothing interesting.

    I mean, really. His friends are going to stop calling him because he's a nerd/shut in. He might as well just have followed up with a mom joke.

    And all that stuff Sbooty said.

  6. I agree Dan. The joke itself wasn't funny. But that's okay, I actually enjoyed looking at the little diagram he did there. I was entertained for the 30 seconds that I spent double checking his math just to make sure he wasn't making stuff up. So, all in all this Comic kept me entertained longer than most.

    CAPTCHA: Erotters... Erotic + Rotting = Erotters. Uhg.

  7. I'm glad to hear Carl say "This comic is fine."! I was worried that this site was going to be full of a bunch of no-men.

  8. Ugh. The underlying joke of this comic is passable if a little trite, but the devil is in the details, as it were.

    First, polygraphs do not use electrodes attached to the head. Randall is probably thinking of an EEG.

    Similarly, the readout of a polygraph is not one bar like an EEG, it's a representation of several things such as blood pressure, respiration, vocal stress et al. this may seem minor, but the similarity of an EEG readout to a seismograph is part of the joke (and they really still don't look similar at all).

    Sedcond, and this is even more minor, but PROTIP: "Protip" is one word.

    Also, why would someone so self-consciously nerdy as Randall use the imprecise "lie detector" in lieu of "polygraph?

  9. Fuck that's not how you use protip

    Yes you too Nymous, you're also doing it wrong

  10. Wow, what a gay blog

  11. PROTIP: To kill the cyberdemon, shoot it until it dies.

  12. Using protip wrong is nothing new to Randall.

  13. Gay blog more like Gae Bulg

  14. more like Gayhae Blarghah, amirite

  15. Ah hah, the lie detector one was actually kind of funny!

    Anyway, about this post, Carl don't you think this comic is still loads better than the tripe that Mandall usually feeds to us? I know we're way past giving him a break for taking a second off producing unadulterated crap to make only slightly below average strips. That said, I'm half expecting some Cuddlefish to blast us with a "you complain when it's bad, then it gets better but you still complain!"

    The esteemed gentleperson making himself known as Sbooty: Sir, you are most assuredly an evident rube and a feeble minded ignoramus. Therefore, allow me to educate you in the finer points of mathematical science. Indeed, the Poisson distribution does not receive a value for an argument such that this argument is less than zero. Though your limited knowledge will undoubtedly bar you from doing so at this time, I shall also supply to you a more symbolical formulation of this concept, that you may revisit it decades from now and be enlightened: ∀x∈ℝ⁻ ∄P(x), P(x) being the value of the Poisson distribution at x.

    Indeed, though you may yet be unequipped in rational capacity prerequisite to questioning unsourced claims, even if they are made by an academically superior person such as myself, allow me to demonstrate: The value of the logarithm naturale over -1337, which you shall observe is negative. The minus sign to its left indicates this fact. The result is, of course, complex, and I suspect far too complex for you to grasp, but no matter. Suffice for you to acquiesce that a result exists.

    Contrarywise, observe the lack of any similar computation being conceivable for the Poisson function. Do not be alarmed by the complex mathematical expression I have input! It is merely the definition of a Poisson function. You will further observe that the parameter x appears in the denominator as "x!". This is not, as you would likely presume, to emphasize its urgency. The exclamation mark indicates a factorial, a mathematical concept wherein one multiplies a sequence of integers starting at 1. No factorial can be computed at the value -1337, a familiarly negative number. Notice how we are informed that "n! is a sequence with integer values for nonnegative n.".

    You might benefit from the some intutive elaboration of the matter to make it clear for your uneducated faculties: A common use of the Poisson distribution is making statistical computations regarding the number of times an event occurs in a given interval. For example, one might study the number of grunts an average shitter produces for a sustained 10 minutes of shitting, where x is the number of grunts and P(x), the value of the Poisson distribution at x, is the probability that number of grunts shall be produced. This would be a very shitty use of the function, yet nevertheless you will concede that it is quite impossible to produce a quantity of "-1337" grunts.

    I wish you luck in your further mathematical adventures! I will suggest this resource to you: Please do not be intimidated by its advanced level! Take heart; with hard work, you shall soon see you are able to grasp the conceps detailed within.

  16. Typical high school mentality from Ar-Pharazôn. I got a chuckle out of your completely unnecessary use of set notation. I can just see you opening up the symbol map and inputting those operators and your little nerd boner rising steadily.

