Re: Some Of Us Really Just Are Terrible At Faces

1

Haven't taken the test yet, but I want to get on the record with the same apology.


Posted by: politicalfootball | Link to this comment | 08-11-11 7:20 PM
horizontal rule
2

Oh my god that was terrible.

20th percentile on real faces, 10th percentile on generated faces. I wanted to just start making up answers to make them go away.


Posted by: Nathan Williams | Link to this comment | 08-11-11 7:42 PM
horizontal rule
3

It boggles my mind to think that anyone could do that.


Posted by: Nathan Williams | Link to this comment | 08-11-11 7:43 PM
horizontal rule
4

80th percentile! I was disappointed in my performance. (But seriously, I was totally making shit up in the group answers. I had no idea.)


Posted by: oudemia | Link to this comment | 08-11-11 7:46 PM
horizontal rule
5

Both of them I got "You scored higher than zero out of every ten people who took this test." And I really tried and had not had a drink tonight.


Posted by: Moby Hick | Link to this comment | 08-11-11 7:46 PM
horizontal rule
6

I honestly thought I was doing above average. I was guessing on only a few.


Posted by: Moby Hick | Link to this comment | 08-11-11 7:47 PM
horizontal rule
7

That was a fun test. I'm very confused by my scores. I sccored in the 20th percentile on generated faces, but 60th on real faces, and until I read the results page I didn't realize that that was the distinction between the two sections. Both times I felt like I was mostly guessing, and completely guessing when they added noise (which elicited a whimper).


Posted by: Eggplant | Link to this comment | 08-11-11 7:49 PM
horizontal rule
8

I'm going to attribute my results to short term memorization of small features.


Posted by: Eggplant | Link to this comment | 08-11-11 7:51 PM
horizontal rule
9

Second decile for real faces, fourth for CG. I have had a couple drinks, but I doubt that has much to do with it. I shouldn't have done it in a dark room - those faces are going to be parading through my nightmares tonight.


Posted by: Minivet | Link to this comment | 08-11-11 7:51 PM
horizontal rule
10

70th on both. I'm generally good on faces and terrible on names.


Posted by: Mr. Blandings | Link to this comment | 08-11-11 7:52 PM
horizontal rule
11

If you scored less than 42 or so on the real faces (and you were sober and paying attention and so on) then you might plausibly have prosopagnosia (face blindness). Congratulations! Also: you have an excuse for not recognizing people! There is tons of information on the topic here, and I'm also happy to drop some links to scientific papers if you'd like them, you kids, you.


Posted by: A Buddy | Link to this comment | 08-11-11 7:52 PM
horizontal rule
12

There are people who get all of the questions on the real faces test correct, such that the researchers involved had to create a test with 20 more difficult questions at the end in order to correctly categorize them. They are called "super-recognizers", for lack of a better word, and they are as much better than the mean (2 standard deviations) as the prosopagnosics are worse than the mean.


Posted by: A Buddy | Link to this comment | 08-11-11 7:55 PM
horizontal rule
13

I got 33 on the real faces (41 on the fake).


Posted by: Moby Hick | Link to this comment | 08-11-11 7:57 PM
horizontal rule
14

13: Would you like to be involved in scientific research? Check out faceblind.org


Posted by: A Buddy | Link to this comment | 08-11-11 7:58 PM
horizontal rule
15

I think that explains why I can't tell Bill Pullman from Bill Paxton.


Posted by: Moby Hick | Link to this comment | 08-11-11 7:59 PM
horizontal rule
16

14: Maybe. It may also just be the end of a long day and I'm not hitting on all cylinders.


Posted by: Moby Hick | Link to this comment | 08-11-11 8:01 PM
horizontal rule
17

Does this mean I have to apologize to all the people I haven't recognized? "I'm not bad with faces, it's just you."


Posted by: Eggplant | Link to this comment | 08-11-11 8:02 PM
horizontal rule
18

If it doesn't meaningfully affect your life then who cares. But 33 is (without knowing about any mitigating factors) impressively low.


Posted by: A Buddy | Link to this comment | 08-11-11 8:03 PM
horizontal rule
19

Hmm, I didn't write down my scores, just my recipes. Still, ow. My wife just got 80th/70th, and claims she felt like she was just guessing.


Posted by: Nathan Williams | Link to this comment | 08-11-11 8:11 PM
horizontal rule
20

Recipies? Deciles. Damn you, Android.


Posted by: Nathan Williams | Link to this comment | 08-11-11 8:12 PM
horizontal rule
21

70 & 60 (59 score on each). Better than I might have thought. But I'm worse than terrible at names. Half the time they never even register upon introduction (or only do so approximately).


Posted by: JP Stormcrow | Link to this comment | 08-11-11 8:18 PM
horizontal rule
22

61 real faces excellent 70 percentile

44 computer generated miserable 9/10 better than i


Posted by: bob mcmanus | Link to this comment | 08-11-11 8:19 PM
horizontal rule
23

Hm, I think I might have figured out who A Buddy is.


Posted by: Josh | Link to this comment | 08-11-11 8:24 PM
horizontal rule
24

So far as I know, not being able to recognize faces has not caused any big problems for me. One time I didn't recognize a guy I was in class with every day for a year, but he went away for six months and lost over a third of his body weight. Otherwise, I only had trouble with people I don't know well.


Posted by: Moby Hick | Link to this comment | 08-11-11 8:24 PM
horizontal rule
25

70th percentile--and I'm usually pretty decent with facile recognition (not so good with names). I hope the researchers are using these results to test eye-witness testimony.


Posted by: Jackmormon | Link to this comment | 08-11-11 8:27 PM
horizontal rule
26

I hope the researchers are using these results to test eye-witness testimony.

