Say someone just asked you to measure the "average daily variations" for the temperature of your town (or for the stock price of a company, or the blood pressure of your uncle) over the past five days. The five changes are: (-23, 7, -3, 20, -1). How do you do it?
The STD of the series is 15.7 and the MAD is 10.8. NNT argues that MAD is more useful in this context.
In fact, whenever people make decisions after being supplied with the standard deviation number, they act as if it were the expected mean deviation...every time a newspaper has attempted to clarify the concept of market "volatility", it defined it verbally as mean deviation yet produced the numerical measure of the (higher) standard deviation.
He says he's seen this mistake made not just in popular articles, but also in financial publications and regulatory documents. The daily temperature is just a contrived example; he's mainly talking about financial analysis (which is his field).
All he is saying is that if you want to tell someone, e.g. what temperature change to expect from day-to-day, then the MAD is the right measure to use.
NNT's post doesn't at all state that MAD should be used instead of STD in any other context.
The "mean absolute deviation" is the mean absolute difference between each point and the mean of the points. Not the mean absolute difference between each point and the following point.
The point of the article is not that the standard deviation is "harder to explain," but that if you want to convey, say, how much the temperature could be expected to change on a day to day basis, then MAD is the better statistic. I'm sure NNT is aware that STD is better in most other contexts (e.g. comparing fits).
I think NNT is saying that the MAD ought to be used when you are conveying a numerical representation of the "deviations" with the intent that readers use this number to imagine or intuit the size of the "deviations." His example is that of how much the temperature might change on a day-to-day basis. According to him, it's not just that the concept is easier to explain, but that it is the more accurate measure to use for this purpose.
Based on his other work I'm sure he understands that the STD is generally superior for optimization purposes, fit comparison, etc.
Not 30 minutes ago I finished building a point model (sort of like this one), which I'll use for head tracking in combat flight sims. I tried using face tracking, but it wasn't reliable enough.
IMHO, head tracking is better than eye tracking. For head tracking, a multiplier is configured so that e.g. turning your head 20 degrees turns the game POV all the way around. Once you're used to holding a controller with your face, it's pretty easy / natural to use. On the other hand, it takes a lot of effort to keep your eyeballs from constantly saccading around the room.
It's good for device manufacturers, and bad for the folks in developing countries for whom a used computer or television is far preferable to a new one. I'm also not sure where it leaves the various charities which export functional but used devices to such countries.
This comment and its child sum this up well. The poster of the linked comment spent $90, a month's wages, on a 4 year old computer, and leveraged that device to eventually become a software engineer earning a lot more than $90 / mo.
That just isn't true. The news outlets he dealt with have been slowly releasing only the most damning documents in a highly redacted form. Thus far, while some programs have been reported on the basis of these documents, no operational or functional details have been revealed - only generalities.
Our insuranced-based aid programs (unemployment, Social Security, Medicare, Medicaid, FDIC deposit guarantees, etc) are contentious only because of 1) anti-government ideology or 2) a belief that the premiums (entitlement taxes) will prove insufficient to cover claims.
Further, most people do not understand how these programs are constituted and funded, so it is easy for politicians to treat them as if they were the same as discretionary spending programs like SNAP (food stamps).
China is a long way away from that stage of their economic rise. First their economy has to become capable of consuming its own outputs, so that they can allow wages to rise, the renminbi to appreciate, etc.
Even then, this wont be bad for the US, though it will certainly change our trade relations with China. A stronger renminbi will imply a weaker dollar, so it is likely the US current account would improve, possibly to the point of restoring a trade surplus. Concomitantly, "re-shoring" will occur, reducing US unemployment. The downside will be that we wont be able to consume Chinese resources and labor (in return for nothing more than US paper, I might add) any more. But we have plenty of our own...and like I said, the implication is increased production and employment in the US.