    Knowledge of first-semester probability is not particularly impressive, and even excusing your over-zealous set notation (it is very common among people trying to show off), your post is still not very impressive.

    All you needed to do was point out the difference between not being defined over an interval and having no real output for an interval. Instead your pedantic post reeked of self esteem issues.

    Work on your math more and other peoples' math less.

  17. Whereas a man correcting another man's maths about another man's maths totally DOES NOT reek of self-esteem issues, guys

  18. I thought this one sucked. I mean, you can only make a stereotypical nerd joke so many times before it becomes completely unfunny.

  19. I was thouroughly unimpressed by the lie detector one. It doesn't even make sense. Yelling at someone to make a bunch of peaks on a lie detector doesn't allow you to detect earthquakes; it just makes a bunch of peaks that would represent an earthquake in a completely different context. It would be much easier to manually jiggle around the needle. Or to just draw the peaks yourself. PROTIP: Your arm and a pen can be a seismograph in a pinch.

  20. somebody having a heart attack could double as a lie detector

  21. Rinnon: You said, "The joke itself wasn't funny. But that's okay"

    Jump off a cliff.

  22. xkcd stopped making sense, so I decided to make my move and stop following it.

    A lie detector can double as a seismograph... how? In reality, lie detectors are unreliable, so it wouldn't double as a seismograph. If the premise assumes that lie detectors do work then... it requires someone who's such a compulsive liar that he'll lie when an earthquake is actually going on that exact moment?

    Quoth Leonard: "PROTIP: Your arm and a pen can be a seismograph in a pinch."

  23. Wait what?

    Mole and Leonard: Are you complaining that the concept doesn't make any sense, while the joke is that this concept doesn't make any sense?

    There are two pieces of absurdity here that make for humor: 1) Lie detectors work at all. 2) A person is lying even when they couldn't possibly know the correct answer.

    In fact, the latter concept is so funny that somebody made an entire movie about it:

  24. Art: absurd is not necessarily funny.

  25. I liked 710. It's what XKCD is trying to be most days - the math is understandable from the comic and there's something of a punchline.

    I don't think it's necessarily indicative of selling out, or at least I don't think he made the shift consciously to wring more money from people (and I don't think that the comic would be better if it were only understandable to a few). XKCD has always had a large number of non-serious-math-based comics which attracted readers without math/science backgrounds - what would be the point of frustrating part of the audience with the math ones?

  26. Mole: "A lie detector can double as a seismograph... how?"

    It can't, that's the point of the comic - whether it's funny or not, the fact that you felt the need to point that out is absurd.

  27. I'm complaining about a nonsensical concept that is confusing and doesn't make me laugh. A Rube Goldberg machine doesn't make much sense, but at least it takes effort to plan out and is amusing. This comic was just Randall going "Hey, these two kinds of graphs are vaguely similar looking. What if you used one instead of the other?" I mean, he could have substituted the stock market for a seismograph (topical!) and it would have been basically the same joke. Other things that look similar: a child's drawing of mountains and a triangle wave LFO; a skateboard quarterpipe and amount of radioactivity over time. Wow, that's week's worth of material.

    Also, Liar Liar isn't about not knowing the correct answer. It's about a guy who gets through every part of his life by telling lies and who suddenly finds himself unable to tell lies.

  28. DINOSAUR COMICS SUCKSMarch 8, 2010 at 6:12 AM

    Yet another exhibition of Dinosaur Comics' much-admired "characters" that XKCD doesn't have because it is such an oh so terrible comic. There's T-Rex, who talks like Ryan North. Then there's that orange dinosaur, who talks like Ryan North. And then there's God, who talks like Ryan North.



    The subject matter is quite good. The question about God creating a rock he can't lift is stupid, so it deserves a lampooning. The alt text is lame though - obviously strongmen want to show off their muscles and wow, a reference to Marvel Comics that has no joke whatsoever except for being a reference? I didn't expect that.

  29. It's even worse because "Protips" are supposed to point out very obvious facts. That's what makes them so funny.

    Once again, Randall has referenced a meme he doesn't completely understand.

  30. I may not be any good at math, but I'm pretty sure that 5*3 =/= 16.

  31. Oh, nevermind, you have to add one. I'm also not good at reading.

  32. ha ha lie detectors detect wrong answers instead of answers that were purposely wrong (or accidentally correct). Old joke.