They are, and it's basically terrible. In fact, they've tested police specifically, to see if they were better at recognizing unfamiliar faces than the average person (who is terrible at it), and they definitively were not.


Posted by: A Buddy | Link to this comment | 08-11-11 8:30 PM
horizontal rule
27

51 on both. I'm using this as one more piece of evidence about how I'm exhibiting signs of Alzheimer's at only 31. The substantial piece of evidence being that I keep forgetting words when I'm speaking. But now I'm forgetting words and I'm bad with faces!


Posted by: LizSpigot | Link to this comment | 08-11-11 8:38 PM
horizontal rule
28

Nearly average = bad? (Or have I forgotten the average ...)


Posted by: JP Stormcrow | Link to this comment | 08-11-11 8:41 PM
horizontal rule
29

I wish there was a breakdown for each section. I was both just guessing and suffering from low attention span at the end of each part and wonder if there was a decline or if it just seemed that way. I also felt like I kept picking "people" I recognized, rather than specific configurations of face+expression, leading me to choose what felt like the same 3 or so repeatedly.

On the last part, I finally got tired of it but didn't want to quit,* so I just hit the number keys rapidly in no particular order with barely enough time to let the faces appear. I think that might have improved my score. Hopefully the survey is set up to catch that behavior and keep it from skewing results.

Also, my low score does not seem to match my personal experience with memory, but people are of course more than faces and I tend to remember people I actually know. I have no idea how many strangers I've never seen before and never talk to I don't remember. So that part of the test could very well be accurate. I certainly feel like more people I don't really know but have definitely seen recognize me than I recognize them.

Just to make a long comment longer: All male faces? Really? Also, part 3 should have the faces appear in the middle of an axe-broken door in a mountain inn closed for the winter.

*This could be real, actual disorder with me. I took an eye test a couple of years ago where you stared at a screen for a while and after what seemed like a long time, I just stopped. Part of that was this was at an optometrist's place attached to a shop and the store part was playing music right next to the seat where I was taking the test. I got them to turn off the music, which was really distracting, and then re-took the test, which didn't seem so long.


Posted by: fake accent | Link to this comment | 08-11-11 8:47 PM
horizontal rule
30

Well, when you're an overachiever...


Posted by: LizSpigot | Link to this comment | 08-11-11 8:47 PM
horizontal rule
31

I should clarify that I only did the number-punching on the last bit of faces that looked like they wanted to be on with Max Headroom, but the reception wasn't working out.


Posted by: fake accent | Link to this comment | 08-11-11 8:50 PM
horizontal rule
32

55 and 56 on real and computer-generated. (40th and 60th percentile, respectively.) It definitely got a bit tiring at the end.


Posted by: Josh | Link to this comment | 08-11-11 8:54 PM
horizontal rule
33

I was both just guessing and suffering from low attention span at the end of each part

Yes, LB should put some kind of ADD warning on the link.


Posted by: JP Stormcrow | Link to this comment | 08-11-11 8:54 PM
horizontal rule
34

Didn't want to put in 25 minutes for the computer-generated one, but I got above-average (75th percentile) on the celebrity faces one. And I second-guessed myself on one of them.


Posted by: k-sky | Link to this comment | 08-11-11 8:54 PM
horizontal rule
35

80th and 90th , real and generated. But I'd forget all their names until I'd met them five times.


Posted by: Stanley | Link to this comment | 08-11-11 8:58 PM
horizontal rule
36

34: I tried that also. I got below average, but not by much. Also, I learned that Fred Astaire looks like David Niven to me.


Posted by: Moby Hick | Link to this comment | 08-11-11 8:59 PM
horizontal rule
37

Scored in the 70th percentile on both. The whole first half of the test was unfair though, because all the subjects were Caucasian and therefore looked the same. And the second half was unfair because Uncanny Valley Skinheads trigger my fight-or-flight response something fierce.

Actually, though, the test taught me vividly what my shorthand is for remembering faces: key in on eyes, lips, and one other memorable detail of the face or its proportions, and that's about it. When I had all those things to go by I could pick out the target faces instantly, like they were people I "knew." But the blurred sections were brutal, and I was guessing at least half the time.


Posted by: DS | Link to this comment | 08-11-11 9:03 PM
horizontal rule
38

I confused JFK with Christopher Noth. But seriously, that picture looks nothing like JFK. I just watched a miniseries about JFK, I should be able to recognize him!


Posted by: LizSpigot | Link to this comment | 08-11-11 9:10 PM
horizontal rule
39

JFK Jr., not JFK. His inclusion on the famous faces test was a bit of a cheat IMO, since his face is not particularly famous, as opposed to his dad's.

I scored 100% on the celebrity faces test, so I guess my pop culture-fu is up to snuff. Whew.


Posted by: DS | Link to this comment | 08-11-11 9:14 PM
horizontal rule
40

Oh thank you, that totally explains it!


Posted by: LizSpigot | Link to this comment | 08-11-11 9:20 PM
horizontal rule
41

Real faces 62/70th decile; CG faces 66/90th decile -- had to be partially luck, b/c as soon as they introduced the noise into the images I was totally lost.

Most of the CG faces looked scary (seconding DS) and I feel I could pick the most thuggish, Caucasian-looking one out of a police lineup with fairly high accuracy. Not a good feeling.


Posted by: lurkey | Link to this comment | 08-11-11 9:42 PM
horizontal rule
42

Holy shit! I scored higher than 9/10 people on both computer-generated (66 right) and real faces (72 right). The latter score looks like it's sort of especially excellent. Yay I am good at something yay.

