Actually, I have heard that prosecutors hate CSI type shows because jurors want some technician to come in and say, "the pesticide oh his wheels indicate that he was at the farm at the time of the crime." Luck breaks like that only happen on... well, CSI.
Yeah, you also have to think about how much of your water comes from drinking water and how much comes from food. Most food is local (especially things like bread), but even produce is often not from that far off (bananas excluded, obviously).
Uh... the USG claims you went to Afghanistan to train with terrorists for a month, you just went to London like your last ticket purchased in the US says... then you use the hair.
If the hair says London, the government knows you either had a massive water truck with you or that you are telling the truth. OTHO, if the hair says Afghanistan then they have circumstantial evidence and might have caught you in a lie. In the end, it's better screening out than in or catching people in lies.
It was energy generation, not distribution, that was deregulated. That can easily be deregulated. It just turns out that the companies start to use monopoly power and jack the prices up. Remember in California they started to do "routine maintenance" on every power generating site to keep supply down and force prices up. It was a travesty.
When government fucks with free markets, the customer loses, always.
Well, except in the case of energy regulation, every state that has deregulated has instantly had massive price spikes (or are these good for the consumer?)... and insurance where the companies kick you out as soon as you file claims unless regulated.
The US government usually asks the market players to regulate themselves and hopes that works (think of movie ratings). It is only after the players show they have no interest in a fair market that it gets regulated.
If the police used your logic, I could beat the crap out of people (but not kill them) with carte blanche until all the more serious criminals were behind bars. Then I could downgrade to something like property damage until all those criminals were behind bars...
Law enforcement... enforces laws. They really shouldn't be responsible for picking which ones to enforce.
This was an open question until it was decided in the courts. Now, linking is also not allowed either.
I'm not really sure I see why so many slashdotters are so worked up about this. This is US property that is having its IP rights violated--it's a valid case, it is settled law. While you might disagree that it is possible to take down every site with ripped content, there are a million other laws that are that way too and still we prosecute offenders. Murder isn't going to go away because we prosecute it, but we still do it.
I'm confused as to what you might loose? What is the offense against you?
Uh, "how stupid the US gov can be sometimes..." I'd not us that for this instance where hindsight is 20-20 whole figuring out on of the first passive resonators is really hard.
Now, the CIA figuring out that Russia was exiting Afghanistan 9 months after Russia held a press conference saying they were leaving Afghanistan, that's stupid. And, that's form a book written by a previous CIA director to trumpet their successes.
That would be fine, but then having to learn a new one every 12 weeks because of a password expiration cycle--that's when it gets impossible. You are always recalling fragments of the old password...
Part of the reason it's confusing is because if you have a boy on a Tuesday and then decide to ask people after the next is born, "I have two children, one is a boy born on a Tuesday" then the probability of two boys is 50%. But, if you would say that if either of your children were a boy born on a Tuesday, then you are back to 13/27.
The reason the person asked the question they way they did matters. I really think it is all about communicative ambiguity.
You can actually set up the Monty Hall problem to be that switching is always bad (per your example) and always good (opposite example), and then using linear combinations, any probability between.
Sometimes, but other times no. Examples: a connection that never leaves my subnet, if someone can launch a MITM on my network, I'm so much more screwed than I thought I should just give up in the first place. OR, a connection that I first initiate on a shared network, store the exception and THEN make remotely. OR, a connection that I verify the signature for over the phone.
Sorry, WRT one- versus two-tailed: you are assuming that you know the difference is in one direction. If you knew that before the experiment, why didn't you tell everyone else and save us the millions of expense on this experiment? If you didn't, then you have to use the two-tailed.
Again, using confidence intervals the way you want to use them I could say that the probability that a penguin named George will win certain baseball games is 95% using the estimator I constructed above.
Alternately, I could say that the probability of a friend eating 2 to 2.2 cups of cereal in the morning is 95% when in fact they have already eaten breakfast and they had 2.4 cups--so the probability is actually zero.
I don't understand. You want to make videos of lawnmowers killing kittens and puppies? Is that seriously your concern?
Keep in mind, congress acted in part because these animals were killed only to make the video. People were not killing puppies and kittens and recording it incidentally--they were killing kittens and puppies to make a video. The killing was tied to the video industry so unless you think animal cruelty is a speech right then I don't understand the concern about Stevens.
If SCOTUS rules that a law is A but not B there is the question: not B because it was unconstitutional or because the law just doesn't say B. They answer this in their opinion.
