Saturday, July 22, 2017

Statistical wellness: the current face of eugenics

Eugenics are back! And in some cases being enforced by governments!

Ok, so that bears some explanation. But I’m going to have to start with some basics.

The first thing to understand is that we don’t know how our bodies work. We see what they do and have some really good ideas about what’s going on, but we’re not really sure. This is why drugs still have to go through trials cause it’s all guessing. Where we differ from the age of leeches, is statistics. We started using data across populations to figure out what worked and what didn’t and we started to correlate different metrics for quantifying a person. Doing this made a leaps and bounds improvement to medicine.The problem is that the low hanging fruit of obvious and useful correlations have all been  identified. Now all the correlations are part of more complicated systems so they are much harder to figure out. The other thing to understand is that most of the work has been targeted and dealing with illness. All the metrics we use are targeted for use when the patient has a problem, as part of a process to guess what the problem is. It’s not a process that thinks it’s definitely going to get it right on the first try, so it isn’t optimized for that. It’s optimized for the statistically best outcome.

The second thing to understand is that we’re not all the same. A person whose growth has been stunted might be taller than a different person in perfect health. We vary. There are members of our population that stand out. Sometimes when we identify these people we learn something. Most of the time we just don’t understand. Averageness and healthiness are disconnected.

With our wellness programs we’ve decided to use our diagnostic tools to evaluate people. Which with my setup leads to, What happens when someone that’s not average goes in for one of these checkups?   They get identified as potentially having a problem and start to get treated for it. In all likelihood they’ll either continue to be treated to cure their healthiness for the rest of their lives, though instead of being treated for one non-problem, they may have a doctor that wants to aggressive track down the non-problem, and keeps changing up their mistreatment. What happens when you keep trying to fix something that isn’t broken? It’s going to break! Wellness programs are by design going to harm the atypical. Then why do we do it? Because it’s statistically “better”. Where better is a metric that includes lifetime cost of care. “Better” might also consider some other factors like life expectancy, but a treatment that kills half the people that take it instantly but slightly more than doubles the life expectancy of the rest is also “better”. Not to mention that some atypical people who would benefit from being checked, won’t because they’re unhealthily (for them) average. The is a system that helps the “average” at the cost of the atypical.

Which leaves us with, the question, if we keep coming back to eugenics, is it really wrong? Objectively, Yes! Favoring an ideal leads to a monoculture. Monocultures are susceptible to extinction level pandemics. We’re in the middle of the 2nd one for commercial bananas. Also for sexually reproducing populations you have the genetic bottle neck which leds to all the genetic problems from inbreeding. Beyond that, you have to remember that we don’t really understand how our bodies work. What ever statistical ideal we pursue, could be tragically flawed. We’re more likely to make a bad choice than a good one. Also having atypical people let’s us learn from their advantages.

Unfortunately, in our world, wellness programs are likely here to stay. But there is a light of hope on the horizon. Big data medicine. With access to comprehensive records of your vital stats, doctors will know how you are doing against your own baseline. And with much much larger populations of People for which there is data, docotors will be able to come up with multiple models and find one that actually fits you instead of basing your prognosis on a bogus one size fits all model.

Sunday, January 15, 2017

Metric system silliness.

I'm not sure even where to begin with the metric system. It was a great idea in principle, but doesn't actually work out in the real world. And the acrobatics that its advocates so through to keep it alive just make things worse, and are rather silly.

So the idea behind the metric system was to have one system that unified measures of distance, volume, mass, temperature, and more based on the natural constants of water. So that any unit can be derived from others based on water. And further to have ranges based on multiples of 10, with set prefixes so that addressing things at different scales need not see inflation of digits. This definition didn't last but you'll still see memes lauding the metric system's superiority because of these relations. 

The water based definitions for the metric system didn't last because they were too naive. The idea was that one or two measures plus water would let you derive any third measure, but it turns out that in the real world, it's not that simple and you really need pretty much all the other measures to derive a missing measure. Not to mention that water is seldom as pure as we'd like.  So the water definitions were dropped. From official adoption in 1795 this idea didn't make it 4 years before being replaced.

But before the metric system was even out of the gates it already started to be crippled. Maybe it's always struck you as odd that the default mass unit has the kilo prefix. It is odd. The original mass unit was the grav. Unfortunately grav sounds like graff, and graff was a title of minor nobility at the time, when nobility was not in favor. However the milligrav has already taken on a more colloquial name of gram. So they went with that instead, skewing the alignment of the base units and their prefixes. 