  33. The question about God creating a rock he can't lift is stupid, so it deserves a lampooning.


  34. A "poisson distribution" is some kind of place that makes fish, right?

  35. Uncivlengr, here's a joke for you: "A dog can double as an alarm clock". Do you like it? I hope you do, because it works on the same principle as today's comic.

    Expanding on the subject: this is the sort of "joke" I read and then ask "yes, and what?" The "pure" joke isn't funny, so I look for more information on how it can be funnier. And there are many things that can be implied in there, but there isn't enough information on the panel... Is Randall implying seismographs are as imprecise as polygraphs(and in the exact same way)? Is it actually working? Also, didn't Randall learn to use Protips yet?

    So many questions, so little funny...

  36. I didn't say it was funny, but pointing out the reality of the situation in the comic is like watching Transformers and pointing out that real cars don't actually transform into alien robots.

    Whether or not the premise is entertaining, you've missed the point.

    Now, would the comic be better if the caption were, "Protip: polygraphs do not make good seismographs"?

  37. Hang on, hang on, can't you take the factorial of a negative number if you use the gamma function generalization? Maybe not, it looks like that only takes complex numbers whose real component is greater than one.

    uncivlengr please endeavor to be less retarded in the future.

  38. What Randy's done here (711) is said "Boy ravens sure are like writing desks amiright? Aye? Aye?" and then closed up shop.

    That's not a joke. It could become a joke, sure, but anyone who thinks it's funny only does so because they are adding the extra material needed to make it funny.
    If you laughed it's not because Randy's a funny guy, it's because you are.

    D+ Incomplete

  39. "pointing out the reality of the situation in the comic is like watching Transformers and pointing out that real cars don't actually transform into alien robots."

    I choose to suspend my disbelief when watching Transformers. Why? Because of the Rule of Awesome: if I accept the premise of transforming robots, I am rewarded with something awesome (i.e. fights between the aforementioned giant robots). There is also a Rule of Funny: if I am willing to accept something unrealistic (e.g. giant, transforming robots) I will be rewarded with something funny (e.g. this: ). But today's XKCD had no funny, and not enough reward to justify the effort of typing the url into my browser.

  40. There are a lot of people here not getting the joke, and although a few have got it right, they didn't do much explaining. This is what I see happening:

    Interrogator asks if there is an earthquake.
    When the man answers no, the polygraph marks him as lying. It does this because there is an earthquake happening somewhere, so the man is lying, even if he doesn't know this. Polygraphs don't work this way, but it is an old joke. Randal uses this joke in addition to noting that the results of the polygraph would be the same as the results of a seismograph.

    This was not funny.

  41. I would like to know why this man is lying about the ongoing earthquake, and why this other man is, must be, constantly asking him if an earthquake is happening.

    The mechanics of the joke are dull. The characters are insane to make it work. It sucks.

  42. Today's breaks the record of the distance between a disembodied head and its body in a xkcd comic.

    Also, this isn't the first time Randal completely misuses the concept of "protip" (see alt-text of

  43. Femalethoth: Why yes, that's interesting. I had not thought of that one.

    After taking another look, I've got good news and bad news: Looks like my math might have been more correct in some sense had I mentioned only the negative natural numbers. Theoretically you could generalize to Gamma (I believe that is the "right" definition of Poisson that math textbooks go with, anyway) and receive a result for, say, -6.9.

    On the other hand, however, the Gamma function still has no value for negative integers. Or rather, to be more precise, it's divergent. It's not even divergent in a well behaved manner, since on one side of the asymptote the thing goes to minus infinity and on the other to plus infinity.

    That said, everyone else, here's why I think the lie detector is funny: The archetypal seismograph is a device with an oscillating needle which draws a waveform like image on the length of paper scrolling underneath, based on seismic input. I'm not sure if they'd actually be -"bent"- sine waves, but hey why not. Anyway, the paper rolls out and the needle makes scribbles on it, and if there's a 'quake then it makes big scribbles, so you point to that and say an earthquake happened at that moment.

    A polygraph, as can be evidenced by the name, also uses a superficially similar setup to generate output. Or, the archetypal polygraph does, since modern ones interface to a computer. The polygraph, unlike the other device, measures not seismic activity of the Earth but various physiological metrics of a human, such as pulse, skin conductivity (think sweat), etc. When these increase, it also makes big scribbles. Incidentally, if you assume lying people are stressed and therefore exhibit greater arousal states, these peak when people lie. You can claim based on the heightened readings registered by the polygraph that at the times corresponding to the spikes the subject person is aroused, therefore lying. You can point to the peak and say the subject lied at that moment.