BTW, IANAL, I've just noticed this type of text in decisions: "congress could make a law that does this, it would be constitutional, but this law does not do this." Or they might throw a law because it is too general and then outline what might (or might not) say what reasonable limits might be. Sometimes SCOTUS is super conservative in terms of what they will rule on. Generally, they have tried to steer clear of the definition of religion in the establishment clause. Even in cases where it appeared that they would have to take it on directly, they managed to not talk about it.
I didn't condition on the null, checkout the OP, that is why Parlyne and I agree that X need not equal Y.
What you wrote is exactly what you can't say. Remember: the interval is random, not the true ratio. Parlyne's statement is about the null hypothesis (it will be this extreme under the null only 5% of the time), I tried to say that the confidence interval will contain the true value 95% of the time (though we don't know the probability that the true value is in the confidence interval).
To see this consider a few cases: (1) a baseball predictor says 19 times out of 20, "exactly one of the two teams playing will win." and the other time, "a penguin named George will will the game." Then, using this method, the actual result is in the predicted set 95% of the time (just like a 95% confidence interval), but you can't say, when it predicts that exactly one of the two teams will win that the probability of one of the two teams winning is 95%. Why? Because the probability of exactly one of the two teams winning is 100%. Same goes for the penguin winning, except George never wins, so the true value is 0%.
As an alternative--imagine I am predicting the final sale value of a friends home and I come up with a model and get a 95% confidence interval. Then I call him and say, there is a 95% chance you home will sell for between x and y dollars. But he tells me that he sold the home and it is not between x and y dollars. So the probability of the event of his home selling for between x and y dollars was actually zero.
Now, consider the statement, "Given my methods, I would predict the value of your home sale is between x and y 95% of the time." Can't argue with that!
Also, you are confused about the interpretation of one versus two tailed tests. They probably used (and should have used) a two-tailed test. But then you "fixed" it to a one-tailed test, which is not valid.
AND often when they make policy, they will layout how the legislative body can change the law if they don't like the outcome. Sometimes they decide on a constitutional ground, but even then they might say something like, "if the legislature had done this... it would have been acceptable."
Slashdot Mirror
I think it would be more, which terrorists were in which countries when. But also useful for someone reentering the US with an unknown travel history.
Actually, I have heard that prosecutors hate CSI type shows because jurors want some technician to come in and say, "the pesticide oh his wheels indicate that he was at the farm at the time of the crime." Luck breaks like that only happen on... well, CSI.
Most people get most of their water from food and food is often approximately local, especially things like the water in soup and bread.
Yeah, you also have to think about how much of your water comes from drinking water and how much comes from food. Most food is local (especially things like bread), but even produce is often not from that far off (bananas excluded, obviously).
Uh... the USG claims you went to Afghanistan to train with terrorists for a month, you just went to London like your last ticket purchased in the US says... then you use the hair.
If the hair says London, the government knows you either had a massive water truck with you or that you are telling the truth. OTHO, if the hair says Afghanistan then they have circumstantial evidence and might have caught you in a lie. In the end, it's better screening out than in or catching people in lies.
It was energy generation, not distribution, that was deregulated. That can easily be deregulated. It just turns out that the companies start to use monopoly power and jack the prices up. Remember in California they started to do "routine maintenance" on every power generating site to keep supply down and force prices up. It was a travesty.
Except every economist who advocates for it?
When government fucks with free markets, the customer loses, always.
Well, except in the case of energy regulation, every state that has deregulated has instantly had massive price spikes (or are these good for the consumer?)... and insurance where the companies kick you out as soon as you file claims unless regulated.
The US government usually asks the market players to regulate themselves and hopes that works (think of movie ratings). It is only after the players show they have no interest in a fair market that it gets regulated.
If the police used your logic, I could beat the crap out of people (but not kill them) with carte blanche until all the more serious criminals were behind bars. Then I could downgrade to something like property damage until all those criminals were behind bars...
Law enforcement... enforces laws. They really shouldn't be responsible for picking which ones to enforce.
ICE raids places all the time, try using google for like five seconds.
This was an open question until it was decided in the courts. Now, linking is also not allowed either.
I'm not really sure I see why so many slashdotters are so worked up about this. This is US property that is having its IP rights violated--it's a valid case, it is settled law. While you might disagree that it is possible to take down every site with ripped content, there are a million other laws that are that way too and still we prosecute offenders. Murder isn't going to go away because we prosecute it, but we still do it.
I'm confused as to what you might loose? What is the offense against you?
Uh, "how stupid the US gov can be sometimes..." I'd not us that for this instance where hindsight is 20-20 whole figuring out on of the first passive resonators is really hard.
Now, the CIA figuring out that Russia was exiting Afghanistan 9 months after Russia held a press conference saying they were leaving Afghanistan, that's stupid. And, that's form a book written by a previous CIA director to trumpet their successes.
