So, how much does the vacuum cleaner really cost?
Also, I wonder if they asked you your age and your street number, would they cancel each other out, or the most recently asked question would rule, or what?
Mainly though, a post with no comments is sad.
Have you tried this? Depending on what kind of person you pick, sometimes it guesses correctly very early.
A post with no comments is unlike a fish without a bicycle.
M/tch, how far is it, in inches, from here to the sun? Also, do you have fifty bucks you can spare?
I am amazed by the extent to which my "decisions" or "beliefs" about things are influenced by what I think other people would find appropriate. Throughout my academic career, including graduate school, I absolutely never read anything for school that wasn't required, and never did any work more than a week ahead of time. I just couldn't convince myself that anybody actually went beyond the minimum required effort, and thought that if I told someone I was doing so, they would think I was some sort of self-important snob. Obviously this was ridiculous at age 25, when I was simultaneously berating myself constantly for not being like all the other students who were working harder than I was, but still, I couldn't get out of the pattern. Even in classes where I enjoyed reading the material.
What comment number is this? Is the next comment number below 150?
I just finished watching Hot Shots II. It's like Tremors, Airplane, or Underworld. If I see them on, I have to watch at least a half hour.
I had always heard this described as "anchoring" rather than "seeding," but I suppose it's the same thing. The primacy effect and the recency effects figure in somewhere as well.
The most frustrating thing about them is watching the slow train wreck of a group of people making decisions based on them, with nothing you can do about it.
9 to the national healthcare debate.
Yeah, I'm not entirely sure Nick is describing exactly the same thing as the anchoring effect, but it's close enough that it might be included.
10, 11: I'm sure that I wasn't being precise. I do know the source through which I encountered the idea so I could look it up but . . . If nothing else posts with obvious errors get more comments, right.
Thesis: Tremors is objectively awesome.
3: That's pretty neat. I picked a pretty obscure computer researcher, and it guessed Linus Torvalds, which was sort of close. It'd be interesting to rank people by the number of info-theoretic bits of required to identify them uniquely. log2(6e9) is 33, so at most 33 questions for every individual on the planet, and much less for some.
14 cont. Of course, that would only make sense relative to a specific compression scheme given a certain distribution of people in the "message". Assuming that a person's frequency in the message is proportional to their fame, that'd simply be a measure of fame. If it's how often the person is played on akinator.com, it's also probably proportional to fame. So, it's not telling us anything new, just giving an interesting number.
Is your character in the Bible? YES
...
Is your character linked to Jesus Christ? YES
Is your character from Southeast Asia? NO
Is your character a member of a dance crew? NO
This is not the line of questioning I expected.
Standpipe, did you ever try omegle? I don't remember you being in that thread.
Heh. It seems to be hung up on Southeast Asia.
I did. I posed as an economist and had an interesting conversation with someone in the book trade.
14: It'd be interesting to rank people by the number of info-theoretic bits of required to identify them uniquely. log2(6e9) is 33, so at most 33 questions for every individual on the planet, and much less for some.
Well, I hadn't even thought of a character and then I did, and it took 20 questions to get it wrong. ... And it took 30 questions to get the next one right. Clearly anime is its baliwick.
max
['Hoy.']
I started that, realized I was supposed to be thinking of a character, realized that that involved thought, and closed the tab.
"Anchoring" or "cognitive anchoring" is usually used to refer to a similar phenomenon in decision making/choice, where we decide to focus on one thing (e.g., how fast a computer is), and then end up having trouble getting away from that thing when making a decision (e.g., which computer to buy).
However, what Nick is describing is pretty much the anchoring and adjustment heuristic of Tversky and Kahneman. I'm sure they'd see his experience as an example of it.
By the way, that linked paper should be required reading. It's 35 years out of date, to be sure, and there's been a lot of work on heuristics and biases since, but it still gives you a good idea of how wedded our brains are to its innate and often deceptive strategies.
