Want to read Slashdot from your mobile device? Point it at m.slashdot.org and keep reading!


Forgot your password?

Comment: Re:We're so screwed. (Score 1) 237

You're technically correct but they are only selectively dependent. If you don't die, the fact that you didn't die that year does not change the likelihood that you will die in the next year. This fact is sufficient for P(X and Y) = P(X) * P(Y) to hold for us because we are looking specifically at the probability that you DON'T die from terrorism over your lifetime. We make the dependency when you DO die irrelevant. That's why I took the path I did to answer the question "What are the chances you'll die from terrorism if you would otherwise live 75 years?"

To illustrate this, lets simplify things and take a look at a classic example: picking colored marbles from a jar. We are going to take a step back from the individual perspective and see what the probability is that a specific marble will get picked out of a larger population.

You have a jar with 3 marbles in it. 1 marble is yellow, {Y} (representing you) and 2 marbles are orange, {O1} and {O2} (representing other people). Every year, 1 marble is removed from the jar by a terrorist (representing death). And at the end of every year, 1 orange marble is added to the jar (representing someone else being born).

In year 1, there are 3 possible outcomes. 1 where the yellow marble is chosen. 2 where an orange marble is chosen.

Outcome 1: {Y}
Outcome 2: {O1}
Outcome 3: {O2}

At the end of year 1, orange marble {O3} is added. In year 2, there are 9 possible outcomes. We can apply the formula from my previous post to this. What does it say?

P(yellow will be picked over 2 years) = 1 - ((1 - (1/3))^2) = 1 - ((2/3)^2) = 1 - 4/9 = 5/9

So we would expect 5 of the outcomes to have yellow picked. Here's a table of the outcomes:

O 1 {Y}__ {O1}
U 2 {Y} __{O2}
T 3 {Y} __{O3}
C 4 {O1}_{Y}
O 5 {O1}_{O2}
M 6 {O1}_{O3}
E 7 {O2}_{Y}
# 8 {O2}_{O1}
_ 9 {O2}_{O3}

As represented by the math, there are 5 out of 9 outcomes in this table where yellow was picked. You can take this to 3 years and beyond. At 3 years, we would expect to see 8 in 27 outcomes where yellow was not picked. This makes sense if you look at the table above because outcomes 5, 6, 8, and 9 are the only ones that can generate a new outcome where yellow was not picked. In the new table, they generate 2 such outcomes each and 2 * 4 = 8.

Hopefully that clears things up a bit.

Comment: Re:We're so screwed. (Score 1) 237

Here's what's wrong: my value is a 0 to 1 probability. Yours is a 0-100%. So my overestimation is actually close to yours. That being said, let me explain why I'm not asking how many times you'll die from terrorism over 75 years.

The line was:

P(Terrorist DOESN'T kill you in your lifetime) = P(Terrorist DOESN'T kill you in a specific year) ^ 75

It works like this:

What is the probability that I will survive this year? 0.999999
In order to survive for 2 years, I have to survive this year and survive next year. In statistics, P(X and Y) = P(X) * P(Y).

So the probability that you'll live for two years is 0.99999^2. And to survive 75 years is 0.999999^75

Comment: Re:We're so screwed. (Score 2) 237

You can figure out the chance you'll die from terrorism with some statistical math if you choose an expected lifetime length and your chance to die from terrorism each year. The math is fairly simple if you assume that your chance of dying from terrorism is roughly constant over your lifetime. It's not perfect but it can give you an idea of the magnitude of the risk involved.

Let P(terrorist kills you in your lifetime) be the probability of a terrorist killing you in your lifetime. Then,

P(terrorist kills you in your lifetime) = 1 - P(terrorist doesn't kill you in your lifetime)

If we assume that the likelihood of dying in a terrorist attack is fairly constant over your lifetime then:

P(terrorist doesn't kill you in your lifetime) = P(terrorist doesn't kill you in a specific year) ^ N

Where N is the number of years you expect to live. Lets overestimate the number of people that die from the kinds of terrorists you see on the news in the United States in a year. I do not see headlines about 100 people dying a year from actual terrorists but still I am going to overestimate and say it's a 1 in a million chance. So around 300 people dead a year in the U.S.A.

