Why "african americans" can never excel at anything relative to the white majority : they can't be the smartest, nor the poorest, nor the best nor the neediest nor the anything else-est. Not ever.

Saturday, 10 March, Year 10 d.Tr. | Author: Mircea Popescu

We've been discussing statistics now and again (and again), but apparently nothing's harder "to understand" for common folk than plain, obvious and common sense items they do not wish to even consider, let alone grok.

Let's go again : black people living in the United States can not be the smartest, poorest, best, neediest nor dominate any other normally distributed category of performance however defined because there's just not enough of them.

That's it, and it's sufficient. Here, with visual aids :


The large white space eating the tiny black space represent the relative populations : there's a factor of about 8 separating white and black populations ; and there's a factor of about 8 separating white and black surfaces.

Because the black distribution is so small, you need roughly speaking three standard deviations to match performance equivalent to one single standard deviation in the larger distribution. This is to say that in order to find a black guy who is better than two thirds of the general population at any given task -- being a fireman, or being indigent, or being just about anything else -- you are looking for a black guy that is better than 99.87% than all black guys at the same task.

Yes, you read that right : only the one-in-a-thousand black swan can compare on equal footing with the one third of the general population. You need the best black guy out of one thousand black guys to get a black guy that can compete on equal footing with a white guy that's the best of... a random sample of three dudes.

Which is why (for instance) having anything but white people on, say, Stanford's list of full scholarships for reason of anything (academic performance, economic indigence, anything whatsoever) this year (or any other year) is strictly obscene, as the white guys are not only the most deserving academically and by such a large margin no Ivy league could possibly admit anyone else in any sort of fair contest -- but they're also the neediest, by equally such a large margin that idem.

Where's your "science" now ?

Are you willing to admit it's just bullshit you made up with no basis in reality whatsoever ? Or not just yet ?

Who's gonna provide "just the facts" carefully curated of any actual facts so the sort of rank nonsense you swim in can continue ?

Hm ?

Comments feed : RSS 2.0. Leave your own comment below, or send a trackback.

25 Responses

  1. Nonsense. A randomly chosen black will deviate with the same probability as a randomly chosen white. Assuming that race and performance are independent (as I may, you said any property), the proportion of blacks in people over 1SD, 2SDs, ... will be exactly the same as their proportion in the entire population. Apply Bayes and see for yourself.

  2. Mircea Popescu`s avatar
    Mircea Popescu 
    Saturday, 10 March 2018

    This is entirely false : an item chosen for property P will not be found as frequently an item' chosen for property P and property Q. As the property "being black" already imposes a 1:8 improbability upon occurence, you will find a competent black policeman ~for every eight competent white firemen you find. Apply whatever you want, there's no such thing as magical pantsuit-fairness.

  3. How can it be entirely false when you just stated pretty much the same thing?

  4. Mircea Popescu`s avatar
    Mircea Popescu 
    Saturday, 10 March 2018

    Well, it's not ; I just misread what you said.

    Anyway, the problem remaining is that both our comments discuss an imaginary world where being black is of the nature of "your ssn ending in 6" or "being born on a tuesday".

    This is eminently not what the pantsuit party claims ; instead they claim "african american" is a group. If indeed this is true (which evidently it is not), then the above discussion in the article exactly applies, and my comment will have to be ammended from

    you will find a competent black policeman ~for every eight competent white firemen you find

    to the finer

    you will find a marginally competent black policeman for maybe every third marginally competent white firemen you find ; but you will find middling competent black policemen extremely rarely, and anything better than that absolutely never.

    This still neglects important long tail effects, which account for interesting natural properties such as "you will not be finding any stars in chemistry labs, notwithstanding that balls of hydrogen of various sizes are often found there", ie, there will never be such a thing as an "african-american renaissance" or industrial revolution, or anything else worth the mention. There is after all a reason the nightlife of Columbus, Ohio is not 10% of the nightlife of New York, notwithstanding that the population is.

