Follow Slashdot blog updates by subscribing to our blog RSS feed

 



Forgot your password?
typodupeerror
×
Math

Journal Journal: My Bias: The Fallacy of "Weirder Shit has Happened"

I was thinking today, that I am guilty of in the past using statistics wrong. More specifically, I accepted "weirder stuff has happened" as a valid reason why I could at least give the time of day to something. The fact is, that it's not at all a reason to do anything. Here's an example

Let's say, Politician X has a .0001% chance of doing something so incredibly stupid that only one in a million people in the same position given freedom to do that action would do it(logical necessity). Suppose then, that Politician Y has a .001% chance of doing something that only one in one hundred thousand people, given freedom to do that action, would do it(again, logical necessity). My reasoning would go as follows;

X did action(X,t,P)
P < P2
:. action(Y,t+k,P2) is reasonable

However, as you can see when it's put in syllogistic form, there is a lot of missing conditions necessary.

Let's make a simpler example. Let's say you are flipping coins, and you flip a properly weighted coin 20 times in a row and it turns up heads. At this point, I could come up to you and bet you 20$ that sometime during the next 100 flips, there will be 10 flips in a row that are all heads. The problem is, there probably won't. The chances of 10 heads in a row in any 10 flips is 2^-10, so in 100 flips, well, you don't have to do the math to know full well that it's not that likely; it's a stacked game. So why is it that when dealing with world events, that it's different?

One possible reason is that sometimes the actor is the same. Just like, when you are not sure whether or not the dice is loaded, and you get 10 6's in a row, you can start to imagine that the dice is loaded, In other words, expressing this statement (stranger things have happened) does not express anything about the odds of the event, it expresses an opinion about the characters in question, but doing so in an obfuscated way that avoids criticism. Is it true that Politician X is an idiot of epic proportions? Does he have a hidden agenda? Perhaps; but does it follow that because X has a hidden agenda that Y does? What if Politicians X1 through X50 all do stupid things(such as claiming to be for family values and then getting caught with his hands in the underage male prostitute cookie jar). Can we start to infer something about the group in general? At what point? Obviously there's no perfect answers, but it's worth pointing out, lest any of you fall into this pattern of reasoning. I'll try to avoid it from now on.

This post has been inspired by Overcoming Bias, and I wouldn't be surprised if they have an article on it already.

Desktops (Apple)

Journal Journal: Human Computer Interaction (HCI) is Pseudoscience

HCI is pseudoscience. Or at least, the stuff that my prof is teaching is pseudoscience. Now, maybe the field of HCI is scientific, and I'm being taught wrong, but it is becoming clearer and clearer as I progress through this field that it is not a science. In fact, the very first sentence of the textbook makes the claim that the field is Not a science, (nor an art, nor does it contain yet the discipline engineering).

Let's see.. At first, I was just kind of unhappy that this course had managed to get 'Computer Science' attributed to it. But it's deeper than that; since there is no objective standards in the field, research into it seem doomed to purely subjective, and inaccurate results. That being said, of all the non-scientific fields to go in, helping people work with computers better is certainly not the worst to go into; but for SERVER's sake, get with the 18th century and join the scientific revolution, or at least drop the pretense that this has anything to do with computer science.

  • the beginning of the course started with a 'there are no facts in this course, no objective truth.' That ought to be a good hint but I refused to believe what I saw.

This list will grow as I think of things. Too bad comments are shut down early. It's kind of sad, in a way. HCI finds gnu/linux, and free software in general very praiseworthy as a thing to spend time doing. Perhaps one of the five-ish pillars of my belief in free software is HCI based; but that pillar has essentially been toppled. Thankfully the others are on much firmer ground.

It's funny.  Laugh.

Journal Journal: On the good side...

On the upside, I figured out where the settings for overclocking my desktop, and got it up to 1.7 GHz, from 1.2 GHz

On the downside, it no longer boots :(

I wonder if it's been permanently fried. Last time it didn't work at all, but since then I put a much beefier power sink/fan setup on the cpu, and have cleaned/closed the case.

Or maybe it was the power outage which directly preceeded my toying around with it(funny, that); I didn't doublecheck to make sure that it actually could boot before trying it...