The system that replaced the water system was prototypes. That is to say they made specific objects and set them aside somewhere to be the definition. Of course they didn't make just one of these, they made many as identical as possible, though a specific one was named to be the actual factual definition. Obviously switching to this system removes the smug superiority from the metric system. It also means that the system is based on items that can get lost or destroyed, which isn't great. It also, as it turns out, was still too naive. These prototypes were replaced at least once, to improve the material science, but that's still too naive, because our technology just wasn't in place. As the duplicates of the prototype when returned and compared to the prototype, were no longer the same. Which really undercuts any faith in the system.

As a next step for the metric system they want to turn them back into natural constants. This has already happened for the meter. They have chosen to define in terms of the speed of light in a vacuum. The consequence of this that I think is the silliest, is that this means you can't measure the speed of light in a vacuum with the metric system. What ever result you get the answer is by definition, not by measurement. This seems like a silly thing to do with such an important constant. Additionally there are just the basic questions about this constant and our knowledge of it. Have we ever has access to a perfect vacuum? Have we had sufficiently fixed endpoints in any sort of vacuum to get a good measure of this? Don't we measure everything else with light, so how can we calculate margins of errors for calculating the speed of light, when those margins depend on the speed of light? And I'm not enough of a physics guy to get deeply into relativity, which fixes it, and makes it so that it the wavelength of light from moving things that change and not the speed, but also has time and space deform. Which when you go from pure mathematics of theoretical physics to the real world still has perhaps some degree of hand waving going on.

For the Kilogram, one idea at least, is to make a prototype in the shape of a sphere out of an atomic molecule, measure it's diameter, calculate the number of atoms in it, and then base the definition of the kilogram based on that. So here material purity rears it's head, and in this case not even ions are allowed. Next you have that they will calculate the number of atoms in a sphere. based on it's diameter, cause nothing says imprecision like including an irrational number. But then you have a similar problem to light, determining the mass of electrons, protons, and neutrons. The mass unit will be defined in terms of the combination of all three of them. So really you can only measure their proportionality. And once you have their proportionality, you can never measure the mass of anything again. All you can do is count it. The only free variables when considering the mass of something will be what it is and how much of it is there. If you have those answers then the mass is by definition, and the proportionality of those 3 particles. Even the mass of other subatomic particles might be in fixed if quarks have fixed masses. This seems monumentally foolish and roundabout to me. 

There is probably more silliness in the history of the metric system. However the question that's been avoided is why is the imperial (or any other) measurement system better? Simply stated because they don't try so hard to be perfect, and certainly no one one goes on about why the imperial system is better based on it's vast and constant technical superiority. The smug adherents to the imperial system normally put forth it's cultural superiority, which is an argument that defeats itself. 

I think really the thing that irritates many (including myself) is that it speaks to the hubris of people ready to declare themselves in nearly full understanding of the universe. Time and time again such people have been found to have been lacking. And even worse sometimes their well intentioned ideas survive and are used as ammunition against improved understanding, despite having been originally put forth to promote understanding. The math used by theorists is precise and perfect, the real world is not, and only fits approximately within precise models. As we learn those approximations improve. They may never be perfect, but that's good since it pushes us to look deeper. The ethos of the metric system is opposed to that, and I guess that it why it irritates me so much and find humor in the foibles of the supposedly perfect system. 

Sunday, January 1, 2017

Why Pluto is still a Planet.

In the early 2000's as we started discovering additional large objects in the outer solar system, astronomical displays started to clump these objects in with Pluto. The rate at which we were discovering them made it seem likely that there would be many. And who wants to memorize 30 different planets? Which if you're lazy and don't want to be embarrassed when little kids know all the planet names but you, who are the actual astronomy professional, don't, means that Pluto has to go. That's it. Don't be fooled everything else is just a smoke screen. Pluto had to go so that a bunch of people who were full of themselves wouldn't have to keep memorizing new lists of planets that might subtly change order. Think about how often these same people, who should know very well that Pluto has a trans neptunian orbit, refer to Pluto as being beyond Neptune.

There is a case for Pluto's demotion. We didn't know how big it was when we discovered it. We thought it was bigger. We didn't know it was in a belt of objects, some of which were big. After it was discovered Ceres was briefly considered a planet, until we realized that it was part of a belt. And while it remains the largest and the only one large enough to pull itself into a sphere, it remained demoted to the rank of asteroid, until this whole kerfuffle, so again there was a reason to be considering these things.

But at the same time as we were finding new things in our solar system we were finding "Planets" in other solar systems, and even looking for ones that weren't captured by a star, that are called rogue planets. This is the back drop of the state of the art science at the time. We were looking for bodies outside of our solar system that didn't orbit stars that we were calling rogue planets. Remember that cause I'm going to come back to it. Also it's very possible that in other start systems we will find bodies in all sorts of exciting and strange orbital configurations.

