But I never have understood the sanity behind jumping out of a perfectly good plane.:(
Women bear the economic price of childbirth. As a consequence, they tend to be conservative and choosy in picking mates and men have to compete for access. In order to succeed, men have evolved to take risks - we see this when comparing the bell curves of women versus men: women tend to have lower standard deviations than men. More women are of average height for women, men tend to have more varied heights. More men are born than women because over the course of their maturity, more men will die from taking risks.
Woman tend to choose men who are successful at taking risks, because those men show capability over other men.
Men tend to get elated by risky activities. It's an emotional cue for a reproduction strategy.
(I'm sorry - were you asking a rhetorical question?)
This tells me that no matter how promising your hardware design and software, I am going to be spending a lot of money before I have anything close to a commercially viable product.
Apropos of nothing, I'm guessing that you have never started a business. Just a guess, mind you...
A cynic would point out that the cost to develop these devices is very high, and companies must recoup their losses.
A mathematician would point out that, from a game-theory point of view, having one group of people come up with safety requirements without any burden of cost for the implementation of those requirements leads to stifling over-regulation.
Specifically, this leads to "safety at any cost", when in reality the cost of safety should be compared to other costs.
There is a cost of spinning reserve and grid stability maintenance. Why shouldn't those who need it or negatively impact it pay for it?
The answer has to do with the Prisoner's Dilemma. What is best for the individual may not be the best for society, and what appears to be a sub-optimal choice for the individual often turns out to be better for the individual in the long run if everyone makes the same choice.
For a concrete example, consider the Polio vaccine. There is a small chance of getting polio from the vaccine, so from the individual's point of view it makes no sense to get your kids vaccinated. If everyone else is vaccinated, you can forego the vaccination and reduce the risk to your child even further... except that if everyone makes that selfish choice, no one is protected.
In this particular case, there is now an incentive for people to purchase batteries and disconnect from the grid (or perhaps purchase batteries and not back-sell electricity). Instead of having the public service managing grid load using the best possible method with economies of scale, the problem is distributed among many small players. It's a narrow-minded view of an externality.
People have an incentive to spend revenue purchasing the batteries, and spend time dealing with installation/setup/maintenance/disposal rather than a few people doing this in bulk. The aggregate loss of productivity to society is much greater. Environmentally speaking, there will be lots of batteries in landfills in 10 years or so.
There is an incentive for less electricity put into the grid to be made available for others. The market has less "liquidity" now, generally considered (economically) to be a bad thing. The utility will need to generate more energy to compensate, which will result in more expenses and more environmental damage, and these changes may cost the utility more than the revenue from the monthly fees.
People without solar panels had the burden of an extra expense. Rather than viewing the expense as unwarranted, you could view it as an incentive to purchase solar panels, which is generally better for society.
Public services are corporations and as such want to make as much money as possible for their stock-holders. They only view this in terms of revenue - they could also consider the expense of externalities, and make decisions to maximize stockholder value rather than narrowly focusing on profit.
On the downloads page you can download the models in various forms - point cloud, mesh, and so on. Different formats, depending on the method used to get the model data (cat scan, laser scan, photographic, &c).
They mention in the about page that it would take 247 years of work 24/7 to capture the entire collection.
We could hire 247 people and get the entire collection online in 3 years (8 hour shifts). At $50,000 per person, that's about $13 million per year(*). Compare to the cost of the Obamacare website currently estimated at around $100 million and it has to be redone.
They've obviously shown "proof of concept" for getting the job done. Can we somehow just give them the money to complete it? Maybe a petition on "We the People"?
(*) Back-of-the-envelope calculation doesn't include cost of scanning equipment or materials, but note that there are a *lot* of museums in this country. We could invest in the infrastructure once and keep 300 people employed for decades putting this great stuff online.
The system is really cool - you can rotate and view a mastodon skeleton (or supernova model) from any angle, it works in the browser, and it's really fast (on my machine, at least).
You can change the material color/texture, change the lighting properties & angle, take a slice of the 3-d image(!), put down a measuring tape, add annotations, then generate a link to that image.
I couldn't find a way to download the model, but you can file->save the web page, and it appears to store all the javascript for the editor on your local system.