In reality domestic and international macroeconomics is not characterized by simple black-or-white, up-or-down relationships. Instead, there are trade-offs between different types of economic roles. Mercantilism (the belief in a fixed quantity of wealth) is false, because incoming solar energy and human ingenuity create more wealth than was represented by Earth's initial allotment of scarce resources. The implications to policy of these facts are critical, because optimal growth for individual nations is no longer obtained via beggar-thy-neighbor approaches.
Great point. Likewise, real terms of trade with China are in our favor. They exchange their real resources and expend their labor in return for nothing more than our notionally valued financial assets (aka fiat money).
Nothing is safer than US Treasury securities in terms of default risk. The reason being that the US is the sole issuer of dollars, borrows solely in dollars, and permits the dollar to float internationally. This means that the US has no commitments as to the total number of dollars which exists, and there is zero risk of a forced US default. This is actually true for any country which is the sole issuer of its currency, borrows solely in that currency and permits that currency to float. Canada, Australia, Great Britain and Japan are further examples, unlike China and Russia (who defend exchange rate pegs against other currencies), or Eurozone nations (who do not issue their currencies).
Of course, there are other important risks, such as inflation risk, but despite what some ratings agencies seem to think, inflation is not a "form of default."
US Securities (and similar) are important to financial markets for a number of reasons, but one of the main ones is that they tell us what the interest rate should be for a zero risk asset. This provides a baseline which helps them determine what rates ought to be earned by assets which do have default risk.
You're conflating the operational aspects of private lending and so-called public borrowing via bond sales. First, the US federal government is the sole issuer of the dollar, borrows solely in dollars and permits the dollar to float against other currencies. Holders of dollar assets (reserves or Treasury securities) have no special rights of exchange or conversion. It is therefore not possible for the US federal government to undergo a forced default.
Second, as a matter of logic, the government as sole currency issuer spends first and taxes or borrows afterwards. In fact, currently the government spends simply by crediting bank accounts. This constitutes a 'reserve add' which tends to push down interest rates and increase inflationary pressures. They counter these pressures by conducting a subsequent 'reserve drain,' either through taxation (which extinguishes private sector assets) or through bond sales (which maintain private sector asset levels but increase the interest rate slightly).
China can and more importantly, is attempting to do
What are you even talking about here? This fake malware attack or selling us stuff?
Real terms of trade with China are unambiguously in our favor as they exchange their precious natural resources and unrecoverable labor time in exchange for our notionally valued financial assets (fiat money).
Again, I think you should explain how exporting all of their resources and labor is supposed to make them stronger. Sure, there is a bigger number next to their name in a computer at the Fed. So what?
your simplistic model does not account for enough variables to draw any meaningful conclusions
There is no "model" in what I stated above - just facts. The fiscal deficit is the difference between government liabilities issued and extinguished in a given period. These liabilities are financial assets in the private sector. In other words, they are slips of paper with a government stamp and some notional value arising from factors such as the government's willingness to accept them in tax payments and their usefulness as a medium of exchange. In contrast, food, minerals and labor are real resources which have an intrinsic value in that we in fact require them to live (as well as to produce goods, develop military power, etc.). Financial instruments and real resources are fundamentally different entities.
I would disagree - we do, in fact, face real inflationary pressures
CPI has been low and stable for decades now. We still see significantly inflationary pressures in food, but energy prices are beginning a long downward trend due to recent exploitation advancements, and the hefty allotment of real resources which we took from the natives^W^W^W^W are lucky enough to possess. In manufactured products and services, there has been strong price deflation due to globalization and technological progress. The financial crisis also precipitated a major deflationary episode worldwide as commodity prices and employment collapsed. Neither the global or national economies have yet made up for what was lost, and remain depressed far below their former output and growth paths.
There has been an effective drop in real wages for decades
Yes - this is a strong deflationary pressure. Outright consumer price deflation has been rare in developed economies, which is explained by behavioral studies demonstrating that people (employers and employees) actually prefer layoffs to wage reduction. Thus we get persistent low inflation and unemployment rather than widespread deflation.