For me the problem was when the mottling came in. Distortion makes things very hard for me. When some blogs make me do the captcha thing they might as well say "please don't comment here ever" because I can't read that shit. Very often I just ask for the audio test for the blind.


Posted by: AWB | Link to this comment | 08-11-11 10:04 PM
horizontal rule
43

12: I don't think a 72 score makes me a super-recognizer, surely. However, I have had like 4 margaritas and a beer in the past few hours. It could mean that I have drunk-recognition evolution. "Like, oh my god I TOTALLY REMEMBER YOU. Don't you remember? Like that one time you were at that bar and you were all blah blah blah?"


Posted by: AWB | Link to this comment | 08-11-11 10:11 PM
horizontal rule
44

The famous faces test was 20/20 very easily. I sweated much harder over the non-famous people. But I think I can credit part of this to my upbringing, which involved a lot of listening to my mother shout out the names of practically unknown people as they flashed across the TV screen. Oh wait! That guy playing the deli guy? He was in All the Pretty Stars in Heaven in 1947 with you know that guy who later was in Seven Angry Navy Seals in 1962.


Posted by: AWB | Link to this comment | 08-11-11 10:20 PM
horizontal rule
45

Scores of 60 and 62 for real and CG, better than 6/10 and 8/10, respectively. 17 gets it exactly right.


Posted by: Gabardine Bathyscaphe | Link to this comment | 08-11-11 10:22 PM
horizontal rule
46

(If A Buddy's lab would like to hire me to get wicked crunk and recognize people sometimes, I would dooo eeeet.)


Posted by: AWB | Link to this comment | 08-11-11 10:23 PM
horizontal rule
47

I did poorly (30th percentile, I guess), but I think it might have been due to the pixelation part. I honestly just couldn't see the faces in those pictures - they reminded me of the magic eyes that I can never ever see because my eyes don't like to track together. In person, I'm generally pretty good with faces, though perhaps this is a complete fiction given my score.


Posted by: Parenthetical | Link to this comment | 08-11-11 10:42 PM
horizontal rule
48

Hm. Maybe, I'm bad with faces but good with other cues.

I scored below average on the celebrity one, too. Most embarrassing was missing Obama, but in fairness, my mental image of him is smiling.


Posted by: Parenthetical | Link to this comment | 08-11-11 10:50 PM
horizontal rule
49

I got 80th percentile on the generated ones and 40th on the real ones, which is an interesting split but basically mirrors the difference in difficulty I found between the two parts.


Posted by: teofilo | Link to this comment | 08-11-11 11:00 PM
horizontal rule
50

I took the other test they have - I'm a test-taking machine when avoiding packing! - and it turns out I have a barely above average short-term memory but a good long-term one (if vocabulary is a decent proxy for long-term memory, that is). That seems about right.


Posted by: Parenthetical | Link to this comment | 08-11-11 11:07 PM
horizontal rule
51

I would totally watch "Seven Angry Navy Seals."


Posted by: Robert Halford | Link to this comment | 08-11-11 11:12 PM
horizontal rule
52

I took the "how many dots" test and did pretty well, 97% correct, better than 82% of the general foolio population. I'd better go to bed soon before I experience a blow to my very high confidence in how neat I am and how much everyone must enjoy having such a neat friend.


Posted by: AWB | Link to this comment | 08-11-11 11:15 PM
horizontal rule
53

And yes, that Obama pic was weird. I immediately recognized him, but then wondered how anyone ever captured him looking so crappy, so I tried to imagine it as someone else and couldn't.


Posted by: AWB | Link to this comment | 08-11-11 11:17 PM
horizontal rule
54

67 score and 90th percentile on computer generated and 70 score and 90th percentile on real.

I had no idea I was so good at that. Is there a job that qualifies me for?


Posted by: F | Link to this comment | 08-11-11 11:24 PM
horizontal rule
55

95% correct and 0.68 seconds per response on the dot test. I'm apparently a lesser AWB.


Posted by: F | Link to this comment | 08-11-11 11:55 PM
horizontal rule
56

62 (80th percentile) on computer-generated and 68 (90th percentile) on real faces. The computer-generated one felt much more difficult to me.


Posted by: essear | Link to this comment | 08-12-11 12:50 AM
horizontal rule
57

Huh; slightly better than average for both, which means I just don't care about what the people I meet look like? Because I can fail to recognize people I've waltzed with. ("I was dizzy!")


Posted by: clew | Link to this comment | 08-12-11 1:30 AM
horizontal rule
58

I would totally watch "Seven Angry Navy Seals."

Or "One Angry Navy Seal and 200 Solemn Faces".


Posted by: | Link to this comment | 08-12-11 3:25 AM
horizontal rule
59

45 for real people, 55 for computer people. I should just go live in Second Life. If it still exists.


Posted by: mcmc | Link to this comment | 08-12-11 4:33 AM
horizontal rule
60

Hah. I saw you comment and wondered if artist would mean unusually good or unusually bad.


Posted by: LizardBreath | Link to this comment | 08-12-11 4:51 AM
horizontal rule
61

"Artist" doesn't seem to matter. Chuck Close is prosopagnosic, and prosopagnosics are generally just as good at evaluating attractiveness (and gender, and age) as anybody else.

There is some evidence that people in some careers (sales, politics) tend to be somewhat better than average at faces, and that people in some careers (software engineer) tend to be somewhat worse than average, but it's not a terribly strong effect.


Posted by: A Buddy | Link to this comment | 08-12-11 5:04 AM
horizontal rule
62

Software engineer represent!