The system is all kinds of awesome. Watching them add to their inventory should be very interesting.
We are also testing the boundaries of the physical universe in a completely new realm.
The number of states in a quantum-entangled set of particles goes exponentially with the number of particles. For 10 particles (entangled) it takes 2^10 states for the universe to represent the possible outcomes. For 1,000 entangled particles, the number of states is 2^1000.
The number of particles in the entire universe is only about 2^80.
Managing 1,000 entangled particles would require the universe to keep track of a staggering amount of information. Does the underlying machinery have information-space this big? No one knows.
For the first time, we can measure the boundaries of the physical mechanism that underlies the universe in a completely different realm: information capacity. This is analogous to a program probing the limits of RAM memory by seeing how much it can allocate.
Here's hoping that we don't find a buffer overflow.
Sure, which means getting rid of all income tax, capital gains, any form of 'service' as in no government, so buy a lot of guns, because the robber barons that were politely robbing you blind privately will be shooting you in your face for all those gold coins you have packed under your matress.
Really dude?
Is this the sort of argument that rates "+4 interesting" on slashdot nowadays?
Stop story-telling! We're nerds - we're better than that.
The value of bitcoins is up at the news of a hearing that is probably going to make it illegal to do 95% of what is done with BTC? That makes sense.
1) It doesn't matter what the US government decides, if people want to use it it will be used. Reference: drugs, prostitution, illegal immigration, abortion, illegal guns, and other world countries.
2) People don't act because they have no alternatives. This is an alternative, but not many people know about it. High-profile senate hearings publicize BitCoin, so that more people will realize how useful it is, and this spurs demand.
3) There are no valid argument against BitCoin. There are economic "story telling" arguments, and predictive "doom and gloom by story telling" arguments, and false equivalences with closely related things (fraudulent exchanges, Silk Road, &c), and outright lies ("it's a Ponzi scheme!!!"), but no actually valid arguments.)
4) All the standard aphorisms apply: buggy-whip manufacturers, the invisible hand, liquidity, privacy and freedom.
BitCoin will become a game-changer (oh, that phrase!) simply because nothing can stop it.
It's narrow mindedness in a government institutions, a common malady.
The DHS sees a need to stop some activity or other, and this makes perfect sense in context. One only has to look at the Syria and Egypt for examples of how this is used in practice - if the US ever descends into armed revolt, the switch will disrupt revolutionary communications and make it easier for the government to regain control. The military has its own, separate channels of communication.
Like all government institutions, it's narrow minded. They only think of themselves and their (DHS's) own needs, without regard to anything outside of their remit. Turning off the internet will have massive consequences to the economy locally and worldwide, but that is considered unimportant. There is no consideration of the action "in context" or the ramifications thereof, it simply achieves the goal.
They need this functionality, and other considerations be damned.
A recent post asked about "fear" and how it's used by the government to control people, but fear works in reverse as well. The government's actions are unsustainable (at the very least, economically) and it's control over the people is rapidly coming to an end. They're terrified of the end-game, and are putting pre-emptive measures into place ahead of time.
As someone previously said, elected officials are cutting everything except checks to supporters. If they stop handouts to supporters their control will fail. If cutting services causes massive unrest, they will fail.
They are between a rock and a hard place and getting squeezed harder every day. It will be interesting to see the end game when it happens.
(Look to the first month-or-two of 2014 as a possible start-date. That's when budget/debt limit talks start anew, and it's when everyone's health insurance costs will double. That's only a possibility, but next summer is looking really good for massive protests.)
What we need is a way to measure the money put into legislation like this. Perhaps a "kickstarter" for political action.
NetFlix probably hasn't put much into the political process. The ISPs have made more campaign donations, so this legislation is pretty-well doomed from the start.
We need a website where people can pledge donations to candidates who vote for or against specific legislation, sort of like Kickstarter for laws. Unlike kickstarter, people (corporations, too!) could pledge a specific amount either "for" or "against" a specific law. This would give lawmakers an easier and much more efficient way of judging which laws are most valuable to society.