Meanwhile, people often "feel" that inflation is high, even though it isn't, because it is actually becoming harder for them to afford purchases. However, this should be blamed on wage suppression, and not inflation which as stated is both stable and low by any historical standard.
There are three major sources for deflationary pressures within the domestic US economy (also commonly called demand leakages). Namely, they are productivity growth, population growth and the trade deficit. The combined inflationary pressures of private credit and federal deficits are insufficient to produce either moderate inflation or reduced unemployment against them.
The deficit would be fact #1 - definitely a resource shortfall there.
A federal deficit is when the federal government issues financial liabilities (such as dollars or bonds) in excess of those extinguished by taxes.
On the other hand, a resource shortfall occurs when there are insufficient real resources (such as food, minerals or labor) given the needs of the people who make up a society.
You have misunderstood the proposition. The fiscal stance of the central government does not imply anything at all with regards to real resources. The fact is that in the US we are currently under-utilizing available real resources (including physical and human capital). For example, we are not suffering from shortfalls in concrete, food or unemployed workers, and there are plenty of outstanding needs (such as infrastructure repair) which could be satisfied by those resources. If we faced real resource constraints, the deficit would lead to price inflation. Inflation is currently low, despite the deficit, in part because of this output gap.
The people I'm addressing belong to a criminal organization
You must either think that it doesn't matter what specific crimes are committed, or that non-violent crimes (like illegal spying) should be punished by death.
Good point, however I believe OP was referring to members of the public who would otherwise support defensive spying but currently see a lack of evidence that the NSA's activities have legitimate defensive purpose.
1) There was never any day of "virtually unlimited budgets."
2) Do you assert that the United States currently faces specific real resources shortfalls, even given the current large output gap? If not, can you propose a specific, realistic mechanism why the United States would currently face fiscal constraints, even given persistently low inflation?
Slashdot Mirror
Not within one day, but across several days.
Say someone just asked you to measure the "average daily variations" for the temperature of your town (or for the stock price of a company, or the blood pressure of your uncle) over the past five days. The five changes are: (-23, 7, -3, 20, -1). How do you do it?
The STD of the series is 15.7 and the MAD is 10.8. NNT argues that MAD is more useful in this context.
In fact, whenever people make decisions after being supplied with the standard deviation number, they act as if it were the expected mean deviation...every time a newspaper has attempted to clarify the concept of market "volatility", it defined it verbally as mean deviation yet produced the numerical measure of the (higher) standard deviation.
He says he's seen this mistake made not just in popular articles, but also in financial publications and regulatory documents. The daily temperature is just a contrived example; he's mainly talking about financial analysis (which is his field).
All he is saying is that if you want to tell someone, e.g. what temperature change to expect from day-to-day, then the MAD is the right measure to use.
NNT's post doesn't at all state that MAD should be used instead of STD in any other context.
The "mean absolute deviation" is the mean absolute difference between each point and the mean of the points. Not the mean absolute difference between each point and the following point.
The point of the article is not that the standard deviation is "harder to explain," but that if you want to convey, say, how much the temperature could be expected to change on a day to day basis, then MAD is the better statistic. I'm sure NNT is aware that STD is better in most other contexts (e.g. comparing fits).
I think NNT is saying that the MAD ought to be used when you are conveying a numerical representation of the "deviations" with the intent that readers use this number to imagine or intuit the size of the "deviations." His example is that of how much the temperature might change on a day-to-day basis. According to him, it's not just that the concept is easier to explain, but that it is the more accurate measure to use for this purpose.
Based on his other work I'm sure he understands that the STD is generally superior for optimization purposes, fit comparison, etc.
Not 30 minutes ago I finished building a point model (sort of like this one), which I'll use for head tracking in combat flight sims. I tried using face tracking, but it wasn't reliable enough.
IMHO, head tracking is better than eye tracking. For head tracking, a multiplier is configured so that e.g. turning your head 20 degrees turns the game POV all the way around. Once you're used to holding a controller with your face, it's pretty easy / natural to use. On the other hand, it takes a lot of effort to keep your eyeballs from constantly saccading around the room.