  44. There are numerous controversies regarding that last application of the polygraph, but that need not concern us.

    As you will deduce, the polygraph should not be able to record earthquakes, because neither it nor the person interviewed have any way of sensing them (unless of course they are right on top of the earthquake). Therefore, if you attempted to reappropriate a polygraph as in the comic, and interrogate the subject, then go by the result, you should not expect to gather statistically significant data. However, in practice, you may find that you do get statistically significant and quite correct readings with this setup.

    This is because in your excitement to receive confirmation when there IS an earthquake, your questions and tone may be -unwittingly to you- more urgent and harsh, which may cause the subject to become aroused, not because he is lying, but because he is intimidated or startled. Since you are able to point at the peaks and say an earthquake occurred at those moments, and be reasonably accurate, you may be led to believe that against all odds, the polygraph IS a good seismological instrument.

    It is not. Obviously there is a hidden relationship distorting the experiment, in this case your knowledge of the answer. You may eliminate this, and obtain correctly statistically insignificant random chance level results from a mendacity-based seismographic device by having an accomplice record seismic data, and informing you of whether an earthquake is occurring only after you have completed the interrogation. This is called a double blind trial, because neither the experimenter nor the subject know the result that are "supposed" to emerge from the experiment.

    The "hidden relationship" is an extremely critical concept for experiment design, especially in the social sciences. It is better known as a "confound". It's not very complicated, I remember this being mentioned as far back as my Psy101 classes. The point is, this comic effectively illustrates, in what I would like to call classic xkcd style, the peculiar scientific concept in naive ignorance of it a priori. Confounds are funny. When you find it in your setup after days, weeks, months, years (possibly centuries, with others' experiments!) it feels like reality pulled a fast one on you. It's almost identical to a magical trick in this sense, that we are amused by the frustration of knowing that we are being deceived, but not how. However, it is doubly amusing in that we are the magician and the audience, and yet oblivious to our own trick.

  45. OH MY GOD does Monday's comic suck. Jesus god in heaven Randall, it was bad enough with the nonsensical God/R2-D2 comic, but this was just so much worse. What's the joke supposed to be, that he's using the guy's knowledge of whether there's an earthquake going on right now to find out... whether there's an earthquake? The whole point of a seismograph is that it makes the strength of the earthquake quantifiable beyond simple human reckoning. Man this was a bad comic.

  46. The Simpsons and Mr. Show both had lie detector jokes that were actual jokes and actually funny.

    Just sayin'.

  47. I think 710 is okay. It introduced me to a math concept I'd never heard of, and an interesting one as well.

    Whatever. Reading this discussion inspired me to implement the "mechanism" of the Collatz conjecture with a Basic Programming cartridge on my Atari. Here's what I got:

    1 If S>0 Then Goto 3
    2 S <- Key Goto 1
    3 If S Mod 2=0 Then S <- S/2 Goto 3
    4 If S=1 Then Goto 4
    5 S <- (S*3)+1 Goto 3

    ("<-" is supposed to be one symbol, the backward assignment arrow)

    The Atari (narrowly) avoids running out of memory while executing this.

    Run it for yourself if you want. If no one cares, that's fine.

  48. "There's T-Rex, who talks like Ryan North. Then there's that orange dinosaur, who talks like Ryan North. And then there's God, who talks like Ryan North."

    Um, can you really not tell the difference between the characters? Like, really? I mean, of all the criticisms it's possible to level against Dinosaur Comics, you're going to go with the one that indicates that you're basically illiterate?

  49. @Femalethoth:

    It sounds like if you did that, you'd be getting close to just using a Gamma distribution.

    The (intuitive) reason Poisson doesn't work for negative numbers is that its input simply refers to the number of occurrences. You can't have a negative number of events occur.

  50. 710 is another one for the picto-blog. Instead of being funny, it's more of a "Math factoid of the day". Randall should stop trying to force a lame joke into everything and just embrace the fact that he wants to make illustrated flashcards.

  51. Ar-Pharazôn, your bullshit only works if we can indeed assume that the person in the chair is unaware of any earthquakes. However, as Randall has failed to give any visual clues either way, we cannot.

    The only clue is that there is a lie detector reacting to a statement, so the simplest explanation is that that statement is being conveyed as a lie. And from that, we know an earthquake IS happening, and the floating headed man has gone through an elaborate method to confirm what he much logically know from the vibrations -- the undrawn vibrations -- he also feels.

