My guess would be that they're performing the dating on once-living objects found in the same strata as the "tools". Since objects in the same strata are approximately the same age, carbon-dating those objects would provide an estimate to when the tools were first in existence.
The science may not be settled yet, but the burden of proof here still lies on the researcher.
Whenever a scientist gets experimental results that are far outside what was previously known and expected, the proper response is to either wait for independent verification (in this case, similar dating results from digs elsewhere in North America at the same depth) or subject the experimental procedure to intense scrutiny. Here, I would expect him to be able to justify
1) That the artifacts really came from the time he claims them to be from (probably easily doable via an independent dating test) 2) That the artifacts really came from the place he claims them to be from 3) That the artifacts are manmade.
Until each of these points is well supported, and barring the independent verification mentioned above, I'd hold out on adjusting the history textbooks.
In Florida in 2000, elections were run by the Secretary of State and co-chair to Bush's election campaign in that state. When the results in Florida were close, her presence tainted the election with the appearance of corruption no matter how partisanly or non-partisanly she behaved.
In Ohio in 2004, elections were run by a Secretary of State who was also heavily involved in Bush's reelection campaign. Fortunately the results in Ohio were not close enough for us to have the same problems, but if they were we would have had the same situation.
I'm not sure exactly who should be running the elections in each state, but it seems like it really shouldn't be officials who run on one party's ticket for statewide office and who may have higher ambitions within that party.
we have to do something about the way house districts are drawn.
Right now (when all that is at stake is a few seat in the house) we've seen court cases galore and the farce that happened in Texas. If we make it so that electoral votes depend on these gerrymanderings as well, things will just get more ridiculous.
Reading the methodology on the site, it seems like they figured out a probability for each state, then ran a simulation where each state independently has its given probability of going each way.
This seems utterly ridiculous! If swing state #1 goes one direction, then it is much more likely that swing state #2 is going in the same direction. Because of this, their model will have results centered artificially close to the expected value (swings in their methodology cancel each other much more than in the real election), and the probabilities of each candidate winning are closer to 50% than the site indicates.
These can be gotten from plugging x=1 in the taylor series for ln(1+x) or arctan(x), but require a fair bit of proof that the series actually converge to what you think they do. It's amazing that series that look so alike can get you pi on one hand and logs on the other!
Imagine that some state has n ballots which are mistakenly marked, but are done so randomly, that is each one is 50% likely to be Democratic instead of Republican or vice versa (I'm simplifying things by using a 2-party system here, but the same argument would hold with more parties).
Each time we miscount a Democratic Vote as a Republican, we mistakenly increase the Republican's margin by 2 votes. The reverse holds true when we miscount the other way. So the error in our margin can be thought of as the sum of n random numbers, each of which is equally likely to be 2 and -2. It is a known result from probability theory that such a "random walk" will be on (root mean square) average only about 2sqrt(n) steps away from 0. In other words, almost all the errors cancel!
Florida had a 537-vote margin in 2000. For that margin to have a 50% chance of occuring would have required roughly 72,000 random errors in counting (probably more, since usually less than half of a set of data is above its Root mean square). On the other hand, a 600 vote sufficiently systematic error would have been plenty.
This disparity only gets worse when we look at margins which would have been larger than 537 votes before error creeps in, but even at this level I'd be more worried about losing 600 votes by Fraud than having 72,000 of them miscounted.
Maine and Nebraska both allocate all their votes (except the 2 for the senator) by congressional district, and those districts are themselves currently decided by the politicians.
Considering how bad the redistricting battles are already, I really don't want to see electoral votes added as an extra jackpot for gerrymandering well.
(Un)fortunately (depending on your point of view), this stands very little chance of being approved, as it would dilute the voting power of people in those big rectangular states in the Presidential election.
I don't remember the exact numbers, but I remember someone working out what would have happened if Congress hadn't locked in the number of representatives, and as it turns out Gore would have won the 2000 election even w/o Florida.