Posted by: Nathan Williams | Link to this comment | 08-12-11 5:12 AM
horizontal rule
63

(Oh, look, I still had the tab with my scores open. 43 on real faces, 27 on CG)


Posted by: Nathan Williams | Link to this comment | 08-12-11 5:13 AM
horizontal rule
64

61.1: "Miss, I have prosopagnosia. If I looked in your eyes, I'd never be able to recognize you."


Posted by: Moby Hick | Link to this comment | 08-12-11 5:24 AM
horizontal rule
65

I misremembered: artists did do better than average. I'm not sure I really trust that study, though.


Posted by: A Buddy | Link to this comment | 08-12-11 5:28 AM
horizontal rule
66

But seriously, I was totally making shit up in the group answers. I had no idea.

Yes! In agreement with all who have said this. I would see the group, and by the second question it would feel like I was just randomly picking someone.

I came out right in the middle, 55 computer/54 real faces.

I found myself keying in on more individual features with the computer images, and more having a general "that guy" take on the real faces. Part of that with the computer images ended up being interpreting some of them as "that guy who could be hispanic" or "that guy who's part asian" and such. And then one of them I was calling "the Polish guy" in my head, which I don't think is because I have an idea of a "Polish" face, but rather because I once knew a guy from Lodz who looks somewhat like him.


Posted by: Blume | Link to this comment | 08-12-11 5:46 AM
horizontal rule
67

Caucasian men are absolutely the worst for me to recognize, celebrity or not, so I'm surprised I did fairly decently with them and then bombed spectacularly on the computer-generated faces. I can't actually keep pictures in my mind and just have to sort of hang on to an adjective cloud, so maybe that's what made the real people easier, that they had more unique specifics or something. Beats me.


Posted by: Thorn | Link to this comment | 08-12-11 5:48 AM
horizontal rule
68

We only dress alike.


Posted by: Moby Hick | Link to this comment | 08-12-11 6:13 AM
horizontal rule
69

So do you have some notion about what it might mean that one recognizes a computer generated face more or less easily than a real one, buddy?


Posted by: mcmc | Link to this comment | 08-12-11 6:15 AM
horizontal rule
70

69: Giant nerd.


Posted by: Moby Hick | Link to this comment | 08-12-11 6:16 AM
horizontal rule
71

Nobody's quite sure what the computer generated faces mean yet. It's very hard to get those tests to correlate well with the real faces test (which is sort of the gold standard of face recognition tests). Also (spoiler alert) there are actually a couple of different versions of the computer generated test, and I have no way of knowing which of you took which one. So it's hard to say. That's not terribly informative, I realize. I will say that I'm personally somewhat surprised that there were people who did better on the real faces than the computer generated faces.


Posted by: A Buddy | Link to this comment | 08-12-11 6:21 AM
horizontal rule
72

Caucasian men are absolutely the worst for me to recognize

I've heard people from various ethnic backgrounds say that about people who aren't the same race as them; Chinese people get Caucasians muddled up and vice versa.


Posted by: ajay | Link to this comment | 08-12-11 6:21 AM
horizontal rule
73

I also felt like I kept picking "people" I recognized, rather than specific configurations of face+expression,

This felt to me like sort of a flaw in the study. (obviously speaking from a position of complete ignorance here, so not trying to criticize the scientific validity so much as just the layman's interpretation of "how good are you at recognizing and remembering faces?") Because, usually when that question is relevant, it's in the context of "have I seen this face before?" (Have I seen them on the street? Have I met them?) But here, after I got a few questions into the test on the group sections, I was clearly recognizing multiple faces in the choices, because the "incorrect" choices were being repeated. So, it became less of a facial-recognition test and more of a straight memory test: I know this face, but was this face one of the original six that I saw, or did I see this person before only as a previous incorrect answer? And I guess, sure, it's fair that part of face "recognition" in the real world isn't just "have I seen this face before", but "where?", although the setup here was artificial enough (without any of the usual context that accompanies seeing a face) that it felt like a different exercise.

Regardless, I was 64 (70th percentile) on real faces and 51 (30th percentile) on cg faces. And in both cases I fucking hated the distortions--I wish the scores had been broken down further to allow me to see how I did separately on the distorted vs. undistorted sections.


Posted by: urple | Link to this comment | 08-12-11 6:27 AM
horizontal rule
74

I will say that I'm personally somewhat surprised that there were people who did better on the real faces than the computer generated faces.

The computers were supposed to be easier? I thought they were much harder... I'd assumed deliberately so.



Posted by: urple | Link to this comment | 08-12-11 6:30 AM
horizontal rule
75

The Bart Simpson one was easy.


Posted by: Moby Hick | Link to this comment | 08-12-11 6:30 AM
horizontal rule
76

72: Yeah, but I'm white. I mean, I can recognize my brothers and other white people I know with no problem, but generic white guy (in real life but even worse in photos) reads as generic white guy to me. Movies are terrible if they're all wearing ties and stupid generic haircuts. I think it's probably that I make more mental effort to distinguish people of races other than mine because that does take more work, but I don't know.


Posted by: Thorn | Link to this comment | 08-12-11 6:32 AM
horizontal rule
77

Yes, I found the computer-generated faces much more difficult than the real ones. I had no feeling for them. With the real faces, I could sort of think about an emotion they were having or how I'd imagine their personalities. I also have a hard time with people who have had facial plastic surgery--I really can't recognize them afterward for a while, and have to start all over having emotional content about them.

And yes, the white guy problem is a big one, especially if you're not particularly used to being around white guys. I usually have one or two in a class of 35. I'm about to go teach at a school that has, well, a lot of white guys. Thank God they have a face book that shows the people enrolled so I can learn them ahead of time.