    The joke is ruined in multiple respects. Note that the floating headed man seems to be shouting. This occupies the readers mind with a sense of urgency, but one that goes completely unfulfilled. Why is he yelling? Why is the floating headed man so desperate to learn what he already knows? Why is the other man, who's spider-sense is going off, attempting to lie about something that is self-evident?

    The joke might work, minimally, if the two men had set it all up just for the sake of seeing it, then waiting for an earthquake. Yet what we actually have is a psychopath demanding answers from a victim oddly reluctant to give them.

  52. Anon 12:48: I'd love to agree with you, but unfortunately, your argument is extremely weak. Math is a field where integrating 22 dimensional vectors in 68 dimensional bent knotty hyperspace is common place. Many mathematical concepts start as extremely simple, limited-domain things and evolved into maddening Lovecraftian monstrosities. Take the sine and the logarithm. A simple ratio of height to hypotenus and a trivial "which power is this" function. Yet you get mindblowing shit like 1+e^i*pi=0, Euler's identity. (Which incidentally got him accused of witchraft, my professor said)

    Ken: You appear to be confused in multiple respects. Let me assert a number of things:

    1.) Output from lie detectors has very little to do with whether what you say is truthful, even for rather original epistemologies. Instead, it measures how excited you are. People become excited when they are having sex, masturbated, fighting, arguing, taking exams (even if they are not making anything up), flying airplanes for the first time, chased by a pack of dogs, and yes, when a figure of authority interrogates them and they feel their answers and/or behaviour may not be satisfactory to the authority figure, regardless of whether it is true or not (a suspect has not planted the bomb, doesn't know where it is, but the officer isn't gonna believe him).

    A cursory Google and/or Wikipedia search should yield ample corroborating information.

    2.) Seismographs are not for measuring activity from an epicenter ten blocks down the street. The walls crumbling around you, the difficulty in staying on your feet, and the general sensation of quaking will be more useful indicators (Fun fact: I have been within 2-3 km of a 6.8 Richter scale earthquake). In any likelihood, with quakes at such proximity you might be better off shopping for a new seismograph.

    The seismograph is for measuring very far, hence imperceptible earthquakes. During North Korea's alleged nuclear weapons tests, among evidence were shown data from seismographs in countries surrounding North Korea, if I remember correctly among them was Japan. For sources, try a Google search for "north korea nuclear weapons test seismography Japan" without quotes.

    3.) The device in the strip records a peak because the needle is moving. The needle is not moving because an earthquake is jerking it. The old movie-style seismographs with needles have those needles set up to be extremely sensitive to the most minor vibrations, and the setup has very complicated mechanisms built in to deal with the problems that causes. I believe (old) polygraph needles are moved by an electrical oscillator, which is modulated by the parameter measured (pulse, conductivity, etc.)

    The joke is not that the polygraph mechanism can be hacked to detect seismic activity. That is ridiculous, but not a joke. The joke that it is easy to mistake psychological confounds associated with lie detectors in general for actual extra sensory perception on the subject's part, or whatever your explanation is.

    The reason why the confounds are so likely to skew your data is that when you have recorded an earthquake, you are very excited to find out whether the subject will, in fact, guess right. Your excitement will spread to the subject, so to speak, and the reading will be "lie", or a peak. When you have not recorded an earthquake and ask the question, the "excitingness" of the issue is technically equal, but human nature is such that you will be unexcited. Thus, by a multitude of mechanisms the subject will be influenced and remain calm, hence register no peak.

    This is not a joke about seismology, it is a mildly amusing observation about the (often amusing, if you have the sense of humor to be able to laugh at your own failure) realities of psychological and psychiatric experimentation. The human factor is, in fact, part the reason why these fields are regarded as "soft" and "imprecise" by those favoring the natural sciences.

  53. troll is not obvious
    what the fuck, Ar-Pharazôn? you actually sound like a typical xkcd mathfanboy.

  54. Note on the comic: The lines indicate trembling of the man due to nervousness. Observe that neither the interrogator nor the device are similarly "in motion". Also, remember that actual earthquakes generate extremely rapid, high frequency vibration- add to that the overall chaos and disorientation that comes from experiencing a vicious natural disaster, and the human perception of an earthquake would most likely be an unidentified, confused blur.

    Footage from secured cameras notwithstanding.