Really it seems like the choice between IRV and Condorcet/Approval is a question of which sacrifices you make.
One one hand, there is the Condorcet Criterion: A candidate which beats every other candidate in a pairwise election (they call it an IDW) "should" win the election when it involves multiple candidates and people vote sincerely. On the other hand, there is Independence of Irrelevant Alternatives: Spoiler Candidates "should" have no effect on the election.
As it turns out, rigid adherance to the IIA criterion is unworkable; Arrow's theorem shows that it can't be applied in all cases without introducing other even more undesirable stuff (like a voter with absolute power). Once you realize this, there are two solutions.
1. Accept that IIA can't always be satisfied, and try to minimize the effect of spoiler candidates as much as possible. This will lead to occasional violations of the Condorcet criterion, but that is a sacrifice IIA supporters choose to make.
2. Castrate the IIA criterion by making it only apply to the so-called "Smith Set", itself a generalization of the Condorcet criterion. On the plus side, the Condorcet criterion will ALWAYS be satisfied.
The makers of the linked site prefer the latter, but they don't seem to realize that there was a choice made in the first place.
Take a look at the technical analysis, and at the criteria they choose:
The "Condorcet" criterion is: If there is an IDW and people vote sincerely, the IDW wins.
The "Generalized Condorcet" criterion is entirely based around the Smith Set, a generalization of the IDW.
The "Strategy Free" criteria really mean: If people vote honestly and there is an IDW, the IDW wins (almost exactly the same as the Condorcet criterion). Note the difference from the standard definition: Here they ONLY concern themselves with Condorcet winners!! An IRV supporter would say: No reasonable system satisfies the Strategy-Free definition, so we will try to violate it as little as possible. Here they just change the definition to make it match the Condorcet condition.
On the other hand, other criteria seem just about worthless except that they're not satisfied by IRV, so they make IRV seem more evil. Take the two "defensive strategy" for example, which state that for each pair (A,B) such that a majority prefers A to B, that majority can choose a strategy so that B is not elected no matter what the rest of the electorate does. The thing is, the "defensive strategies" for two different candidates can't be employed simultaneously, so all such a strategy guarantees is the ability for the population to choose between the two candidates they dislike the most. Again, note that this "defensive strategy" has been defined in terms of the same pairwise voting used to define the Condorcet criterion.
All their complicated "technical evaluation" seems to show is that Condorcet is preferable to IRV if you consider prefer 2) to 1). In other words, if you're willing to focus mostly on the Condorcet condition, the Condorcet method is the way to do so. What a surprise!
...in a three+ party election which other than a "dictatorship" (an election where there is a voter whose chosen candidate will win the election regardless of how everybody else votes)
This is a theoretical result proved by Gibbard and Satterthwaite which is somewhat similar to Arrow's theorem (which gives other conditions under which an election system must be a dictatorship).
Of course, this isn't to say that there won't be FEWER opportunities for tactical voting under an IRV or Condorcet system than under the current system. On the other hand, to assert that Condorcet is completely strategy-proof (or that IRV is bad just because it allows strategic voting) is foolish.
If seems like the Libertarian party has a chance of being on the ballot of every U.S. State EXCEPT New Hampshire. Of all the states for the party to be off the ballot in, I would have placed the "Live Free or Die" state right at the bottom of the list.
Is the NH signature requirement just that much harder than all the other states, or is there some other reason?
They can play classical music or maybe some old jazz tunes from 80 years back?
Although the music itself may be in the public domain, most performances of it would not be. Perhaps some conservatory students could start putting out free performances?
It's not that the puzzles are individually that difficult, but that there are so many of them. Several of the puzzles (a word search with some letters missing, for example, or some of the "matching pictures") can be brute forced in time. The catch is, there really isn't that much time...the winner averaged a puzzle solved every 6 minutes 49 seconds!