This would also make political advocacy more efficient. Rather than donating to PACs or lobbyists, the public could send money directly to the pockets of the legislators involved. In economics-speech, It adds "liquidity" to this particular market - eliminating middlemen (who are only rent-seekers) and passing the savings along to the end-user.
The "invisible hand" of economics is often touted as the most modern and efficient way to solve a complex problem, yet we labor under an antiquated 250-year-old political system which is slow and inefficient.
Let's upgrade to more modern methods. We need a kickstarter system for laws.
It clearly can't be based on floating-point binary representations. These are not scale-free. What I mean is, under one scale, you might have an error of 0.5 and you might say that needs 2 bits to store. Under another scale the same error would be 0.3333333 which takes infinite bits to store, under the same storage scheme. Even if you solve that, you'd get that a small but complex error (e.g. 1/pi^100) is worse than a huge but simple one (e.g. 2^64). That's just ridiculous.
So we have to use a metric where larger errors give a larger metric. A monotonic function, at least. So let's choose log2(error), right?
Ah, damn! If only you weren't anonymous, I'd love to continue this conversation offline!
That level of insight! I didn't expect anyone to be familiar enough with the foundations of statistics to make this observation.
You were correct up to the point where you chose log2(error) as a metric. There's a function which can be deduced from first principles, and which solves all the above-mentioned problems. Can you find it?
Back-edit the statistical methods and use the new error function instead of least squares. This leads to a solution to the fixed-width floating point problem, the regression problems mentioned above (including mixtures of gaussians mentioned in a side-post), and an algorithmic solution to the front-page article issue. (As a side-note, it's the first step towards hard-AI.)
I'm writing up my results right now - we'll see if my solution holds up under scrutiny.
The problem I have with least squares is that I don't like the definition of the "error". If you have two things that are correlated, one isn't necessariy a function of the other that includes some variability. If you flip the X and Y axes over - plot height against weight, rather than weight against height - then the least squares regression gives a different line. If the two errors are both minimised, but different, then neither of them is the "real" error.
Wow - brilliant insight! Thanks for that - things like this are why I come to Slashdot.
It seems like you found a lot of problems with what is taught in a stat 101 class. This is good; there are many. However, there are also solutions to these problems which you would find if you took a higher level course.
Actually, there is a really good reason to use least-squares regression. A model that minimizes squared error is guaranteed to minimize the variance of error, obviously.
This is the wrong place for an argument (you want room 12-A) and I don't want to get into a contest, but for illustration here is the problem with this explanation.
A rule learned from experience should minimize the error, not the variance of error.
It's a valid conclusion from the mathematics, but based on a faulty assumption.
Do outliers skew the results? If the outliers are biased, then that may tell us something about the underlying population. If they aren't biased, then their effects cancel.
There's no algorithm that will identify the outliers in this example.
But random data would generate statistically insignificant correlation coefficients. Also, the 95% confidence intervals used to predict values are wider for random data.
What value of correlation coefficient distinguishes pattern data from random data in this image?
Okay, here's the real problem with scientific studies.
All science is data compression, and all studies are are intended to compress data so that we can make future predictions. If you want to predict the trajectory of a cannonball, you don't need an almanac cross referencing cannonball weights, powder loads, and cannon angles - you can calculate the arc to any desired accuracy with a set of equations that fit on half a page. The half-page compresses the record of all prior experience with cannonball arcs, and allows us to predict future arcs.
Soft science studies typically make a set of observations which relate two measurable aspects. When plotted, the data points suggest a line or curve, and we accept the linear-regression (line or polynomial) as the best approximation for the data. The theory being that the underlying mechanism is the regression, and unrelated noise in the environment or measurement system causes random deviations of observation.
This is the wrong method. Regression is based on minimizing squared error, which was chosen by Laplace for no other reason that it is easy to calculate. There's lots of "rationalization" explanations of why it works and why it's "just the best possible thing to do", but there's no fundamental logic that can be used to deduce least squares from from fundamental assumptions.
Least squares introduces several problems:
1) Outliers will skew the values, and there is no computable way to detect or deal with outliers (source).
2) There is no computable way to determine whether the data represent a line or a curve - it's done by "eye" and justified with statistical tests.