Posted by: AWB | Link to this comment | 08-12-11 6:37 AM
horizontal rule
78

So, it became less of a facial-recognition test and more of a straight memory test: I know this face, but was this face one of the original six that I saw, or did I see this person before only as a previous incorrect answer?

Well, it is called the Cambridge Face Memory Test.

Less glibly, you've settled on something that people struggle with, which is how to test face recognition without also testing face memory, if they're even separable. That said, the reason this test is standard is that it's very strongly predictive of real-world problems face processing/face recognition (with the caveat that your individual results on a test on the internet don't necessarily mean anything about anything). And the repeated faces and the distortions are both deliberate; a section-by-section breakdown would probably not tell you anything you didn't already know: you started out doing really well, and by the end you were just guessing.


Posted by: A Buddy | Link to this comment | 08-12-11 6:39 AM
horizontal rule
79

76: huh. That's interesting. You must absolutely hate war films (our cast includes seventy short-haired white guys in identical clothing!)


Posted by: ajay | Link to this comment | 08-12-11 6:42 AM
horizontal rule
80

stupid generic haircuts

For the most part, white guys with shaved heads all look pretty much the same to me.


Posted by: apostropher | Link to this comment | 08-12-11 6:42 AM
horizontal rule
81

79: Yes, yes, yes! I'm not sure how much of it is pacifism and how much of it is that it's like watching a bunch of indistinguishable plastic army men, but I can't tolerate war films at all.


Posted by: Thorn | Link to this comment | 08-12-11 6:45 AM
horizontal rule
82

For the record, black women who switch wigs/weave often can be hard for me to uniquely identify, too, but at least I pay more attention because usually if there's that much hair changing it involves bad fake hair and I notice that.


Posted by: Thorn | Link to this comment | 08-12-11 6:48 AM
horizontal rule
83

With the real faces, I could sort of think about an emotion they were having or how I'd imagine their personalities.

I didn't read emotion on any of them, but the second part seems a bit like what I meant by identifying them as "that guy". I had several of them in a sort of particular kind of character actor role, although not that well defined.


You must absolutely hate war films (our cast includes seventy short-haired white guys in identical clothing!)

I immediately thought this as well! I find it particularly hard with older movies. Though maybe that just means that newer movies are more likely to have more races or ethnicities in their set of military dudes.


Posted by: Blume | Link to this comment | 08-12-11 6:50 AM
horizontal rule
84

So, I did it again* and got exactly the same (41) on the computer faces. I did much better (51, just a bit below average) on the real faces. Does that mean last night was an aberration or today's score is unfairly inflated with practice?

*There was a box to check for if I'd done it before and I checked it to avoid screwing up the research.


Posted by: Moby Hick | Link to this comment | 08-12-11 6:55 AM
horizontal rule
85

how to test face recognition without also testing face memory, if they're even separable.

What about having someone meet and talk to another person for five minutes or so, and then later (an hour? week? month?) have them attempt to pick that person out of a lineup of similar-looking people (the rest of whom the test subject hadn't previously met)? That seems more realistic.


Posted by: urple | Link to this comment | 08-12-11 6:57 AM
horizontal rule
86

85: And when you write the grant, be sure to mention the pony.


Posted by: Moby Hick | Link to this comment | 08-12-11 6:58 AM
horizontal rule
87

"The Name of the Rose" can't have been much fun either.

82: I have the same problem if it's a particularly dramatic haircut. People with purple hair tend to just get filed as "person with purple hair" and if they then cheat! and change hair colour it confuses my brain.


Posted by: ajay | Link to this comment | 08-12-11 6:59 AM
horizontal rule
88

Retaking the test the expectation would be that you would do a bit better (it's not as reliable a measure the second time around, because you start to learn the faces), but that said your first score was low enough (and enough below your second score) that unless you commonly fail to recognize members of your immediate family it was probably an aberration.


Posted by: A Buddy | Link to this comment | 08-12-11 7:00 AM
horizontal rule
89

||
NMM to the Cherry Pie guy, though I suppose the woman in the video is still fair game.
|>


Posted by: apostropher | Link to this comment | 08-12-11 7:00 AM
horizontal rule
90

88: There's only two other people in my immediate family and they are differ by gender and generation. That's not a very tough standard.


Posted by: Moby Hick | Link to this comment | 08-12-11 7:03 AM
horizontal rule
91

89: One wonders if the woman is even necessary.


Posted by: Moby Hick | Link to this comment | 08-12-11 7:03 AM
horizontal rule
92

85: People have done things similar to that, but it's very time- and labor-intensive. Also, in real situations people tend to use lots of cues -- hair, voice, gait, mannerisms -- to recognize people, not just the face. What's interesting -- and what this test is in part designed to probe -- is that despite all that there appears to be a brain region that is explicitly and primarily for differentiating faces.


Posted by: A Buddy | Link to this comment | 08-12-11 7:04 AM
horizontal rule
93

91 to 90.


Posted by: apostropher | Link to this comment | 08-12-11 7:04 AM
horizontal rule
94

90: Severe prosopagnosics will often report failing to recognize their spouse, or approaching the wrong child at the playground (which goes over well, needless to say).


Posted by: A Buddy | Link to this comment | 08-12-11 7:06 AM
horizontal rule
95

93: It was an MTV promo from back in the day.


Posted by: Moby Hick | Link to this comment | 08-12-11 7:06 AM
horizontal rule
96

94: The people I know with those types of problems were alcoholics.


Posted by: Moby Hick | Link to this comment | 08-12-11 7:09 AM
horizontal rule
97

Moby, we can be twins! Except that I got 49 on the real faces, so you're potentially better than I am.