  5. I'm confused by the graph here. If white and black populations had the same distribution for a trait, then your graph should be a vertical scale only: the tails are equally long. So maybe you are saying something different?

  6. Mircea Popescu`s avatar
    Mircea Popescu 
    Thursday, 22 March 2018

    What does this mean, "vertical scale only" ?

  7. So again, I'm assuming the probability distribution of whatever trait is the same in whites and blacks. Let's say the distribution is T(x). If you plot them where the vertical axis is probability and the horizontal the trait, they overlap perfectly. Now to get the distribution of the trait in the populations of magnitude W and B, you just multiply for WT(x) and BT(x); a vertical scale, where the vertical axis is now population. If you take any piece of the trait, say between 2 sigma and 3 sigma, and divide the integrals of WT(x) and BT(x) taken over the piece, you're left with W/B. This is exactly what xuqtvifr is saying above. But this is in conflict with your graph.

  8. Mircea Popescu`s avatar
    Mircea Popescu 
    Friday, 23 March 2018

    In normal distribution of a population, the vertical axis is count and the horizontal axis measured value. Because the counts of the African-American population are smaller than the counts of the white population, the height of the curve at all measured points will be shorter.

    Try numerically, if that's not obvious : 1.5% of every population is going to be perfectly mediocre. Thus the middle line for white people (all say 300mn of them) is going to be 4`500`000 units tall, whereas the middle line for African-American people (all say 22 mn of them) is going to be 330`000 units tall -- ie much shorter. Going further out, if exceptional performance however defined is 1 ppm, then the white line at that spot will be 300 units tall, whereas the African-American line 22 units tall. And even further out, where the concentration is less than the total population, the African-American line is not going to be at all, whereas the white line is going to be however tall it gets to be.

    See ?

    (If you don't follow the pantsuit idiocy, and agree there is not nor can ever be such a thing as "African-American", but merely black people, then the image drawn can not exist -- the only proper draw of the distribution is a single, somewhat grayer curve. Which is the point -- that by believing the Clinton & democrat gargle, a person whose skin happens to be black is doing themselves the worst disservice conceivable -- amusingly just so today as 200 years ago.)

  9. Given that the probability distribution of the trait is the same standard normal in the white and the black population, the population distributions of the whites and blacks looks like this. Your graph is wrong on the width of the distribution, (notice that mine is a simple vertical scale), which is leading you to think this:

    Because the black distribution is so small, you need roughly speaking three standard deviations to match performance equivalent to one single standard deviation in the larger distribution.

    Again, assuming the probability distributions are equal, this is incorrect. To match the performance for one standard deviation in the larger population, you only have to be above one standard deviation in the smaller.

    You also said above:

    Anyway, the problem remaining is that both our comments discuss an imaginary world where being black is of the nature of "your ssn ending in 6" or "being born on a tuesday". This is eminently not what the pantsuit party claims ; instead they claim "african american" is a group.

    Do you mean this to imply that the "african american" group has differences in probability distribution for the relevant traits? Because if they don't, then being african american is exactly like having your ssn end in 6.

    On the flip side, if the probability distributions for all relevant traits are the same in whites and blacks, then the pantsuits have no ground to stand on for calling blacks a discernible group, just as they have no ground for "people with ssn ending in 6" being a group. Of course, the distributions are not the same on many traits (wealth, incarceration, education level, etc.), but I'm not seeing it from tail effects.

  10. Mircea Popescu`s avatar
    Mircea Popescu 
    Sunday, 25 March 2018

    Except in your graph, the proportion of mediocre blues to exceptional blues is MUCH smaller than the proportion of mediocre reds to exceptional reds. Your graph is a) incorrect and b) politically motivated (it translates in fanmaths the proposition that the blues are substantially better than the reds, ie that it takes fewer of them to hit a certain exceptional objective standard). This may of course be your political slant, but you don't get to claim it has anything to do with statistics.

    In order for the populations to be of the same kind, the gaussian WILL HAVE THE SAME CURVE.