3) The resultant function frequently looks "off" to the human eye, humans can frequently draw better matching curves; meaning: curves which better predict future data points.
4) There is no way to measure the predictive value of the results. Linear regression will always return the best line to fit the data, even when the data is random.
The right way is to show how much the observation data is compressed. If the regression function plus data (represented as offsets from the function) take fewer bits than the data alone, then you can say that the conclusions are valid. Further, you can tell how relevant the conclusions are, and rank and sort different conclusions (linear, curved) by their compression factor and choose the best one.
Scientific studies should have a threshold of "compresses data by N bits", rather than "1-in-20 of all studies are due to random chance".
First of all, it's Professor David Nutt, not "Nut"
Second of all, it's the same Professor who was a British government advisor, who was sacked for "criticising politicians for distorting research evidence and claiming alcohol and tobacco were more harmful than some illegal drugs, including LSD, ecstasy and cannabis."
Seems like a scientist with integrity. Perhaps this is less the risible ramblings of a madnam, and more he's at the "...then they laugh at you" part of fighting the good fight.
(Unless, of course, you think LSD and cannabis are more damaging than alcohol and tobacco, in which case feel free to poke fun.)
The Earth-Moon system orbits around a point which is not at the center of the Earth, sort of like a barbell with small and large ends will balance on a point which is not in the center of the big end. The point of rotation is still inside the Earth, just not at the center of the Earth.
This is why we have two tides [roughly] each day - the side near the moon gets attracted to the moon, while the other side gets swung around on the outside and experiences centrifugal force, pulling it away from the center.
The moon should experience this same effect. I would expect tidal forces to draw mass away from the center, resulting in more mass on the far side.
Slashdot Mirror
Nobody expected that!
But I never have understood the sanity behind jumping out of a perfectly good plane. :(
Women bear the economic price of childbirth. As a consequence, they tend to be conservative and choosy in picking mates and men have to compete for access. In order to succeed, men have evolved to take risks - we see this when comparing the bell curves of women versus men: women tend to have lower standard deviations than men. More women are of average height for women, men tend to have more varied heights. More men are born than women because over the course of their maturity, more men will die from taking risks.
Woman tend to choose men who are successful at taking risks, because those men show capability over other men.
Men tend to get elated by risky activities. It's an emotional cue for a reproduction strategy.
(I'm sorry - were you asking a rhetorical question?)
This tells me that no matter how promising your hardware design and software, I am going to be spending a lot of money before I have anything close to a commercially viable product.
Apropos of nothing, I'm guessing that you have never started a business. Just a guess, mind you...
A cynic would point out that the cost to develop these devices is very high, and companies must recoup their losses.
A mathematician would point out that, from a game-theory point of view, having one group of people come up with safety requirements without any burden of cost for the implementation of those requirements leads to stifling over-regulation.
Specifically, this leads to "safety at any cost", when in reality the cost of safety should be compared to other costs.
Here's a medical device that meets none of the standards of today, and it has bugs that need to be worked out.
There's a difference between "meets none of the standards" and "compliant, but untested".
Please don't sell something short by making unwarranted accusations.
The study of medicine has only one goal. Improve the life expectancy of human beings.
I believe the flaw in your argument is in this statement.
There is a cost of spinning reserve and grid stability maintenance. Why shouldn't those who need it or negatively impact it pay for it?
The answer has to do with the Prisoner's Dilemma. What is best for the individual may not be the best for society, and what appears to be a sub-optimal choice for the individual often turns out to be better for the individual in the long run if everyone makes the same choice.
For a concrete example, consider the Polio vaccine. There is a small chance of getting polio from the vaccine, so from the individual's point of view it makes no sense to get your kids vaccinated. If everyone else is vaccinated, you can forego the vaccination and reduce the risk to your child even further... except that if everyone makes that selfish choice, no one is protected.
In this particular case, there is now an incentive for people to purchase batteries and disconnect from the grid (or perhaps purchase batteries and not back-sell electricity). Instead of having the public service managing grid load using the best possible method with economies of scale, the problem is distributed among many small players. It's a narrow-minded view of an externality.