I could solve some of the easiest puzzles in that time, but the more difficult puzzles would take me (and most other solvers) MUCH longer than 7 minutes to figure out.
When the Astronomer is talking about the second scenario, he sees the critical description of that scenario as that "there are other Words in other languages to other cultures". According to Christian Theology (as quoted by him from the opening lines of the book of John), the Word of God existed before humanity did. In other words, the aliens we encounter will have already experienced God, that "need to overcome evil in the world".
He doesn't necessarily think that he's going to be converting them in this scenario. As I see it, he thinks that they will have already encountered some form of Christianity, perhaps in a form completely different from the one seen here on Earth, and that Christians may be able to learn from their encounters with (what he believes is) the same God.
In 2000 there were 8.75 million registered voters, most of whom were born between 1920 and 1980. Therefore on average >350 registered voters were born on a given day. If you removed everyone who shared a birthday with any felon in Florida, you'd be removing a heck of a lot more than 40,000 names.
...is who Slate chose as the author of the article. Looking at the "by the same author" at the end of the article, it seems like Slate decided to assign its 'Parenting' columnist instead of any sort of science writer. Is it surprising that she then decided to focus on the "nerdiness" and "looks like a jock" aspects rather than the projects themselves?
Slashdot Mirror
My guess would be that they're performing the dating on once-living objects found in the same strata as the "tools". Since objects in the same strata are approximately the same age, carbon-dating those objects would provide an estimate to when the tools were first in existence.
The science may not be settled yet, but the burden of proof here still lies on the researcher.
Whenever a scientist gets experimental results that are far outside what was previously known and expected, the proper response is to either wait for independent verification (in this case, similar dating results from digs elsewhere in North America at the same depth) or subject the experimental procedure to intense scrutiny. Here, I would expect him to be able to justify
1) That the artifacts really came from the time he claims them to be from (probably easily doable via an independent dating test)
2) That the artifacts really came from the place he claims them to be from
3) That the artifacts are manmade.
Until each of these points is well supported, and barring the independent verification mentioned above, I'd hold out on adjusting the history textbooks.
In Florida in 2000, elections were run by the Secretary of State and co-chair to Bush's election campaign in that state. When the results in Florida were close, her presence tainted the election with the appearance of corruption no matter how partisanly or non-partisanly she behaved. In Ohio in 2004, elections were run by a Secretary of State who was also heavily involved in Bush's reelection campaign. Fortunately the results in Ohio were not close enough for us to have the same problems, but if they were we would have had the same situation. I'm not sure exactly who should be running the elections in each state, but it seems like it really shouldn't be officials who run on one party's ticket for statewide office and who may have higher ambitions within that party.
we have to do something about the way house districts are drawn.
Right now (when all that is at stake is a few seat in the house) we've seen court cases galore and the farce that happened in Texas. If we make it so that electoral votes depend on these gerrymanderings as well, things will just get more ridiculous.
Note that Boston just won the World Series in Busch stadium.
Reading the methodology on the site, it seems like they figured out a probability for each state, then ran a simulation where each state independently has its given probability of going each way.
This seems utterly ridiculous! If swing state #1 goes one direction, then it is much more likely that swing state #2 is going in the same direction. Because of this, their model will have results centered artificially close to the expected value (swings in their methodology cancel each other much more than in the real election), and the probabilities of each candidate winning are closer to 50% than the site indicates.
1-1/2+1/3-1/4+1/5-...= ln(2)
s html/ for a couple examples)
1-1/3+1/5-1/7+1/9-...=Pi/4
These can be gotten from plugging x=1 in the taylor series for ln(1+x) or arctan(x), but require a fair bit of proof that the series actually converge to what you think they do. It's amazing that series that look so alike can get you pi on one hand and logs on the other!
On the other hand, for sheer drop-your-jaw-in-awe-ness it's hard to beat some of the formulas from Ramanujan's notebooks (see example #5 at http://www.cut-the-knot.org/do_you_know/fraction.