So with a reason to consider the designation of things in our solar system, and the desire to limit how often school children embarrass them, the IAU set out. The definition that they cam up with is "(1) A planet is a celestial body that (a) is in orbit around the Sun, (b) has sufficient mass for its self-gravity to overcome rigid body forces so that it assumes a hydrostatic equilibrium (nearly round) shape, and (c) has cleared the neighbourhood around its orbit".

Of the planet definition criteria b is a reasonable criteria. It is a clear dividing line that objects throughout the universe will be on one side of or another, and as such will be useful with all the extra-solar astronomy that is going on as well as with what's going on in the Solar system. This criteria is very broad though and in addition to what had been considered planets before it also captures stars and many (but not all) of the moons in the solar system. 

Criteria a defines planets as being only in orbit around the Sun, and as a consequence there are only planets in our solar system. No extra solar planets, no rogue planets. Planets are just something we have. This is rubbish of course. Further it is complicated by just what that means. In orbital dynamics there is something called the Hill sphere, which is used when looking at two bodies to see if objects nearer the smaller one are orbiting it or the larger one. Right now the moon is orbiting the Earth, and the Earth the Sun. As little as ~1.27 million years ago when the moon was closer, the earth orbited the moon not the sun. Without wanting to do the math I would guess that for Pluto and Charon it is the case that they orbit each other and it's only their system that is in orbit around the sun. So this is the sort of things we should expect to see out there, so if we apply this definition to other systems and other suns it still will eventually break. This criteria wasn't forward thinking towards improving a definition for science. It's stated really naively so you have to guess that they are talking about Hill sphere's for the criteria, otherwise everything in the system orbits the sun. Also this criteria is part of the dwarf planet definition, so if Pluto and Charon are in orbit around each other then they are not even Dwarf Planets, or the framers we being even more naive and less scientific than that. 

After what I did to criteria a, I'm sure that you are expecting that I think criteria c is a pure mess. It is. It's so bad and so naive that people keep coming out with "interpretations" of it, to fix the problems. So let's begin with the big one. Pluto and Neptune. Pluto crosses Neptune's orbit. How can you have cleared your orbit when there is something that big that crosses it? Then you have things like Earth and Cruithne, an asteroid in a 1:1 orbital resonance with the earth that crosses our orbit but isn't technically in orbit around us. And add on to that trojans. The original trojans were in Jupiter's orbit, but it's such a thing that we've generalized the term for all the planets, and they are objects in the same orbit, at the same orbital period, that are in the right spot to roughly never get cleared out of an orbit by a planet. That's right we know, and knew when the definition was made, that no planet could completely clear it's orbit. This definition is a lie. It's a lie that was put in place to kill Pluto and keep the rest of them out. And beyond the problems in our solar system there are lots of possible interesting orbital dynamics that we might find out there, that won't fly with this either. Sure a lot of them won't be super stable, but what's the threshold of stability? Nothing in our solar system is actually stable, so it can't be forever, and then after that you let in a lot of interesting things that we are going to find out there.

So the current definition is unscientific and garbage. As such I refuse to accept it. And revert to the previous state in which Pluto was a planet. If some one tries to argue for the demotion of Pluto, there is a simple question you can use to break them, "What is the difference between a rogue planet and a rogue dwarf planet?". You don't have to give them anything harder to think about than that question which the framers of the new definition, left unanswerable, which counter to the mandate for even considering the definition. In response to searching for and finding lots of new things they revised the definition to not be applicable or otherwise be nonsense to the new? No. Just No. If a future definition gets put forth that's good and demotes Pluto, I'll accept it. But I refuse to accept that trash just like I refuse to accept other garbage that wants to claim to be science.

What is required for a good definition? It has to consider all the types of things we're looking for, so it should consider objects in other sun systems as well as those in between them. It should be based on the intrinsic properties of a body not extrinsic ones. We're not going to have perfect knowledge of everything in a solar system when we first see a body, so it's designation shouldn't rely on other bodies that aren't clearly playing a major role in what's happening with that body. Also the definition should be very specific. It should start as a hard computable thing that is technically precise that can be summarized for humans, rather then something naively high level that might melt down when you actually get into the details. 

I don't know if a body will ever put out a good definition that takes off. But I have some predictions. I predict that we'll define some new term like planetoid to account for objects between solar systems. Of objects that pull them selves into spheres, I think the divisions are more like to keep Pluto with Earth, than Earth with Jupiter. I think that after we define terms like planetoid (potentially with sub groupings) we will then define planet and moon based on orbital relations so that the Earth and the Moon can be the same type of object but will also have terms to describe them in terms of a start system. (I think moon and satellite will be split in this situation). And I think that in systems of mutuality like Pluto and Charon, the two are likely to hold the same status.