And I hope I'm not offending any of the white guys on this thread. I'd be able to recognize any of you if we met and then I saw you again, because I actually care about you and have all sorts of things you've said her plugged into my brain already. I just can't manage strangers well.


Posted by: Thorn | Link to this comment | 08-12-11 7:10 AM
horizontal rule
98

Good morning, guys! The sun is shining, the birds are singing, and all my favorite people are here at the bestest blog in all the world! And best of all: I MADE THE NEWS!!!!!

Sure, they spelled "Troll" wrong ("Torroll"? What's that?) but still, I'm famous! And I was under a bridge and everything! Who wants an autograph?


Posted by: Pauly Shore | Link to this comment | 08-12-11 7:14 AM
horizontal rule
99

97.2: I actually try to look as much like a generic white guy as possible, so no offense.


Posted by: Moby Hick | Link to this comment | 08-12-11 7:15 AM
horizontal rule
100

97: It's because we've faced the horrible truth that we are esentially indistinguishable that we are so violent and single-minded in our pursuit of fame and wealth and power.


Posted by: peep | Link to this comment | 08-12-11 7:16 AM
horizontal rule
101

I can recognize my brothers and other white people I know with no problem, but generic white guy (in real life but even worse in photos) reads as generic white guy to me. Movies are terrible if they're all wearing ties and stupid generic haircuts.

Ooh, I'm totally the same way. I've lost count of the number of films where I've been totally unsure of which character's which unless it's really famous actors. This may be one of the reasons I tend to watch films I like again more often than I watch new films, which is an awful habit.

25, 26: Vaguely on this topic, Mind Hacks just noted an Economist story on experimental studies of false confessions.


Posted by: Ginger Yellow | Link to this comment | 08-12-11 7:21 AM
horizontal rule
102

83: I thought things like, oh he looks vaguely sensual and ironic and possibly French, or this one looks a bit like a statue of a medieval saint, or that guy has a goofy look like he's about to be very surprised.


Posted by: AWB | Link to this comment | 08-12-11 7:24 AM
horizontal rule
103

I feel like I've said this before around here somewhere but I can't find it. Anyway, all the freshmen girls at Deaf U dye their hair rainbow colors and then shave their heads, over the course of about 4 weeks every fall. I am shocked and annoyed, every time, when I suddenly can't recognize half of my students, even though I know it's going to happen so I do all kinds of "remember this facial structure, not just the pink and green stripes above it". It's crazy how much taking away the hair disrupts my recognition process.

I bet this practice is why I did so great on the test.


Posted by: E. Messily | Link to this comment | 08-12-11 7:37 AM
horizontal rule
104

100: that's why it's been so difficult to prosecute anyone for the fraud involved in the subprime crisis. All the witnesses are like "yes, I'd recognise him anywhere, he was a generic white guy in a suit" and the detectives sigh inwardly and go back to arresting inner-city gang members who are easier to tell apart.


Posted by: ajay | Link to this comment | 08-12-11 7:38 AM
horizontal rule
105

all the freshmen girls at Deaf U dye their hair rainbow colors and then shave their heads, over the course of about 4 weeks every fall.

No doubt out of solidarity with the trees.


Posted by: ajay | Link to this comment | 08-12-11 7:38 AM
horizontal rule
106

All the witnesses are like "yes, I'd recognise him anywhere, he was a generic white guy in a suit"....

And this is why we must teach the children about pocket squares.


Posted by: Flippanter | Link to this comment | 08-12-11 7:45 AM
horizontal rule
107

Anyway, all the freshmen girls at Deaf U dye their hair rainbow colors and then shave their heads,

They could save on dye if they did it the other way around.


Posted by: Ginger Yellow | Link to this comment | 08-12-11 8:01 AM
horizontal rule
108

104: Imagine the Crimewatch, though. "Did you see a 5'1" quant leaving Canary Wharf on the eveing of August 8? Call us on 0800 555 5555."


Posted by: Ginger Yellow | Link to this comment | 08-12-11 8:03 AM
horizontal rule
109

And this is why we must teach the children about pocket squares.

Pocket squares = gang colours for generic white guys.


Posted by: ajay | Link to this comment | 08-12-11 8:12 AM
horizontal rule
110

Historically speaking, the anomaly is not that British politicians are very close mates with corrupt bankers and hideous muckrakers and blackmailers, but that they are not now, as far as I know, similarly close mates with violent gang members*. Things were different in the days of Queen Anne.


(*except for the Bullingdon Club, of course)


Posted by: ajay | Link to this comment | 08-12-11 8:14 AM
horizontal rule
111

Puts Gordon Gekko's, "Buy a decent suit. You can't come in here looking like this." in a different light.


Posted by: JP Stormcrow | Link to this comment | 08-12-11 8:19 AM
horizontal rule
112

108: Damn dirty quants. Get a real job! Stop telling me how sophisticated Excel is!


Posted by: Flippanter | Link to this comment | 08-12-11 8:24 AM
horizontal rule
113

Damn dirty quants.

You blew it up! Damn you! Damn you to hell!


Posted by: ajay | Link to this comment | 08-12-11 8:31 AM
horizontal rule
114

An then there was the day that to a man they all stood up to SEC investigators and said, "I am Joseph Cassano!"


Posted by: JP Stormcrow | Link to this comment | 08-12-11 8:32 AM
horizontal rule
115

You blew it up! Damn you! Damn you to hell!

"Your loved ones are dead and forgotten for twenty centuries. Twenty centuries! Even if you could get back, they'd think you were something that fell out of a tree."


Posted by: OPINIONATED CHARLTON HESTON | Link to this comment | 08-12-11 8:36 AM
horizontal rule
116

"You follow mortgage-backed securities, you'll get mortgage brokers and mortgage originators. But you start to follow the money, and you don't know where the fuck it's gonna take you."