People have an incentive to spend revenue purchasing the batteries, and spend time dealing with installation/setup/maintenance/disposal rather than a few people doing this in bulk. The aggregate loss of productivity to society is much greater. Environmentally speaking, there will be lots of batteries in landfills in 10 years or so.
There is an incentive for less electricity put into the grid to be made available for others. The market has less "liquidity" now, generally considered (economically) to be a bad thing. The utility will need to generate more energy to compensate, which will result in more expenses and more environmental damage, and these changes may cost the utility more than the revenue from the monthly fees.
People without solar panels had the burden of an extra expense. Rather than viewing the expense as unwarranted, you could view it as an incentive to purchase solar panels, which is generally better for society.
Public services are corporations and as such want to make as much money as possible for their stock-holders. They only view this in terms of revenue - they could also consider the expense of externalities, and make decisions to maximize stockholder value rather than narrowly focusing on profit.
It's about 10^80, not 2^80.
Good catch. I misremembered the number from my original source, which also lists it as 10^80.
I'm going to use my 3D printer to print out my OWN artifacts and have my OWN Smithsonian and collect all that delicious entrance fee money...
With blackjack? And hookers?
Okay, it's even more awesome.
On the downloads page you can download the models in various forms - point cloud, mesh, and so on. Different formats, depending on the method used to get the model data (cat scan, laser scan, photographic, &c).
They mention in the about page that it would take 247 years of work 24/7 to capture the entire collection.
We could hire 247 people and get the entire collection online in 3 years (8 hour shifts). At $50,000 per person, that's about $13 million per year(*). Compare to the cost of the Obamacare website currently estimated at around $100 million and it has to be redone.
They've obviously shown "proof of concept" for getting the job done. Can we somehow just give them the money to complete it? Maybe a petition on "We the People"?
(*) Back-of-the-envelope calculation doesn't include cost of scanning equipment or materials, but note that there are a *lot* of museums in this country. We could invest in the infrastructure once and keep 300 people employed for decades putting this great stuff online.
Try some of the links from the article.
The system is really cool - you can rotate and view a mastodon skeleton (or supernova model) from any angle, it works in the browser, and it's really fast (on my machine, at least).
You can change the material color/texture, change the lighting properties & angle, take a slice of the 3-d image(!), put down a measuring tape, add annotations, then generate a link to that image.
I couldn't find a way to download the model, but you can file->save the web page, and it appears to store all the javascript for the editor on your local system.
The system is all kinds of awesome. Watching them add to their inventory should be very interesting.
We are also testing the boundaries of the physical universe in a completely new realm.
The number of states in a quantum-entangled set of particles goes exponentially with the number of particles. For 10 particles (entangled) it takes 2^10 states for the universe to represent the possible outcomes. For 1,000 entangled particles, the number of states is 2^1000.
The number of particles in the entire universe is only about 2^80.
Managing 1,000 entangled particles would require the universe to keep track of a staggering amount of information. Does the underlying machinery have information-space this big? No one knows.
For the first time, we can measure the boundaries of the physical mechanism that underlies the universe in a completely different realm: information capacity. This is analogous to a program probing the limits of RAM memory by seeing how much it can allocate.
Here's hoping that we don't find a buffer overflow.
Sure, which means getting rid of all income tax, capital gains, any form of 'service' as in no government, so buy a lot of guns, because the robber barons that were politely robbing you blind privately will be shooting you in your face for all those gold coins you have packed under your matress.
Really dude?
Is this the sort of argument that rates "+4 interesting" on slashdot nowadays?
Stop story-telling! We're nerds - we're better than that.
The value of bitcoins is up at the news of a hearing that is probably going to make it illegal to do 95% of what is done with BTC? That makes sense.
1) It doesn't matter what the US government decides, if people want to use it it will be used. Reference: drugs, prostitution, illegal immigration, abortion, illegal guns, and other world countries.
2) People don't act because they have no alternatives. This is an alternative, but not many people know about it. High-profile senate hearings publicize BitCoin, so that more people will realize how useful it is, and this spurs demand.