Update: On the good side, the problem isn't heat. The CPU tempurature is running in the low 50's. Which is definitely warm, but only around as hot as it usually is during the heat of the day(it's a canadian night, so I have cooling coverred). I changed the max tempurature to 66 from 70. Something else is causing this. I can sort of get into the BIOS; it seems very flaky. Software bug? Some sort of cpu skew?

Update: still not sure the cause, but at least things are working(now that overclocking is off...)

Hardware Hacking

Journal Journal: Homemade Battery a Success!

I just got my voltemeter working, and tested my homemade battery, and sure enough it works!
Homemade Battery
Required: 2 plastic containers(I think I used old honey/margarine containers), 2 empty glass jars(orange juice containers worked), two seperate metals (I used lids from coffee containers(some tin alloy?), and aluminum(?) foil, vinegar/salt+water (vinegar/salt is probably optional, but supposedly helps).
Place the two glass jars in the middle of the two plastic containers, and fill the plastic containers with the vinegar/salt solution. You may need to weigh the glass jars with water so they don't float away. In one end of the container place the first metal, in the other, place the second metal. Do not allow the metals to touch.
Sure enough, I got .2 V. Not much, but I don't think my choice of metals was very good. The point is that there is voltage running.
Now the next question; what can actually use .2 V?
Slashdot.org

Journal Journal: Slashdot Network Circle Game

I just thought of a game to play, and successfully finished a round(I think, anyway).

Next time you reply to someone on slashdot, see if you can identify who you are replying to. See if they know anyone who knows anyone who knows anyone who ...etc knows you, and vice versa.

curing boredom, one mind at a time
User Journal

Journal Journal: 16777216 1

The 16777216th slashdot comment is comming up. Keep your eye out :D
User Journal

Journal Journal: TRIPS, Canada, 1998, the EC and WTO trade Dispute 153 (

Interestingly wink has been more helpful than Google in this search.
This Trade Dispute needs a better website. Right now the only comprehensive website that I can find on it is the WTO's own site on the matter, and by the looks of their site the issue has not been resolved (is this possible?)

The idea as I understand it is this: The EU notices in around 1997 that Canada's Patent Act is not up to spec with TRIPS, which aparently Canada signed. So Canada
a) fixes the Patent Act to comply
b) and files a dispute with the WTO regarding the European Commission's patents, and how they aparently discriminated on the basis of technology in regards to patents.

Then...the story stops.
What happened? Google has some papers which I'm digging through but it's mostly on stuff that relates to this but not in a direct way.
Space

Journal Journal: g 1

Using a variable length pendulum(duct tape, watch and sewing thread thread) and a stopwatch, I took the following data

f(s^-1) L(m)

0.767 0.396
0.767 0.396
0.733 0.396
0.733 0.396
0.733 0.396
0.733 0.396
0.767 0.404
0.767 0.404
0.733 0.404
0.733 0.404
0.733 0.404
0.733 0.404
1.000 0.190
1.033 0.190
1.000 0.190
1.033 0.190
1.000 0.198
1.033 0.198
1.000 0.198
1.033 0.198
1.033 0.192
1.033 0.192
1.033 0.200
1.033 0.200
0.600 0.615
0.617 0.615
0.633 0.615
0.600 0.615
0.600 0.615
0.600 0.623
0.617 0.623
0.633 0.623
0.600 0.623
0.600 0.623

which according to my calculations means g should be here at 8.6 +/- .2 There might have been air resistance going on around my pendulum, and resistance going on between where the pendulum attatched to the table. Outside of that I'm not sure.

Slashdot.org

Journal Journal: Themusicgod1 finds girlfriend through Slashdot 1

Thank you slashdot, without you I'd still be single and lonely. As I theorized earlier, there are single women on slashdot and a few of them might just be interested in you. As it stands now, I'm happily coupled with a blogger-geek named tricksy, whom I adore. No this isn't a joke. Now comes the tricky part: keeping my entire life from falling apart, and keeping her head over heels in love. Feel free to comment :)
Censorship

Journal Journal: bugfix - divert around slashdot censorship

"The Internet views censorship as damage and routes around it"(Gilmore, n.d.)

Too many people I know on slashdot have had their ability to post suspended due to their political direction. Enough is enough. I'm going to try to make a change here.

If you cannot post something on /. due to account or subnet blacklisting, fear of account or subnet blacklisting, or fear of reprisal from the legal system of your country of residence, or any other country, etc, email me and I will post it: recursive.genepool at gmail.com.