Posted by: ajay | Link to this comment | 08-12-11 8:38 AM
horizontal rule
117

I didn't do that test, but I had a professional evaluation done for identifying emotions. I'm kind of middle to low middle, but in real life I'm fine when you can see the rest of someone's body.


Posted by: Bostoniangirl | Link to this comment | 08-12-11 8:43 AM
horizontal rule
118

Sure, but naked people tend to have a restricted set of emotions.


Posted by: Moby Hick | Link to this comment | 08-12-11 8:46 AM
horizontal rule
119

Restricted but highly obvious.


Posted by: ajay | Link to this comment | 08-12-11 8:58 AM
horizontal rule
120

117: Yeah, for real people I know, I rely on body (posture, movement) as much as face to recognize them.


Posted by: LizardBreath | Link to this comment | 08-12-11 9:01 AM
horizontal rule
121

||
This is very NSFW, but so awesomely hilarious that you should consider going home over lunch to see it. Art Thoughtz: The Female Gaze.
|>


Posted by: apostropher | Link to this comment | 08-12-11 9:05 AM
horizontal rule
122

Holy shit, apostropher. I'm still cracking up.


Posted by: Megan | Link to this comment | 08-12-11 9:17 AM
horizontal rule
123

He kills me. His YouTube channel is worth a bookmark; there is a lot more where that came from.


Posted by: apostropher | Link to this comment | 08-12-11 9:23 AM
horizontal rule
124

10th percentile for computer generated, 20th percentile for real faces. I am, in fact, terrible at recognizing faces. Names I'm fine with--I'll always recognize the name of someone I've talked to or emailed with. But I can never match them to a face.

The hilarious part? I work in politics.* I meet *a lot* of people. I've gotten good at saying "great to see you" and never "nice to meet you". Because god knows if I've ever met them before.

*Staff, not elected. Thank god.


Posted by: Roadrunner | Link to this comment | 08-12-11 9:32 AM
horizontal rule
125

They're called "obliques", btw.


Posted by: Megan | Link to this comment | 08-12-11 9:32 AM
horizontal rule
126

56 (60th percentile) on the computer generated
47 (20th percentile) on the real faces.

I got the real faces second, and was definitely more tired on that one. My real faces score might have been improved if I got that one first (I assume the order is randomized.)

Did people who did worse on the CG faces get that test second?


Posted by: rob helpy-chalk | Link to this comment | 08-12-11 9:37 AM
horizontal rule
127

"You follow mortgage-backed securities, you'll get mortgage brokers and mortgage originators. But you start to follow the money, and you don't know where the fuck it's gonna take you."

I'm picturing Bunk and McNulty poring over a prospectus and saying "Fuck" a lot.


Posted by: Ginger Yellow | Link to this comment | 08-12-11 9:37 AM
horizontal rule
128

126: I got computer generated faces second, and did better on them.


Posted by: mcmc | Link to this comment | 08-12-11 9:44 AM
horizontal rule
129

127: the series ends on a sour note when Daniels tells McNulty that the investigation of Stringer Bell has been called off; Barksdale Hathaway Investment Management has been classified as "too big to fail".


Posted by: ajay | Link to this comment | 08-12-11 9:54 AM
horizontal rule
130

129: I don't know if this is true of current PBS Kids shows but the Between the Lions dvds we get from the library always note that the presentation is funded by a grant from the Barksdale Reading Initiative and I crack up every time about where Stringer's been funneling funds.


Posted by: Thorn | Link to this comment | 08-12-11 10:53 AM
horizontal rule
131
Posted by: | Link to this comment | 08-12-11 11:39 AM
horizontal rule
132

A bit slow in this thread so I'll put a link to a new mix in my pseud.

Chicken Bones - John Grant
Doctor My Eyes - Jackson Browne
An Old Photo of Your New Lover - The One AM Radio
Shell Games - Bright Eyes
King of Diamonds - Motopony
Me And Lazarus - Iron & Wine
Waltz (Live) - Ian Axel
Will Do - TV on the Radio
Little Red Corvette - Prince
Guess I Lied - The Qualia
Tropicalia - Eliane Elias
The Cold War - The Bruces
Any Other Day - EZ Tiger
We Want You To Stay - Bill Fay

enjoy!


Posted by: k-sky | Link to this comment | 08-12-11 11:46 AM
horizontal rule
133

Forwarded that video on to my friend Chris. He hasn't had time to watch yet, but did just ask me what that acronym I used meant. As acronyms go, "NSFW" is a useful one.


Posted by: Megan | Link to this comment | 08-12-11 4:04 PM
horizontal rule
134

132: Looks good, I'll download that. Thanks, k-sky.

A mix to reciprocate. I know some peeps here will have some of this music, since the raw ZIP file containing it has seen more than a few downloads from my shared account. (If it's not Unfogged I'm not sure who it would be, anyway, since this is the only place I post that address.) Some of it will be new, and who knows but some of it may have appeared on a previous mix, I've honestly lost all track of that. But FWIW:

">Not Just a Face in the Crowd

01. Black Sheep (The Clash at Demonhead)
02. Fly Feather Fly (Donita Sparks & The Stellar Moments)
03. Midnight Rider (The Allman Brothers Band)
04. Good Morning Li'l Schoolgirl (Muddy Waters)
05. Don't Let Me Be Misunderstood (Nina Simone)
06. Shot Me in the Heart (Adrian Younge & The Black Dynamite Sound Orchestra)
07. Tainted Love (Gloria Jones)
08. Cut Me Out (MNDR)
09. 15 to 20 (The Phenomenal Handclap Band)
10. The Israelites (Apache Indian)
11. Guramayle (Gigi)
12. Ayappa Harivarasanam (Tamil Hymn)
13. The Sicilian Clan (Ennio Morricone)
14. Custer Died For Your Sins (Floyd Crow Westerman)
15. Jessica (The Allman Brothers Band)
16. The Free Design (Stereolab)
17. Comfortably Numb (The Bad Plus)
18. Ghost of Love (David Lynch)
19. The Girl With the Flaxen Hair (Isao Tomita) BONUS TRACK
20. Black Sheep (Metric) BONUS TRACK
21. One Night in Bangkok (Murray Head) BONUS TRACK


Posted by: DS |
Link to this comment | 08-12-11 4:47 PM
horizontal rule
135

Whoops, link error. Here it is.