    The interrogator is yelling, because he is attempting to assert dominance and inspire compliance in the subject. This is a common, non-brainer, intuitive interrogation tactic. It is a natural reflex for all but the most strictly mannered of people. Further supporting my take on this strip, the experimenters behaviour demonstrates one of the very significant problems involved with polygraph testing: With someone yelling at you like that, while you sit there, submissive, tied to a chair, and your very thoughts are (in a sense) exposed, who wouldn't be scared, telling the truth or not?

    You have noted yourself, moreover, that the reader's mind is occupied with urgency. This is just a stick-figure drawing with a humorous premise. Imagine a middle-aged, harsh-mannered colonel in military uniform, who has previously ordered you to be beaten in your cell several times, shouting at you to guess if an earthquake is occuring: An impossible task, failure of which might very well lead to your death. Although in practice most polygraph sessions are likely not so extreme, this should better illustrate the situation for you.

    As should already be clear, there is, of course, no spider sense. The subject is merely frightened by the interrogator, and attempting to appease him in any way possible. Although it is impossible for him to answer the question asked, he still attempts to do his best by guessing.

    Indeed, the particular response is correct, although probably ill-chosen. He has answered "No", yet you can see a peak is recorded, which the interrogator will likely interpret as the subject telling a lie, and conclude that an earthquake did, in fact, happen. The point is that polygraph tests are unreliable such that they essentially allow the interrogator have free reign over what is true or what is not, regardless of the subject's objections. Of course confirmation bias will easily instill in him confidence that the method is viable. That said, I think a response like "Yes... Yes! I CONFESS, YES THERE IS!" would be less confusing.

  55. Is "Seismograph" the smallest xkcd ever? Randall is getting lazier and lazier and this one is almost forum-avatar sized. The only time he makes a "big" comic is when he wants to sell a poster.

  56. There's a whole lot of tl;dr in this thread.

  57. ar-pharazon is still in the phase of life where he thinks that the person who writes the longest post with the biggest words automatically wins.

  58. William Monty Hughes syndrome

  59. @Ken: You said: "The joke might work, minimally, if the two men had set it all up just for the sake of seeing it, then waiting for an earthquake. Yet what we actually have is a psychopath demanding answers from a victim oddly reluctant to give them."

    I don't know why, but when I read that last line, I couldn't help but laugh. I just found the situation to be extremely funny to imagine. Shame the comic wasn't as funny as your description of the comic.

  60. @Rinnon:

    You sound too much like Ryan North!

  61. What a lame site

  62. yeah xkcd is really shitty

  63. Ar-Pharazon: how many hours have you spent analyzing this comic and spent posting replies about it? Do you always talk in such a ridiculously pretentious fashion? Such verbiage may lead one to infer that while your musings may be prolific, your social life is quite likely limited in nature.

  64. Shh, let him live in his self-congratulatory fantasy world.

  65. Anonymous 4:48:
    Just felt like asking, isn't having a pretentious attitude dependent on whether or not the speaker knows what he's talking about?

    Like, a scientist isn't pretentious, because he knows what he's saying. It's not dependent on the listener, right? So to say that Ar-Pharazon is pretentious makes it sound as if he's wrong in some aspect, which you haven't shown.

  66. @TheMesosade

    His pretentiousness has to do with his tone and not with the subject matter.


    Anon 12:48's argument was not intended to be rigorous, only intuitive. It is how to explain the reason that Poisson(x;lambda) is undefined for x < 0 in one sentence to a person who doesn't know what it is and what its mass function is. Of course if you wanted to show why symbolically, you would point to the factorial and discuss its domain.
    That being said, you seem to be saying that you can't write off the Poisson distribution's possibility for utility outside of its normal purpose (with your sine and logarithm examples).
    I would argue that in the statistics side of math, extending any one distribution or looking at general cases usually ends up producing a separate but similar distribution with more general cases. Take the Gamma distribution in comparison to Poisson for example.
    So, even though Poisson can evolve into crazy things, I think it's safe to limit the discussion of it to its intended use because any other use gets "rebranded" into another distribution.

  67. you might as well have used the sine inverse function for all i care

  68. The annoying thing about this is that graphing Collatz is for computers to do

  69. @ TheMesosade

    Anyone with an undergrad degree in math, who actually paid attention in class, can see Al-Pharazon for the pretentious amateur that they are. "Mind-blowing shit like 1+e^i*pi=0" indeed; those of us who moved beyond Wikipedia are not startled by that equation, because we can give it proper context. (For example: what is actually impressive is that

    1 - pi^2/2! + pi^4/4! - pi^6/6! + ... = -1,

    not some bullshit about "the most important numbers in mathematics omfg".)