I reserve the reserve the right to rot13 it (or base64...ascii art will be a lot less goatse in base64), and/or post from an account other from this one. I will do my best to post it. Probably with some sort of sig ----- THE FOLLOWING IS NOT MY OPINION IT HAS BEEN FORWARDED TO ME SEE MY SIG (url) ---- (hrm...better get on that rot13 program)

After first thinking of this idea, I realized: Oh god, I now have a moral obligation to post to slashdot; To propegate this sig/meme. Oh well. For those of you out there reading this please forward it along to other people in case they are having issues with being blocked for their activities here on slashdot. I will not watch idly as this society censors unpopular opinion...this forum is becoming too prominent in technology and F/OSS and too important to give up yet.

Oh and yes I will crapflood...but I will crapflood very...very slowly so don't even bother. If you flood my email you can pretty much guarantee I won't get your message.

The Almighty Buck

Journal Journal: Death and working your ass off

syntax:
10,000 is 10 thousand with 2 significant digits.

Canada
630000 million $ debt
33 million people
19000$/capita debt
2900 hours @ saskatchewan minnium wage
2700 hours @ my current wage

saskatchewan
6650$ million $ debt
1 million people
6650$/capita debt
10,00 hours @ saskatchewan minnimum wage
950 hours @ my current wage

Total:
3900 hours @ saskatchewan minnium wage
3700 hours @ my current wage

That's a lot of hours and it's just the cost that is handed down from my parents generation to mine.

I'd provide source links but I'm not at home so I'm kind of stuck for resources on a shitty first generation iMac.

Slashdot.org

Journal Journal: 42,500 single women on slashdot 7

idea
(gack! webserver dies!!)

assumptions :
that my friends/fans are not biased to any gender
that they aren't lying, or that it isn't blatantly obvious which gender
they are

female 4
male 18
unkown 15
total 37

pihat=4/37=.1081

confidence 65%
1-alp=.65
-alp=.35
alp=.35

Z(alp/2)

[phat-Z(alph/2)(sqrt(phat(1-phat)/n)),phat+Z(alph/2)(sqrt(phat(1-phat)/n))]
.1081-Z(.175)sqrt(.1081(1-.1081)/37)),.1081+Z(.175)sqrt(.1081(1-.1801)/37))
( .1081-.45sqrt(.1081(.8919)/37),
.1081+.45sqrt(.1081(.8919)/37) )

assuming sample is random and unbiased

.1081 +/- .0230
ie
the population of females is 10.81% +/- 2.30%,
within 65% accuracy

confidence 65%
alph=.35
phat =15/37=.4054

P= .4054+/-.45(sqrt(.4054(1-.4054)/37))
.4054+/-.45sqrt(.4054(.5946)/37)
.4054+/-.45sqrt(.000651489)
.4054+/-.0115

so if we add the uncertainty within the total
to the uncertainty of the
population of females we get...

10% +/- 43.99%, within 65% accuracy

now assuming that slashdot's total population
is 850,000
there should be
85,000 +/- 373915 females, 65% accuracy

assuming that http://www.nothingisreal.com/girlfriend/
is on the right track
then there should be
42,500, +/- 186958 single females, ~65% confidence;

"that is not a small number"-strongbad. So...what do you think? Am I full 'O shit? On the wrong track? thoughts?

are YOU one of these 42,500? Email me at themusicgod1@jabberwocky.hn.org with a breif hello and I could be yours for the low cost of a few dozen emails and a plane ticket to whatever country you live in.
Slashdot.org

Journal Journal: weaning off slashdot not successful

One of my new years resolutions was to post to slashdot less. Not to quit, but to post less. This hasn't been going well. Oh well, could be worse. I could be committing suicide or something.
Slashdot.org

Journal Journal: Public Key

If you have a public key on /. Can the public have access to it? Is there a link somewhere in slashdot which leads to a users public key? I'd like to get into pgp/gpg/etc more but, honestly I don't think I know anyone who is into it and that kind of limits its usefulness. I noticed that slashdot has a public key field but I can't for the life of me find either my own or other peoples keys.

Slashdot Top Deals

"The one charm of marriage is that it makes a life of deception a neccessity." - Oscar Wilde

Working...