  11. > amusingly just so today as 200 years ago

    I thought you blamed Lincoln for that.

  12. Mircea Popescu`s avatar
    Mircea Popescu 
    Sunday, 25 March 2018

    Let's delve, why not.

    Today's pantsuit discourse, in the universal and unchanging pantsuit manner, purports to enact the false "elect a can of soda" choice to all vulnerable minorities -- be they foreigners, blacks, women, whosoever. TLP calls it "accepting the form of the question while debating the results", but it's the same thing : as long as you accept it's your job to "pick" between two cans of soda, no harm is going to come to them out of you picking either one.

    In this vein, they're telling women [that are dumb enough to go for it] that, literally, "Hey, women! Did you know that before the Pantsuited Hilarity gave you your civil rights, you were living in slavery ?"

    Entirely predictably, they're also telling blacks that

    Hey, this group over there may be claiming that the tiny black gauss curve representing you is actually slid to the left wrt to the greater white curve! Whereas we, only right and true goodpeople agree with you that the tiny black gauss curve representing you is just as perfectly alligned with the much greater and better white curve as you'd like! (Except of course for the portion of "renegades" among us who claim that the two graphs are distributed randomly so for some criteria [such as being smart] the blacks will center more to the left whereas for some other criteria [such as running really quick and fucking filthy white bitches] it's centered more to the right, but even they agree it all evens out in the end.) We're the only ones that love you, and so much so some of the uglier among us are even willing to take one for the team by fucking some of your womenz! (Notrly the other way around aaaaite ?)

    The problem with this falsehood (other than its tendency to feed niggers their delicious fried chicken, of course) is that if you believe it, you lose. Mental representations may not be capable of altering objective reality, but they sure as fuck are capable of altering the actor's own behaviour, and identity is behaviour over time and nothing more. Consequently if you believe "this is everything", like the pantsuit always and everywhere believe, you do end up needing the pinoy to come discover Zanzibar, because crossing that narrow strait's too much strain for you.

    The amusing happenstance with the falsehood, however, is that it's exactly the same structure as the nonsense put forth by the exact same pantsuit in the mid 1800s. Make no mistake about it, there is no such thing as "Southern" Democrats. The Democratic party of 1800 is, exactly and entirely same, identical to the Democratic party of 2000 : where reason goes to die, a collection of volitively stupid marauding idiots. Volitive in the same ways, stupid in the same ways, marauding endlessly towards the same goalposts on the same exact field.

    So, the position of Republicans wrt slavery was both rational and correct, in the mid 19th century. The position of Democrats -- all of them -- was batshit nonsense, proudly pseudoscientific, offensively stupid. Then exactly as now. But admire the political expediency -- the Democrats feigned a "split", which is what they always do. It's the ultimate disavowal, the equivalent of victorian priss "passing out". This put the minority Republicans in charge, and Lincoln made the unutterable mistake of delivering Democrat goals at his own life's price. Yes it was correct to be against slavery as practiced and self-justified in the South (primarily because of the fugitive clauses -- you're more than welcome to keep your women as chattel, but you're not fucking welcome to impose on my time or my tools to "take back" your "runaway" women, derp) ; but it was deeply incorrect to be against States' rights and for centralization.

    This was the hidden gambit there : the idiot Democrats yakked and yakked for three decades, entirely by and among themselves. Then a rational minority was "forced to action", but in a context and on the terms "forced" upon them by the entirely spurious spam of the previous decades. Then an irrational (and utterly clueless) minority rose up to oppose what problems ~they perceived~ (on the basis of the exact same Democratic yak) with the proceedings of the Republicans. Then the two clashed, as they predictably would have, and nobody got anything out of it -- the black man suffered worse after than before ; the economy didn't come out of the shitter until gold was found in California towards the end of that century (and had gold never been found...) and the Republic was lost forever. The only winners, of course, the miserable pantsuit scum.