    To paraphrase Rob, he has all the trappings of someone who wants to look like a mathematician, but, for all that, you would not ever mistake him for one.

    Protip: no one expresses the domain of a function as "∀x∈ℝ⁻ ∄P(x)". It is ugly and clashes with the common usage of first-order logic.

    @ Al-Pharazon

    In the entirety of the XKCD forums, there are half a dozen people who actually understand math beyond calculus-for-engineering. You are not one of them. (Then again, neither is Randall.)

  70. Holy shit, dudes. I just went to the forums for Seismograph. This is heartening: "Didn't think the comic was very funny, but I lol'd at the title txt." "didn't find it that funny either..." "really? Randall's been getting lazy." "meh." "...what? Randall, you've been off your game the last couple of weeks." "The worst part is the unsettling feeling that he came up with this as a result of watching Numb3rs (which this past Friday had a lie detector plotline) and the Chilean earthquake." "Got to agree this one wasn't even remotely as good as the other xkcd comics :cry: , except if we are all missing some hidden joke/reference, although I fear there isn't one." "A cognitive leap does not a good comic or degrees of separation game make." "Randall must be sick, the last few comics haven't been up to standards."


    And once again the fanboys can't agree on what the comic is about, but that's not surprising given they also mostly hate it.

  71. ArPharazon, are you the same guy by that name who's watching through Buffy over on GameFAQs?

  72. Speaking of the forums...sigh.


    I almost want to register to post now because watching this guy talk about fucking Merriam-Webster like it has anything to do with it is painful. And no one else seems to be able to just link the image. One guy linked to Urban Dictionary but that site isn't very useful at times (like this time).

  73. I love the way some xkcd forumites are starting to doubt the comic.

    This upped my faith in humanity a bit:

    "Lately I've noticed the figure's heads aren't completley on their bodies."

  74. DINOSAUR COMICS SUCKSMarch 9, 2010 at 5:41 AM

    Man, this must be the most insubstantial one ever. It makes the ice cream the dinosaurs are all rambling about seem really, really substantial.

    Basically, Ryan North has noticed that it's nearly spring. If Mr XKCD did the same thing --ie drew one of his stickmen making obvious observations about the weather-- Carl would be all over him. Of course, some XKCD fans might like it, but that's the point really: with XKCD, as with Dinosaur Comics, sometimes the authors rely on the fans to go with a somewhat tenuous idea. That's not necessarily a bad thing.

    And in this one, they all DEFINITELY talk the same.

  75. dcs - sometimes. sometimes they talk the same.

    The thing is, not every comic is expected to be funny.

  76. I think I found a way to make the Seismograph comic funny.

    Panel 1:

    You see man 1, and he asks if there is an earthquake. In the lower right corner is a panel-in-panel displaying a flat output.

    Panel 2:

    Still see man 1, but hear man 2 say "no". The panel-in-panel output goes spiky. (Reader goes "huh, that doesn't match?")

    Panel 3:

    You see the whole picture, only then learn that the output graph was from a polygraph. The room, with other objects in it, is obviously shaking. Man 1 calls Man 2 a liar.

    Is it great? Probably not. But I think it's the best that could have come from Randall's idea.

  77. DCS - Keep it up. We wouldn't want to discourage anyone from analysing a webcomic.

    But 'insubstantial' well now that's a word for you to use alright.

  78. @DCS: There's more to it than that. Dinosaur Comics isn't about stick figures; it's about dinosaurs. So there is a great deal of dramatic irony when two members of extinct species discuss how great the future is going to be, followed by a mock-final "The End".

    I don't know if they really talk the same. The Utahraptor only gets in about four words, so it's kind of hard to tell one way or the other.

  79. "The opposite of love is not hate, it's indifference." - Elie Wiesel

  80. "a witty saying proves nothing" -Voltaire

    also from your blog, you seem to hate creationists.

    (apologies if I've misunderstood your comment, it's very vague. care to clarify?)

  81. I tried to read through some posts here. The entire site is TL;DR. You could simply post "I don't find this amusing" after each comic and be done with it. Seems like a lot of wasted effort for so little analysis, none of which was interesting or unique.

    And even more of a waste of effort to try and slog through it all.