Posted by: DS | Link to this comment | 08-12-11 4:48 PM
horizontal rule
136

"One Night In Bangkok" rules.


Posted by: Moby Hick | Link to this comment | 08-12-11 4:52 PM
horizontal rule
137

One of my favorites. Almost as good as that Floyd Crow Westerman track.


Posted by: DS | Link to this comment | 08-12-11 5:05 PM
horizontal rule
138

I have no idea who that is.


Posted by: Moby Hick | Link to this comment | 08-12-11 5:06 PM
horizontal rule
139

(Pretty much the Native American country star.)


Posted by: DS | Link to this comment | 08-12-11 5:10 PM
horizontal rule
140

k-sky, a password is needed to unzip your files.


Posted by: DS | Link to this comment | 08-12-11 5:16 PM
horizontal rule
141

His pants, however, require no such thing.


Posted by: Megan | Link to this comment | 08-12-11 5:23 PM
horizontal rule
142

83: I thought things like, oh he looks vaguely sensual and ironic and possibly French, or this one looks a bit like a statue of a medieval saint, or that guy has a goofy look like he's about to be very surprised.

I genuinely thought at first that this was meant to be about the guy from Warrant.


Posted by: One of Many | Link to this comment | 08-12-11 6:01 PM
horizontal rule
143

Wait, what? Really? That sucks. Email me if you are having this problem.


Posted by: k-sky | Link to this comment | 08-12-11 6:24 PM
horizontal rule
144

As detailed in 140. Not 141.


Posted by: k-sky | Link to this comment | 08-12-11 6:24 PM
horizontal rule
145

Wait, the guy from Warrant died? Really? That sucks. Email me if you are having this problem.


Posted by: Cryptic ned | Link to this comment | 08-12-11 6:27 PM
horizontal rule
146

Finally just took the test. 54/40th percentile for CG, 61/70th percentile for real.

It helped a lot that the real faces was the second test, and so I knew how the test was going to go. Also I gave them all names like Droopy and Corners and Bright Eyes, and that seemed to help a lot too.


Posted by: heebie-geebie | Link to this comment | 08-12-11 6:36 PM
horizontal rule
147

Once I'd watched enough old movies, I found I was starting to recognize people based on voice. But despite my low, low scores, I generally do ok recognizing people in movies. Partly this is because I worry I won't, and so I'm thinking about it, and partly because I usually pay attention.

I've confused significant characters a few times, but in pretty much every case it was because I was commenting/playing a game on the internet while watching something at home. I can't help it, Bloxorz is way more interesting than The Departed.


Posted by: fake accent | Link to this comment | 08-12-11 6:41 PM
horizontal rule
148

146: "Creepy Sunken-Eyed Voldemort Dude"


Posted by: oudemia | Link to this comment | 08-12-11 6:44 PM
horizontal rule
149

145: People call him "The Cleaner."


Posted by: Moby Hick | Link to this comment | 08-12-11 6:45 PM
horizontal rule
150

hideous muckrakers and blackmailers

I was talking to a Canadian about this recently - "muckrakers" seems to have a much more negative connotation in the commonwealth than in the US. It was a derogatory term when Roosevelt coined it, but even then it didn't refer to scandal sheet stuff like blackmailers. Actually, some of the muckraking magazines helped expose scandal sheet blackmailing.

Muckraking was more like radical reformist investigative journalism; Roosevelt just thought they were getting irresponsible and going too far (possibly even toward socialism--the horror). Even now, among Americans, it seems associated much more with establishment-style investigative journalism rather than tabloid-style stuff.


Posted by: fake accent | Link to this comment | 08-12-11 6:48 PM
horizontal rule
151

148: We must have had the same version!


Posted by: heebie-geebie | Link to this comment | 08-12-11 6:57 PM
horizontal rule
152

"Pouty Lips Dude" was the one I felt most confident about getting. But to illustrate my real powers of observation in action, here's where I confess that I had no idea that one set was real and the other computer-generated.


Posted by: JP Stormcrow | Link to this comment | 08-13-11 6:37 AM
horizontal rule
153

I just took another test on the same site -- emotion discrimination and facial recognition. I came out slightly better than average on identifying emotions, and still twentieth percentile on recognizing faces (in this test, you see two pictures with different facial expressions for half a second each, one right after the other, and have to say same person or different person. Surprisingly hard.)

I wonder if I'm antisocial and misanthropic because I'm so bad at recognizing people, or vice versa?


Posted by: LizardBreath | Link to this comment | 08-13-11 7:12 AM
horizontal rule
154

I'm so misanthropic that I think being good at recognizing people would make one more antisocial.


Posted by: JP Stormcrow | Link to this comment | 08-13-11 7:28 AM
horizontal rule
155

I am loving these mixes! Thank you k-sky and DS!


Posted by: ursyne | Link to this comment | 08-13-11 8:13 AM
horizontal rule