    This is the problem there -- and I very much doubt Lincoln failed to appreciate his tragic situation. The Republicans thought they're fighting "against slavery" on the extremely flimsy basis of the Democrats having stuffed the report of the Dec 1860-Jan 1861 "committee of the 33" 100% full of nothing but slavery nonsense. Buying into the "representative" nonsense, they defensibly if incorrectly decided hey, "the South is all about slavery". The South wasn't, but it did serve the Democrats' international aggenda, so those of them living in the south happily sold their constituencies for "the greater glory" -- but the greater glory of the pantsuit mothership. Meanwhile the people who actually fought that war, on the Southern side, didn't either care or even read the Democrat agitprop. They were interested in entirely different points, such as free trade (ie, no "Commerce Clause"), no tax-and-spend, etcetera.

    So there you have it : I blame Lincoln for doing the stupid, but I'd gladly hang any US Democrat for no further cause than simply that. Questions ?

  13. Except in your graph, the proportion of mediocre blues to exceptional blues is MUCH smaller than the proportion of mediocre reds to exceptional reds.

    But this isn't correct: since all I did was vertically scale one graph to another, the proportions are the same.

  14. Then you see no problem with slavery?

  15. Mike Truck`s avatar
    Mike Truck 
    Sunday, 25 March 2018

    Here is the tru story esthlos: Racetruth.jpg.

    In yellowish, the numerous but very deeply mediocre Chinese. In black, the few and similarily inept African. In blue, the not so numerous but much, much, much more diverse White. We base this on ten thousand years of history, during which five hundred generations the Chinese and the Africans happily rediscovered how to do nothing at all every third or so. That's why they're not only narrower but actually the extremes of their performance are documented on the graph. The blues however, you don't know how far they can go, and likely never will.

    Your graph is bullshit not only because you have nfi what e^-pix^2 is, but because you then take the line fitted to white and pretend it's the niggers. It's not the fucking niggers, aaaaite? Stop going around in whitey's cars and boats pretending like your rasta skunk had anything to do with it.

  16. Mircea Popescu`s avatar
    Mircea Popescu 
    Sunday, 25 March 2018

    My my this is a disputatious topic.

    @esthlos And here's a picture of a woman and her vertically smushed self / midget sister :


    Clearly they have the same proportions as all I did was altering the proportions, hurr. What are you, some kind of nut ?

    Take remedial geometry.

    @nbd I don't see a problem with slavery in general, obviously, which is why I practice it. I see a problem with anything/everything else (admitting these are distinct), which is why I don't practice "something else".

    I do however see two problems with the dumbfoundingly unexamined implementation of slavery as seen in the period Southern US (and more generally everywhere else, except perhaps the Romans, as a very contentious and altogether dubious point) : the foisting and the unprincipled exception.

    As to the first : it is batshit insane to expect that you can ask me / society in general to manage your ineptitude for you. If your slave runs away, it's your fucking problem. Make it so they don't, however you do that, and get out of my hair ; or else learn to live with the running -- but in any case, get the fuck out of my hair.

    As to the second : it permeates the intellectual history of the antebellum South (such as it barely was), a fine example being found right above -- these fine gents were for low taxes and no public services, but still expected the state to chase their runaways for them. Well hurr fucking durr, which is it! This nonsense rises to the level of world-ending where they invented a notion of "wife" that's somehow not held exactly the same as the slave. Why the fuck not have the exact same arrangements for the women and the servants ? And don't fucking tell me "they had children with the wives", jesus christ they mostly did not.

    Notice how "optionality" doesn't enter into any of this -- as it well fucking shouldn't, for the simple reason that it can't. Not anymore than the boogeyman can bugger your wife, at any rate.

    @Mike Truck Honestly, I don't see how you can have differently fitted graphs for "different" human populations, but whatever.

    Nice name tho ; and desmos sure kicks ass dunnit.

  17. Mike Truck`s avatar
    Mike Truck 
    Sunday, 25 March 2018

    > desmos sure kicks ass dunnit.