Chris! The thing is, I thought anchoring was when you started with one number explicitly, and then fail to adjust enough in some new direction towards the right number. Isn't there a separate name for the phenomenon where the number you start out with in the first place comes from a priming effect?
BTW, anyone interested in heuristics and biases should totally read OB and LW.
Oh, and of course, the master list of cognitive biases.
I'm unimpressed -- it thought Steven Jay Gould was Carl Sagan, and did ask about both Southeast Asia and dance crews. There was a version of that limited to sitcom characters and dictators which was cool; when it missed one, it asked you for a question that would distinguish your target from the wrong guess, and by the time I found it it was very hard to stump.
Yeah, I stumped it quite often, but I was impressed occasionally by the ones it got right. It seems to have a massive library of anime and manga characters, and it does pretty well with major actors and musicians. It seems more random in the order of its questions than it used to be, as 16 suggests. One gets almost all the way to the end of a series of questions about an author (verified British, male, dead, real, married, etc.) and then gets, "Is this a cartoon character?"
13: yes. "Just a few household chemicals in the proper proportions."
Isn't there a separate name for the phenomenon where the number you start out with in the first place comes from a priming effect?
What about priming? That's what it's called when it influences word choice, rather than numbers. No particular reason for it to be different.
I presume that if you started by asking them their street address, more people would answer "more" to the second question.
YMMV. In the UK, most street addresses are lower than 150.
The game linked at 3 is addictive. It wasn't able to guess Eartha Kitt, however. I gave it 2 chances and it came up with Tina Turner and Whitney Houston. Puhleeze.
That's what it's called when it influences word choice, rather than numbers.
I wasn't aware that distinction was made. Wikipedia seems to support me, though they don't mention numbers in the article: Priming in psychology occurs when an earlier stimulus influences response to a later stimulus.
I tried the thing linked in 3 with Martin Lawrence, Kurtis Blow, and Julia Child. It was 0 for 3, but came pretty close with Martin Lawrence.
Here's the Dictator/Sitcom Character thing. It's a restricted domain, but it's very hard to stump.
OT: We're Number 37! (Pretty good song & video re: U.S. ranking for health care.)
The game in number three got Marlene Dietrich and Al Gore, but guessed Thomas Mann for Paul de Man (not too shabby) and some truly random game designer for Thomas Pynchon.
The game in three easily guessed Cyndi Lauper and Haruki Murakami, but didn't even hazard a guess for Albert Schweitzer. At one point it asked if he was French and I said "partly" -- because it's all confusing over there! -- so maybe I screwed it up.
The game in three gets cheeky if you pull the old 20Q stunt of just saying yes until it makes a guess. I got worried when it established that I was thinking of a black female American singer and then asked "is your character from Belarus?" but it turned out cute.
Jimmy Carter has been played 1159 times previously.
5.1: 5,892,480,000,000. I might have some extra zeroes in there, though. I did it on a calculator with an eight-digit display, so I had to do the scientific notation in my head.
3: That's a weird game. I started off thinking of Santa Claus. (I thought of him because of a college philosophy class in which I determined that Santa Claus "exists" because he's based on a real person. Reading comments here I got the impression that unconventional theories of meaning might throw this program for a loop. No such luck, though.) Starting with the sixth question I got asked in succession if my character was fat, has white hair and wears a red suit. I answered yes to all of them, I'm impressed that it got there so quickly, and expect the next one to be the guess. Nope. Question number nine: "Is your character a poet?" The following two questions came out of left field too, but it got the right answer after the 12th.
Leaving my desk right now, but I'm going to try to do it again and see if it can get to the correct answer any more quickly.
Ridley Scott in 29 questions, previously played 114 times. He's my go-to stumper -- once beat a van full of traveling improv comedy players. (I was stumped among them.)
||
On-topic Emerson cyber-stalking!