3) There are no valid argument against BitCoin. There are economic "story telling" arguments, and predictive "doom and gloom by story telling" arguments, and false equivalences with closely related things (fraudulent exchanges, Silk Road, &c), and outright lies ("it's a Ponzi scheme!!!"), but no actually valid arguments.)
4) All the standard aphorisms apply: buggy-whip manufacturers, the invisible hand, liquidity, privacy and freedom.
BitCoin will become a game-changer (oh, that phrase!) simply because nothing can stop it.
It's narrow mindedness in a government institutions, a common malady.
The DHS sees a need to stop some activity or other, and this makes perfect sense in context. One only has to look at the Syria and Egypt for examples of how this is used in practice - if the US ever descends into armed revolt, the switch will disrupt revolutionary communications and make it easier for the government to regain control. The military has its own, separate channels of communication.
Like all government institutions, it's narrow minded. They only think of themselves and their (DHS's) own needs, without regard to anything outside of their remit. Turning off the internet will have massive consequences to the economy locally and worldwide, but that is considered unimportant. There is no consideration of the action "in context" or the ramifications thereof, it simply achieves the goal.
They need this functionality, and other considerations be damned.
A recent post asked about "fear" and how it's used by the government to control people, but fear works in reverse as well. The government's actions are unsustainable (at the very least, economically) and it's control over the people is rapidly coming to an end. They're terrified of the end-game, and are putting pre-emptive measures into place ahead of time.
As someone previously said, elected officials are cutting everything except checks to supporters. If they stop handouts to supporters their control will fail. If cutting services causes massive unrest, they will fail.
They are between a rock and a hard place and getting squeezed harder every day. It will be interesting to see the end game when it happens.
(Look to the first month-or-two of 2014 as a possible start-date. That's when budget/debt limit talks start anew, and it's when everyone's health insurance costs will double. That's only a possibility, but next summer is looking really good for massive protests.)
What we need is a way to measure the money put into legislation like this. Perhaps a "kickstarter" for political action.
NetFlix probably hasn't put much into the political process. The ISPs have made more campaign donations, so this legislation is pretty-well doomed from the start.
We need a website where people can pledge donations to candidates who vote for or against specific legislation, sort of like Kickstarter for laws. Unlike kickstarter, people (corporations, too!) could pledge a specific amount either "for" or "against" a specific law. This would give lawmakers an easier and much more efficient way of judging which laws are most valuable to society.
This would also make political advocacy more efficient. Rather than donating to PACs or lobbyists, the public could send money directly to the pockets of the legislators involved. In economics-speech, It adds "liquidity" to this particular market - eliminating middlemen (who are only rent-seekers) and passing the savings along to the end-user.
The "invisible hand" of economics is often touted as the most modern and efficient way to solve a complex problem, yet we labor under an antiquated 250-year-old political system which is slow and inefficient.
Let's upgrade to more modern methods. We need a kickstarter system for laws.
Can I discuss some ideas with you offline? thon dot 9 dot okianwarrior at spamgourmet dot com
It clearly can't be based on floating-point binary representations. These are not scale-free. What I mean is, under one scale, you might have an error of 0.5 and you might say that needs 2 bits to store. Under another scale the same error would be 0.3333333 which takes infinite bits to store, under the same storage scheme. Even if you solve that, you'd get that a small but complex error (e.g. 1/pi^100) is worse than a huge but simple one (e.g. 2^64). That's just ridiculous.
So we have to use a metric where larger errors give a larger metric. A monotonic function, at least. So let's choose log2(error), right?
Ah, damn! If only you weren't anonymous, I'd love to continue this conversation offline!
That level of insight! I didn't expect anyone to be familiar enough with the foundations of statistics to make this observation.
You were correct up to the point where you chose log2(error) as a metric. There's a function which can be deduced from first principles, and which solves all the above-mentioned problems. Can you find it?
Back-edit the statistical methods and use the new error function instead of least squares. This leads to a solution to the fixed-width floating point problem, the regression problems mentioned above (including mixtures of gaussians mentioned in a side-post), and an algorithmic solution to the front-page article issue. (As a side-note, it's the first step towards hard-AI.)
I'm writing up my results right now - we'll see if my solution holds up under scrutiny.