P(terrorist doesn't kill you in a specific year) = 1 - P(terrorist kills you in a specific year) = 1 - 0.000001 = 0.999999

So the formula looks like this:

P(terrorist kills you in your lifetime) = 1 - ((1 - P(terrorist kills you in a specific year)) ^ N)

For a lifetime of 75 years:

P(terrorist kills you in your lifetime) = 1 - (0.999999 ^ 75) = 7.4997 x 10^-5

Which is 2 (a.k.a a couple) orders of magnitude lower than your 1 x 10^-3.

Comment: Re:We're so screwed. (Score 2) 237

When that next truck bomb detonates at a sporting event or mall, or when that next muslim fan goes on an indiscriminate killing spree through a church, know in your heart that you have allowed that to happen.

I'll enjoy my freedom, thankyouverymuch, even if it does come with an 0.001% chance of dying by terrorist.

0.001%? That's insanely high. The real rate is a couple orders of magnitude lower. It just goes to show how completely terrible human beings are at estimate the risk of extremely rare events.

Comment: Re:1000 times (Score 2, Informative) 622

by TheNastyInThePasty (#49529773) Attached to: Cheap Gas Fuels Switch From Electric Cars To SUVs

You don't need to have a crystal ball. The price drop only happened because Saudi Arabia wanted to assert its dominance in the global market. With higher fuel prices, North American companies were investing in more expensive extraction methods that only become profitable when prices are high.

Saudi Arabia has been keeping its production down to drive up fuel prices and decided that enough was enough. They didn't even ramp up production to full capacity and it's been causing oil companies in North America to shutdown sites and lay off workers. Once Saudi Arabia decides that the oil companies get the picture, they will cut production again.

Then, even when prices rise, investors will think twice about risking their money to support oil extraction.

Comment: Re:well.. (Score 5, Informative) 760

I think the point is not for the police departments to get Teh Phat Lootz, but to equalize the pain of violating the rules.

You can't have one without the other. Unless you deny the entire government the money from the fines, the rich will become the only ones targeted by traffic cops. It's already bad enough that police departments prioritize money over safety. It could perhaps become bad enough that the cops ignore anyone without an extremely nice car because the revenue is not worth it.

Comment: Re:There is no legitimate reason to show it. (Score 1) 645

"Look at those evil Muslims killing innocent people to further their political goals! They are barbaric demonspawn! NUKE THE ENTIRE MIDDLE EAST!"

Of course, people with this view (I know a few) are completely unaware of the incredible irony...

Comment: Senator John McCain (Score 5, Insightful) 772

by TheNastyInThePasty (#48558351) Attached to: CIA Lied Over Brutal Interrogations

"I know from personal experience that the abuse of prisoners will produce more bad than good intelligence. I know that victims of torture will offer intentionally misleading information if they think their captors will believe it. I know they will say whatever they think their torturers want them to say if they believe it will stop their suffering. Most of all, I know the use of torture compromises that which most distinguishes us from our enemies, our belief that all people, even captured enemies, possess basic human rights, which are protected by international conventions the U.S. not only joined, but for the most part authored."

From a Republican even.

Comment: Re:America is a RINO (Score 4, Informative) 588

by TheNastyInThePasty (#48319501) Attached to: Marijuana Legalized In Oregon, Alaska, and Washington DC

In most states over 10% of the voters register as Independent. How do you gerrymander those to vote Republican?

It doesn't matter what they're registered as. What matters is what they vote for and most will vote predictably.

Democrats cluster in large cities. How do you evenly distribute their votes out into Republican districts on the other side of the state?

You don't have to distribute the democratic votes in the major cities. You assign as many as you can to majority Republican districts and then fit the rest into a district that is as close to 100% Democrat as you can.

Imagine a state with 800 people. Let's ignore the geographical distribution for simplicity. 59% (470) of the people vote purple, 41% (330) will vote orange, and you are in charge of drawing 4 districts such that the orange politicians remain in power. How will you do it?

3 districts with 110 orange people and 90 purple people (that's a 10% lead in elections which is plenty).
1 district with 200 purple people.

Congratulations! The orange people get 3 seats and the purple people get 1 despite the purple voters being a clear majority of the total. Here is a good illustration on wikipedia that also illustrates drawing the borders around geographically distributed voters.

Comment: America is a RINO (Score 5, Insightful) 588

by TheNastyInThePasty (#48317617) Attached to: Marijuana Legalized In Oregon, Alaska, and Washington DC

Yesterday's election was a message to Washington that America wants conservatives to represent them! Also, they want legalized pot, increased minimum wage, the right to have an abortion, insurance-provided contraception, and required paid time off at work!

Wait, what?

E = MC ** 2 +- 3db