From comments at DeLong:
Alternatively, you're on an island where half of the people are Chicago school economists and the other half are non-Chicago-school economists, but there is no way of knowing which is which. You are at a crossroads, one fork of which leads to a ravening monster and the other fork of which leads to socialism. How do you phrase your question so that no matter which kind of economist you're talking to, you end up on the road to socialism?...JE
|>
some truly random game designer for Thomas Pynchon.
Oooh. now I'm curious how many game designers it knows. I doubt I could answer many questions about Gary Gygax, actually, but there are other options . . .
31:I wasn't aware that distinction was made.
That's the point I was trying to convey. I'm familiar with priming from linguistics, where obviously it usually refers to words. But given that it seems to be the same effect, why not use it in this context?
I asked it Ed/ward Wit/ten and it came up with Step/hen Haw/king, which is not bad, except for my having answered yes to "Is your character an American citizen?". Taken in by the health-care propaganda, I suppose.
At least we can be sure the game is learning to tailor its questions to the sort of people who use the site. Every time I am asked at least two of "Is your character Japanese?", "Is your character from anime or manga?", and "Is your character from an animated series or movie?"
now I'm curious how many game designers it knows.
It guessed Gary Gygax, for the first game designer I tried (Nigel Findley), and he is the only game designer on the resulting list. But it may have other living game designers.
Freaky. It got Peter Molyneux correct, even though most of the questions it was asking, apart from the two about computers, suggested it was way off track. I suspect it would have guessed Molyneux if I'd used any other British videogame designer of about the same age. Although there aren't all that many internationally famous ones
Also, I was rather weirded out when, after I told it that my character was a real person, it asked me: "Does your character come from the internet"? I had to think about that for a while. It's pretty deep, man.
45: That is an easy mistake to make, you know Hawking has a pitch perfect american accent.
Failed for Barbara McClintock. I bet it fails for 90% of people famous for nonfiction written work
Why isn't she on the currency?. Her, JW Gibbs, Alexander Graham Bell....
I was just thinking that "Fight the Power" is dated not just for the 1989 line but also for the reference to stamps.
It failed to get Eamon de Valera (guessing a Czech president instead) but successfully guessed the main character of "Girl with a Dragon Tattoo".
I find this phenomenon fascinating -- but, of course, brains aren't silly. They're the most fabulous machines ever. This ill-effect is a side effect of our ability to function, make enormous leaps of logic and creative thought in an utterly confusing world.
You might be intrigued by this cite, in which Bruner and Potter (1964) showed that guessing about the content of an image when it was blurry slowed the eventual recognition of the image (that is, the same phenomenon you're demonstrating with your trivia calendar).
http://www.ahs.uwaterloo.ca/~kin356/bpdemo.htm
36: It guessed Yossarian for Slothrop. Close!
26: Here's the Dictator/Sitcom Character thing. It's a restricted domain, but it's very hard to stump.
I stumped the first time out with 'Julius Caesar'.
max
['I did not help it, though.']
It got Caesar for me just now -- is 'not helping' like the herring joke?
56: No, max just gave it the silent treatment.
I'm probably the last to see this, but that doesn't make me like it any less.
I can't seem to link the order denying reconsideration, and giving Taitz a chance to show cause why a $10k sanction isn't appropriate.
This'll do. Click through to the decision on reconsideration.
56: It got Caesar for me just now -- is 'not helping' like the herring joke?
Apparently I was answering one the questions 'wrong'. Three more attempts got me to Caesar.
58: I'm probably the last to see this, but that doesn't make me like it any less.
Wait - are you saying it's bad they denied the restraining order?
max
['I'm glad they used sanctions on her.']
On the OP, I just got an example of "seeding." My teacher was casting about for some occupation beginning with B. First he said "baseball player," then decided against it, and his next pseudo-random choice was "basement technician."
62: I think you have it backwards. It confused me for a moment.
Have you tried this? Depending on what kind of person you pick, sometimes it guesses correctly very early.
Hmm. It came up with Dagny Taggart when I was thinking of Emmeline Grangerford. That probably says something about the interwebs.
YOU SPELLED "MICHAEL" WRONG