Invisible Man is one of my favorite books. Isn't about time you started stealing electricity and smoking joints to Satchmo?
It's good for device manufacturers, and bad for the folks in developing countries for whom a used computer or television is far preferable to a new one. I'm also not sure where it leaves the various charities which export functional but used devices to such countries.
This comment and its child sum this up well. The poster of the linked comment spent $90, a month's wages, on a 4 year old computer, and leveraged that device to eventually become a software engineer earning a lot more than $90 / mo.
released every fucking piece of information
That just isn't true. The news outlets he dealt with have been slowly releasing only the most damning documents in a highly redacted form. Thus far, while some programs have been reported on the basis of these documents, no operational or functional details have been revealed - only generalities.
Our insuranced-based aid programs (unemployment, Social Security, Medicare, Medicaid, FDIC deposit guarantees, etc) are contentious only because of 1) anti-government ideology or 2) a belief that the premiums (entitlement taxes) will prove insufficient to cover claims.
Further, most people do not understand how these programs are constituted and funded, so it is easy for politicians to treat them as if they were the same as discretionary spending programs like SNAP (food stamps).
This hurts our entire economy.
Maybe a little. About 33% of the Gripen is sourced from the US [1].
China is a long way away from that stage of their economic rise. First their economy has to become capable of consuming its own outputs, so that they can allow wages to rise, the renminbi to appreciate, etc.
Even then, this wont be bad for the US, though it will certainly change our trade relations with China. A stronger renminbi will imply a weaker dollar, so it is likely the US current account would improve, possibly to the point of restoring a trade surplus. Concomitantly, "re-shoring" will occur, reducing US unemployment. The downside will be that we wont be able to consume Chinese resources and labor (in return for nothing more than US paper, I might add) any more. But we have plenty of our own...and like I said, the implication is increased production and employment in the US.
In reality domestic and international macroeconomics is not characterized by simple black-or-white, up-or-down relationships. Instead, there are trade-offs between different types of economic roles. Mercantilism (the belief in a fixed quantity of wealth) is false, because incoming solar energy and human ingenuity create more wealth than was represented by Earth's initial allotment of scarce resources. The implications to policy of these facts are critical, because optimal growth for individual nations is no longer obtained via beggar-thy-neighbor approaches.
Great point. Likewise, real terms of trade with China are in our favor. They exchange their real resources and expend their labor in return for nothing more than our notionally valued financial assets (aka fiat money).
Nothing is safer than US Treasury securities in terms of default risk. The reason being that the US is the sole issuer of dollars, borrows solely in dollars, and permits the dollar to float internationally. This means that the US has no commitments as to the total number of dollars which exists, and there is zero risk of a forced US default. This is actually true for any country which is the sole issuer of its currency, borrows solely in that currency and permits that currency to float. Canada, Australia, Great Britain and Japan are further examples, unlike China and Russia (who defend exchange rate pegs against other currencies), or Eurozone nations (who do not issue their currencies).
Of course, there are other important risks, such as inflation risk, but despite what some ratings agencies seem to think, inflation is not a "form of default."
US Securities (and similar) are important to financial markets for a number of reasons, but one of the main ones is that they tell us what the interest rate should be for a zero risk asset. This provides a baseline which helps them determine what rates ought to be earned by assets which do have default risk.
You're conflating the operational aspects of private lending and so-called public borrowing via bond sales. First, the US federal government is the sole issuer of the dollar, borrows solely in dollars and permits the dollar to float against other currencies. Holders of dollar assets (reserves or Treasury securities) have no special rights of exchange or conversion. It is therefore not possible for the US federal government to undergo a forced default.
Second, as a matter of logic, the government as sole currency issuer spends first and taxes or borrows afterwards. In fact, currently the government spends simply by crediting bank accounts. This constitutes a 'reserve add' which tends to push down interest rates and increase inflationary pressures. They counter these pressures by conducting a subsequent 'reserve drain,' either through taxation (which extinguishes private sector assets) or through bond sales (which maintain private sector asset levels but increase the interest rate slightly).