Imagine that some state has n ballots which are mistakenly marked, but are done so randomly, that is each one is 50% likely to be Democratic instead of Republican or vice versa (I'm simplifying things by using a 2-party system here, but the same argument would hold with more parties).
Each time we miscount a Democratic Vote as a Republican, we mistakenly increase the Republican's margin by 2 votes. The reverse holds true when we miscount the other way. So the error in our margin can be thought of as the sum of n random numbers, each of which is equally likely to be 2 and -2. It is a known result from probability theory that such a "random walk" will be on (root mean square) average only about 2sqrt(n) steps away from 0. In other words, almost all the errors cancel!
Florida had a 537-vote margin in 2000. For that margin to have a 50% chance of occuring would have required roughly 72,000 random errors in counting (probably more, since usually less than half of a set of data is above its Root mean square). On the other hand, a 600 vote sufficiently systematic error would have been plenty.
This disparity only gets worse when we look at margins which would have been larger than 537 votes before error creeps in, but even at this level I'd be more worried about losing 600 votes by Fraud than having 72,000 of them miscounted.
Maine and Nebraska both allocate all their votes (except the 2 for the senator) by congressional district, and those districts are themselves currently decided by the politicians.
Considering how bad the redistricting battles are already, I really don't want to see electoral votes added as an extra jackpot for gerrymandering well.
(Un)fortunately (depending on your point of view), this stands very little chance of being approved, as it would dilute the voting power of people in those big rectangular states in the Presidential election.
I don't remember the exact numbers, but I remember someone working out what would have happened if Congress hadn't locked in the number of representatives, and as it turns out Gore would have won the 2000 election even w/o Florida.
Really it seems like the choice between IRV and Condorcet/Approval is a question of which sacrifices you make.
One one hand, there is the Condorcet Criterion: A candidate which beats every other candidate in a pairwise election (they call it an IDW) "should" win the election when it involves multiple candidates and people vote sincerely. On the other hand, there is Independence of Irrelevant Alternatives: Spoiler Candidates "should" have no effect on the election.
As it turns out, rigid adherance to the IIA criterion is unworkable; Arrow's theorem shows that it can't be applied in all cases without introducing other even more undesirable stuff (like a voter with absolute power). Once you realize this, there are two solutions.
1. Accept that IIA can't always be satisfied, and try to minimize the effect of spoiler candidates as much as possible. This will lead to occasional violations of the Condorcet criterion, but that is a sacrifice IIA supporters choose to make.
2. Castrate the IIA criterion by making it only apply to the so-called "Smith Set", itself a generalization of the Condorcet criterion. On the plus side, the Condorcet criterion will ALWAYS be satisfied.
The makers of the linked site prefer the latter, but they don't seem to realize that there was a choice made in the first place.
Take a look at the technical analysis, and at the criteria they choose:
The "Condorcet" criterion is: If there is an IDW and people vote sincerely, the IDW wins.
The "Generalized Condorcet" criterion is entirely based around the Smith Set, a generalization of the IDW.
The "Strategy Free" criteria really mean: If people vote honestly and there is an IDW, the IDW wins (almost exactly the same as the Condorcet criterion). Note the difference from the standard definition: Here they ONLY concern themselves with Condorcet winners!! An IRV supporter would say: No reasonable system satisfies the Strategy-Free definition, so we will try to violate it as little as possible. Here they just change the definition to make it match the Condorcet condition.
On the other hand, other criteria seem just about worthless except that they're not satisfied by IRV, so they make IRV seem more evil. Take the two "defensive strategy" for example, which state that for each pair (A,B) such that a majority prefers A to B, that majority can choose a strategy so that B is not elected no matter what the rest of the electorate does. The thing is, the "defensive strategies" for two different candidates can't be employed simultaneously, so all such a strategy guarantees is the ability for the population to choose between the two candidates they dislike the most. Again, note that this "defensive strategy" has been defined in terms of the same pairwise voting used to define the Condorcet criterion.