The problem I have with least squares is that I don't like the definition of the "error". If you have two things that are correlated, one isn't necessariy a function of the other that includes some variability. If you flip the X and Y axes over - plot height against weight, rather than weight against height - then the least squares regression gives a different line. If the two errors are both minimised, but different, then neither of them is the "real" error.
Wow - brilliant insight! Thanks for that - things like this are why I come to Slashdot.
It seems like you found a lot of problems with what is taught in a stat 101 class. This is good; there are many. However, there are also solutions to these problems which you would find if you took a higher level course.
That brought a smile to my face. Thanks.
Actually, there is a really good reason to use least-squares regression. A model that minimizes squared error is guaranteed to minimize the variance of error, obviously.
This is the wrong place for an argument (you want room 12-A) and I don't want to get into a contest, but for illustration here is the problem with this explanation.
A rule learned from experience should minimize the error, not the variance of error.
It's a valid conclusion from the mathematics, but based on a faulty assumption.
Do outliers skew the results? If the outliers are biased, then that may tell us something about the underlying population. If they aren't biased, then their effects cancel.
There's no algorithm that will identify the outliers in this example.
But random data would generate statistically insignificant correlation coefficients. Also, the 95% confidence intervals used to predict values are wider for random data.
What value of correlation coefficient distinguishes pattern data from random data in this image?
Okay, here's the real problem with scientific studies.
All science is data compression, and all studies are are intended to compress data so that we can make future predictions. If you want to predict the trajectory of a cannonball, you don't need an almanac cross referencing cannonball weights, powder loads, and cannon angles - you can calculate the arc to any desired accuracy with a set of equations that fit on half a page. The half-page compresses the record of all prior experience with cannonball arcs, and allows us to predict future arcs.
Soft science studies typically make a set of observations which relate two measurable aspects. When plotted, the data points suggest a line or curve, and we accept the linear-regression (line or polynomial) as the best approximation for the data. The theory being that the underlying mechanism is the regression, and unrelated noise in the environment or measurement system causes random deviations of observation.
This is the wrong method. Regression is based on minimizing squared error, which was chosen by Laplace for no other reason that it is easy to calculate. There's lots of "rationalization" explanations of why it works and why it's "just the best possible thing to do", but there's no fundamental logic that can be used to deduce least squares from from fundamental assumptions.
Least squares introduces several problems:
1) Outliers will skew the values, and there is no computable way to detect or deal with outliers (source).
2) There is no computable way to determine whether the data represent a line or a curve - it's done by "eye" and justified with statistical tests.
3) The resultant function frequently looks "off" to the human eye, humans can frequently draw better matching curves; meaning: curves which better predict future data points.
4) There is no way to measure the predictive value of the results. Linear regression will always return the best line to fit the data, even when the data is random.
The right way is to show how much the observation data is compressed. If the regression function plus data (represented as offsets from the function) take fewer bits than the data alone, then you can say that the conclusions are valid. Further, you can tell how relevant the conclusions are, and rank and sort different conclusions (linear, curved) by their compression factor and choose the best one.
Scientific studies should have a threshold of "compresses data by N bits", rather than "1-in-20 of all studies are due to random chance".
First of all, it's Professor David Nutt, not "Nut"
Second of all, it's the same Professor who was a British government advisor, who was sacked for "criticising politicians for distorting research evidence and claiming alcohol and tobacco were more harmful than some illegal drugs, including LSD, ecstasy and cannabis."
Seems like a scientist with integrity. Perhaps this is less the risible ramblings of a madnam, and more he's at the "...then they laugh at you" part of fighting the good fight.
(Unless, of course, you think LSD and cannabis are more damaging than alcohol and tobacco, in which case feel free to poke fun.)
The Earth-Moon system orbits around a point which is not at the center of the Earth, sort of like a barbell with small and large ends will balance on a point which is not in the center of the big end. The point of rotation is still inside the Earth, just not at the center of the Earth.
This is why we have two tides [roughly] each day - the side near the moon gets attracted to the moon, while the other side gets swung around on the outside and experiences centrifugal force, pulling it away from the center.
The moon should experience this same effect. I would expect tidal forces to draw mass away from the center, resulting in more mass on the far side.