I have discovered a truly marvelous proof of this, which this comment is too narrow to contain.
China can and more importantly, is attempting to do
What are you even talking about here? This fake malware attack or selling us stuff?
Real terms of trade with China are unambiguously in our favor as they exchange their precious natural resources and unrecoverable labor time in exchange for our notionally valued financial assets (fiat money).
Again, I think you should explain how exporting all of their resources and labor is supposed to make them stronger. Sure, there is a bigger number next to their name in a computer at the Fed. So what?
Yeah, their dastardly plan to trade all of their resources and labor away in exchange for our paper is really crushing us.
your simplistic model does not account for enough variables to draw any meaningful conclusions
There is no "model" in what I stated above - just facts. The fiscal deficit is the difference between government liabilities issued and extinguished in a given period. These liabilities are financial assets in the private sector. In other words, they are slips of paper with a government stamp and some notional value arising from factors such as the government's willingness to accept them in tax payments and their usefulness as a medium of exchange. In contrast, food, minerals and labor are real resources which have an intrinsic value in that we in fact require them to live (as well as to produce goods, develop military power, etc.). Financial instruments and real resources are fundamentally different entities.
I would disagree - we do, in fact, face real inflationary pressures
CPI has been low and stable for decades now. We still see significantly inflationary pressures in food, but energy prices are beginning a long downward trend due to recent exploitation advancements, and the hefty allotment of real resources which we took from the natives^W^W^W^W are lucky enough to possess. In manufactured products and services, there has been strong price deflation due to globalization and technological progress. The financial crisis also precipitated a major deflationary episode worldwide as commodity prices and employment collapsed. Neither the global or national economies have yet made up for what was lost, and remain depressed far below their former output and growth paths.
There has been an effective drop in real wages for decades
Yes - this is a strong deflationary pressure. Outright consumer price deflation has been rare in developed economies, which is explained by behavioral studies demonstrating that people (employers and employees) actually prefer layoffs to wage reduction. Thus we get persistent low inflation and unemployment rather than widespread deflation.
Meanwhile, people often "feel" that inflation is high, even though it isn't, because it is actually becoming harder for them to afford purchases. However, this should be blamed on wage suppression, and not inflation which as stated is both stable and low by any historical standard.
There are three major sources for deflationary pressures within the domestic US economy (also commonly called demand leakages). Namely, they are productivity growth, population growth and the trade deficit. The combined inflationary pressures of private credit and federal deficits are insufficient to produce either moderate inflation or reduced unemployment against them.
The deficit would be fact #1 - definitely a resource shortfall there.
A federal deficit is when the federal government issues financial liabilities (such as dollars or bonds) in excess of those extinguished by taxes.
On the other hand, a resource shortfall occurs when there are insufficient real resources (such as food, minerals or labor) given the needs of the people who make up a society.
You have misunderstood the proposition. The fiscal stance of the central government does not imply anything at all with regards to real resources. The fact is that in the US we are currently under-utilizing available real resources (including physical and human capital). For example, we are not suffering from shortfalls in concrete, food or unemployed workers, and there are plenty of outstanding needs (such as infrastructure repair) which could be satisfied by those resources. If we faced real resource constraints, the deficit would lead to price inflation. Inflation is currently low, despite the deficit, in part because of this output gap.
You must either think that it doesn't matter what specific crimes are committed, or that non-violent crimes (like illegal spying) should be punished by death.
Good point, however I believe OP was referring to members of the public who would otherwise support defensive spying but currently see a lack of evidence that the NSA's activities have legitimate defensive purpose.
1) There was never any day of "virtually unlimited budgets."
2) Do you assert that the United States currently faces specific real resources shortfalls, even given the current large output gap? If not, can you propose a specific, realistic mechanism why the United States would currently face fiscal constraints, even given persistently low inflation?
Too true. I would also nominate Richard Posner (even though his politics are somewhat contrary to mine). No idea what Alsup's politics are.