That would be fine, but then having to learn a new one every 12 weeks because of a password expiration cycle--that's when it gets impossible. You are always recalling fragments of the old password...
Part of the reason it's confusing is because if you have a boy on a Tuesday and then decide to ask people after the next is born, "I have two children, one is a boy born on a Tuesday" then the probability of two boys is 50%. But, if you would say that if either of your children were a boy born on a Tuesday, then you are back to 13/27.
The reason the person asked the question they way they did matters. I really think it is all about communicative ambiguity.
You can actually set up the Monty Hall problem to be that switching is always bad (per your example) and always good (opposite example), and then using linear combinations, any probability between.
You couldn't patent pi if you came out and said you were patenting pi. You would do, "A method of finding a circle's circumference using its radius."
Sometimes, but other times no. Examples: a connection that never leaves my subnet, if someone can launch a MITM on my network, I'm so much more screwed than I thought I should just give up in the first place. OR, a connection that I first initiate on a shared network, store the exception and THEN make remotely. OR, a connection that I verify the signature for over the phone.
Sorry, WRT one- versus two-tailed: you are assuming that you know the difference is in one direction. If you knew that before the experiment, why didn't you tell everyone else and save us the millions of expense on this experiment? If you didn't, then you have to use the two-tailed.
You can say
Pr(interval contains true value) = 0.95
Not
Pr(true value in interval) = 0.95
Again, using confidence intervals the way you want to use them I could say that the probability that a penguin named George will win certain baseball games is 95% using the estimator I constructed above.
Alternately, I could say that the probability of a friend eating 2 to 2.2 cups of cereal in the morning is 95% when in fact they have already eaten breakfast and they had 2.4 cups--so the probability is actually zero.
I don't understand. You want to make videos of lawnmowers killing kittens and puppies? Is that seriously your concern?
Keep in mind, congress acted in part because these animals were killed only to make the video. People were not killing puppies and kittens and recording it incidentally--they were killing kittens and puppies to make a video. The killing was tied to the video industry so unless you think animal cruelty is a speech right then I don't understand the concern about Stevens.
If SCOTUS rules that a law is A but not B there is the question: not B because it was unconstitutional or because the law just doesn't say B. They answer this in their opinion.
BTW, IANAL, I've just noticed this type of text in decisions: "congress could make a law that does this, it would be constitutional, but this law does not do this." Or they might throw a law because it is too general and then outline what might (or might not) say what reasonable limits might be. Sometimes SCOTUS is super conservative in terms of what they will rule on. Generally, they have tried to steer clear of the definition of religion in the establishment clause. Even in cases where it appeared that they would have to take it on directly, they managed to not talk about it.
I didn't condition on the null, checkout the OP, that is why Parlyne and I agree that X need not equal Y.
What you wrote is exactly what you can't say. Remember: the interval is random, not the true ratio. Parlyne's statement is about the null hypothesis (it will be this extreme under the null only 5% of the time), I tried to say that the confidence interval will contain the true value 95% of the time (though we don't know the probability that the true value is in the confidence interval).
To see this consider a few cases: (1) a baseball predictor says 19 times out of 20, "exactly one of the two teams playing will win." and the other time, "a penguin named George will will the game." Then, using this method, the actual result is in the predicted set 95% of the time (just like a 95% confidence interval), but you can't say, when it predicts that exactly one of the two teams will win that the probability of one of the two teams winning is 95%. Why? Because the probability of exactly one of the two teams winning is 100%. Same goes for the penguin winning, except George never wins, so the true value is 0%.
As an alternative--imagine I am predicting the final sale value of a friends home and I come up with a model and get a 95% confidence interval. Then I call him and say, there is a 95% chance you home will sell for between x and y dollars. But he tells me that he sold the home and it is not between x and y dollars. So the probability of the event of his home selling for between x and y dollars was actually zero.
Now, consider the statement, "Given my methods, I would predict the value of your home sale is between x and y 95% of the time." Can't argue with that!
Also, you are confused about the interpretation of one versus two tailed tests. They probably used (and should have used) a two-tailed test. But then you "fixed" it to a one-tailed test, which is not valid.
Sure, go to Cornell's SCOTUS decisions page and search for "congress can". search this page for a single example.
I think your friends were talking about lower courts.
Uh, if the calibration is wrong, then the p-value is irrelevant because it is based on poor assumptions.
Also, if they have a prior, then I wonder where it came from.
AND often when they make policy, they will layout how the legislative body can change the law if they don't like the outcome. Sometimes they decide on a constitutional ground, but even then they might say something like, "if the legislature had done this... it would have been acceptable."