    Did the colors gave it away?

    > I don't see how you can have differently fitted graphs for "different" human populations

    Can you see how you can have differently colored groups called "the same species"? If it happens to ducks you call them Cygnus cygnus and Cygnus atratus but when it happens to humans suddenly everyone's coscoroba.

    Niggers aren't the same species, that's all there is to it.

  18. Re: "slave runs away, it's your fucking problem" -- this AFAIK is actually how the slave territories worked, from the very first slave boat unloading, ~until~ the 1850 "Fugitive Slave Act". The latter, quite arguably, was a "TILT" in response to "one too many kicks of the dog."

    The abolitionist strategy was, approx, "make scorpion bite himself." And it worked. Which is why the same algo is used to this very day, by the same people, against every other organized resistance to pantsuit.

  19. Mircea Popescu`s avatar
    Mircea Popescu 
    Sunday, 25 March 2018

    @Mike Truck Lol. Yes, the colors gave it away.

    Anyway, speciation in humans is a topic I'm not really interested in discussing. Yes you could take the scientific view and declare all the major haplogroups h.h. subspecies, even genuses if you insist. If Anatidae can be Anser whatever and Brants bernicla and so on then certainly Hominidae can be Alb Gothic and Bantu blabla.

    What'll that do ? Just move the whine parade exactly as it is now into a discussion of these new species you enacted. That's the problem here, that disputes about human subgroups are never ever scientific, nobody is ever interested in finding out anything, nor does anyone perceive subjective cluelessness as any kind of a barrier to enter.

    They're always political, everyone got something to say because everyone wants things to be a certain way, so they'd just move right on with their tiresome windbagging on new terminology without batting an eyelash (maybe even thank you for giving them a whole new pile of words to yak and yammer). Meanwhile the psychology of why the fuck all the idiots want to "participate in this discussion" is so fucking intricate and so fucking not worth the time... it's a sort of "which soccer team is better", carried in the context and with the means available to 1980s Romanians. It really and exactly is, and I'll be fucked if I'm going to entertain the nonsense.

    (Cygnus isn't ducks by the way. Yes, yes, it's funny, I get the point. I'm just stuck, gotta say something for the young'uns. Congrats for writing the most-rakes-left-for-MP-to-be-forced-to-step-on-later comment in the annals of this here lemma, I guess.)

    @Stanislav Datskovskiy Myeah. You can't have a Republic if you won't defend it ; and the defending can take mighty incredible and unexpected forms.

  20. Clearly they have the same proportions as all I did was altering the proportions, hurr. What are you, some kind of nut ? Take remedial geometry.

    Poor choice of words on my part. Yes, the proportions of the graph have changed. I meant the proportions of the center to the tails, in the following way: To get the ratio of the extremes to the mediorce, you divide the integral of the extremes by the integral of the mediocre (duh). Vertically scaling the graph doesn't affect that quotient.

    Smybolically: int_A f / int_B f is equal to int_A cf / int_B cf

  21. Mike Truck`s avatar
    Mike Truck 
    Sunday, 25 March 2018

    Could say the same about scaling it horizontally instead. But you didn't, did you. Why not?

  22. "One thing must be perfectly apparent to every intelligent man. This abolition movement must be crushed or there is an end to the Union."
    Pierce identifies pantsuit "abolitionists", 1835.

  1. [...] know whether Mr. Lincoln to-day stands, as he did in 1854, in favor of the unconditional repeal of the Fugitive-Slave law. I desire him to answer whether he stands pledged to-day, as he did in 1854, against the admission [...]

  2. [...] the point remains : given a normally distributed group, the average of the elite (under any relevant criterion) will far surpass the average of the [...]

  3. [...] needed for whatever nonsensical purpose any of five thousand white boys'd have done just as good (in fact better) at. So yes, I understand white people setting themselves to making black people appear like actors [...]

Add your cents! »
    If this is your first comment, it will wait to be approved. This usually takes a few hours. Subsequent comments are not delayed.