  82. And boy what wasted effort posting that observation on top of it all BOY

  83. Kinda hard to make your point when you critique xkcd's early, formative comics.

  84. Actually, there does seem to be some research linking (almost imperceptible) earthquakes with panic attacks in sufferers, albeit a somewhat tenuous link as yet. Sooo... All of a sudden, the comic becomes possibly true!

    Naah... I didn't laugh at this one. Didn't really see the point. The comic didn't make any particular possible joke clearly enough. I'm glad the more rabid fans (i.e. many forumites) seem to have recognised this. Not everything Randall puts out is automatically comedy gold. The Collatz one was fine, though.

  85. Hey Carl, I totally got an idea you should totally listen to:

    What if you buy the rights to release the new xkcd one hour before randall does? That way the hardcore fans would either have to wait 1 hour or read the comic here. What if buy the right to release it one day earlier?
    If I was a webcomic author, I would totally do this, just to fuck with my fans. Then again, I am a massive asshole.

  86. then one noble "Xkcd" fan would take it upon himself to sacrifice his own wellbeing for the good of the fora. By reposting the image there.

  87. but they would still all know where it comes from

    they would know, and though they would deny it, it would change them.

  88. Oh DCS. You're my new favorite troll. Don't ever change.


    Chris just slams together a bunch of random stupidity and hopes that somewhere, somehow something in there will end up being funny. He fails.

    Today's Achewood is almost as lame as Dinosaur Comics

  90. AS: sorry, your trolling is derivative. find a new schtick.

  91. @ Rob

    This is a blog for trolls. It's all trolling. It's like a boring version of /b/

    We need to liven it up! xkcd is dull, we need to troll more comics, troll xkcdsucks itself, let's make this place suck less.

  92. I'm not anti-trolls. I'm anti-BORING trolls. The only reason DCS is amusing is how clear it is he's a butthurt XKCD fanboy who thinks he's come up with a way to get under Carl's skin despite not actually reading Dinosaur Comics before his trolling began. AS is just derivative of a troll who is mostly only amusing because he's so bad at it. We don't need his kind here.

  93. Rob Mason: trolling snob

  94. I'm not anti-trolls but this isn't a trolling blog. Disagreeing with you≠trolling.

  95. @ Rob

    But YOU'RE a boring a troll.

    At least post some tits to cancel the boredom.

  96. Please do not encourage Rob's exhibitionist tendencies, believe me you will regret it.

  97. guys do you seriously want to see Rob's hairy mantits because believe me he will show them

  98. Rob already showed them:

  99. Rob wishes he had tits that nice.

  100. That's nothing, he showed the full monty here.

  101. Today's Nedroid is terrible. You can just tell Anthony Clarke is just making up for his own insecurities by making these transparent fantasy comics in which the main character is unbelievably full of himself.

  102. Today's comments are terrible. You can just tell Anonymous is just making up for his own insecurities by making these transparent posts in which the main words are unbelievably full of letters.

  103. I see tough guys like Rob and I slap the shit out of them!

  104. There is no joke in Dinosaur Comics today(1669). Yeah, bike rides are nice. I know because I went on one today. I also like iced cream. But why does Ryan insist on pointing out the obvious? Why does Ryan even try anymore? I mean, WTF? PLEASE STOP RUINING DINOSAUR COMICS!

  105. It took 300 posts for the Anons to start doing the "counter-trolling"?


    CAPTCHA: Sagan. I fucking WIN!

  106. I have counter-trolls on my kitchen counters, they steal my sandwiches

  107. Rob, you have a new favorite troll every comic. Why is your love so fickle?

  108. i know I'm feeding you right now but dinosaur comics has never been about the jokes, more about the tone. if you know what i mean. it's funny sometimes, but mostly you read it because it makes you happy.

    i didn't really like this one either though, to be honest.

  109. I remember when I anonymous...accusing carl of being randall

  110. See, this is one of those comics that could have been funny as some kind of live-action sketch. It'd be at the very least a bit of slapstick.

    It just doesn't work as a single page comic because the joke itself isn't that funny. The idea that "haha, seismographs and polygraphs both use the same readouts in pop culture, let's combine them" just doesn't make sense when you set it up like that. Never mind the missed "PROTIP" joke.

    The "haha, let's combine two pop culture references with a tenuous link that are otherwise incongruous" joke is the basis for most of the recent XKCD comics. The lack of any other joke in favour of just making nerdy references is why people have a hard time looking for the joke.

    - Dr. Dugong.