All their complicated "technical evaluation" seems to show is that Condorcet is preferable to IRV if you consider prefer 2) to 1). In other words, if you're willing to focus mostly on the Condorcet condition, the Condorcet method is the way to do so. What a surprise!
...in a three+ party election which other than a "dictatorship" (an election where there is a voter whose chosen candidate will win the election regardless of how everybody else votes)
This is a theoretical result proved by Gibbard and Satterthwaite which is somewhat similar to Arrow's theorem (which gives other conditions under which an election system must be a dictatorship).
Of course, this isn't to say that there won't be FEWER opportunities for tactical voting under an IRV or Condorcet system than under the current system. On the other hand, to assert that Condorcet is completely strategy-proof (or that IRV is bad just because it allows strategic voting) is foolish.
If seems like the Libertarian party has a chance of being on the ballot of every U.S. State EXCEPT New Hampshire. Of all the states for the party to be off the ballot in, I would have placed the "Live Free or Die" state right at the bottom of the list.
Is the NH signature requirement just that much harder than all the other states, or is there some other reason?
So outsourcing is a labor market version of trickle-down economics?
They can play classical music or maybe some old jazz tunes from 80 years back?
Although the music itself may be in the public domain, most performances of it would not be. Perhaps some conservatory students could start putting out free performances?
January through November 2003 U.S. Sales:
PS2: 4.41 million
Gamecube: 2.12 million
Xbox: 2.03 million
Seems like a significant player to me
(This is according to http://www.vgpro.com/news/4121 ; I'd welcome a second source)
The 2004 data really isn't that good to go by yet, as the Democrats just went through a long primary season that the Republicans didn't.
Dropping that year, the trend doesn't really exist.
It's not that the puzzles are individually that difficult, but that there are so many of them. Several of the puzzles (a word search with some letters missing, for example, or some of the "matching pictures") can be brute forced in time. The catch is, there really isn't that much time...the winner averaged a puzzle solved every 6 minutes 49 seconds!
I could solve some of the easiest puzzles in that time, but the more difficult puzzles would take me (and most other solvers) MUCH longer than 7 minutes to figure out.
"The star's original mass is not known, so there's a roughly equal chance that the remaining central object is a neutron star or a black hole"
Just because you don't know whether or not an event occurs doesn't mean that it stands a "roughly equal" chance of occurring and not occurring
since 2^(odd number)+1 is always a multiple of 3
He's also found the largest known perfect number, 2^(24,036,583-1)*((2^24,036,583)-1)
ask Harvey Haddix, who pitched a perfect 12 innings, only to lose in #13.
...or at least it's not how I interpret it.
When the Astronomer is talking about the second scenario, he sees the critical description of that scenario as that "there are other Words in other languages to other cultures". According to Christian Theology (as quoted by him from the opening lines of the book of John), the Word of God existed before humanity did. In other words, the aliens we encounter will have already experienced God, that "need to overcome evil in the world".
He doesn't necessarily think that he's going to be converting them in this scenario. As I see it, he thinks that they will have already encountered some form of Christianity, perhaps in a form completely different from the one seen here on Earth, and that Christians may be able to learn from their encounters with (what he believes is) the same God.
In 2000 there were 8.75 million registered voters, most of whom were born between 1920 and 1980. Therefore on average >350 registered voters were born on a given day. If you removed everyone who shared a birthday with any felon in Florida, you'd be removing a heck of a lot more than 40,000 names.
...is who Slate chose as the author of the article. Looking at the "by the same author" at the end of the article, it seems like Slate decided to assign its 'Parenting' columnist instead of any sort of science writer. Is it surprising that she then decided to focus on the "nerdiness" and "looks like a jock" aspects rather than the projects themselves?