Friday, December 22, 2017

Naive Generational Influence

"Good Warriors"? Who chooses these names?

Reading yet another article trying to assign generational responsibility for the current state of America I realized that I didn't have the best grasp of generation definitions or population distribution across the decades. That inspired me to whip up a quick spreadsheet to compute a simple measure of political influence across time, under the premise that ultimately responsibility lies with the electorate. Since generation definitions are debatable and birth rates can vary a lot I did it year by year and then turned it into a generational overview afterward. The goal isn't to directly make any responsibility assignment (an arguably pointless aim) but to examine what people's contribution over time would've been minus any idiosyncratic events (i.e. all other factors being equal) and therefore provide a better starting intuition for such factors.

Births in a given year: I neglected the issues of infant mortality, woman's suffrage, slavery, and voter suppression generally as this is intended as a rough analysis only and those can arguably be included in the historical idiosyncrasies I mentioned earlier. Here is the birth data by year, which required going back as far possible so that boundary issues are minimized.

Years of political activity: As an estimate of the years during which an individual can have political influence on society, I chose a starting age of 18 and an ending one of 75. All years are counted, not merely election years as political activity can (and should) extend beyond the mere act of casting a ballot. This chart shows the number of voters in each generation of interest that are in the age range of 18 to 75 over time, with preceding generations lumped together.


Political Inertia: To reflect the time it takes for the course of events to be altered, I employed an exponential decay model. Here is the generational influence with a couple choices of decay rate:


Since exactly what decay rate to use is difficult to determine, I plotted the share of generational responsibility for the state of affairs at the end of 2017 across the entire range of possible values:


The values approach a limit fairly quickly and quite a bit of the full range is rather extreme anyway, so here is the area with a decay rate less than ten percent:


Even given the simplistic nature of this analysis, it's pretty clear that one group shares little responsibility for the current state of affairs (for good or ill): New Boomers, or "Millennials" to use the more common label. Additionally, for all their numbers, Baby Boomers maximum value of impact for any choice of decay rate is 32.5%. While certainly not a small share, previous generations had a vast, ongoing influence.

More generally, I would argue that this basic model serves as a useful basis for considering more nuanced historical, societal, and psychological influences. I also would strongly advise against assigning more significance to it beyond that. Indeed, the major takeaway is that previous generations didn't simply evaporate once newer ones came of age. Patently obvious, but not properly reflected in the tone of many articles on the topic I come across.

Wednesday, August 17, 2016

Pop Up Politics


Live debates are a terrible way to vet a president, but it's a tradition we appear to be stuck with. One way to improve them, and political speeches generally, would be a more integrated method of fact checking, perhaps along the lines of VH1's Pop Up Video format. In addition to the videos, a website with verifiable fact checking sources would be required for true credibility. There should be more than facts, though, as keeping it entertaining through humor will help more people stay engaged. I believe this is the driving force behind why Daily Show viewers have hisotrically had a high level of news knowledge: because they watch consistently. The popularity of that (I submitted the suggestion to them in 2008 and again in 2012, but never received a response) or a similar program would help offset diminished interest stemming from the fact that the political events would have already aired and they'd have the resources to achieve a rapid turnaround time. 

Saturday, July 4, 2015

Turn Signal Enforcement

If we only had hand signals there'd be an actual excuse.
Failure of drivers to signal is one of my pet peeves (the ratio of personal to public benefit being so small) so I was reflecting on what it would take to actually discourage since it is widespread and considered minor enough that human based methods are impractical. Of course this is purely a hypothetical exercise as such a universal system would be difficult to implement and likely meet with widespread condemnation (a combination of properties which seem to be the unofficial theme of my blog).

While it is certainly possible in theory to have a GPS based system that would attempt to catch every infraction, this not only would create all sorts of problems and be needlessly expensive, but is also unnecessary. All that's required is to ensure that drivers are not failing to signal chronically. To that end, a system that detects whether or not a signal was used before a maneuver was taken and keeps track of the overall ratio would be adequate. The difference between a hard turn and a lane shift would need to be accounted for and very low speed failures likely excluded but such hardware should not be difficult (and is something we've had the capability to implement for decades). Of course the device would need to be tamper resistant to the extent it discourages most drivers from attempting it.

What would be done with such information? While the obvious use would be as a penalty, a carrot and stick approach would likely be more effective. Given the coarse nature of the data, perhaps a three tiered system: chronic offenders get a punishment, reasonable signalers are left alone, and fastidious ones rewarded with a reduction in insurance rates. The magnitude of those two extremes should be informed by research into the relationship between signaling and accident rates.

Not only will this never happen but I suspect it's too late in the car's life cycle for it to make sense now even were the political will there. With luck, in the next few decades self-driving cars will largely end the signaling problem.

Sunday, May 3, 2015

Cosmic Ray Mitigation

I do enjoy a good pun.
I have mentioned previously that I am opposed to manned space exploration as it now exists, but am absolutely in favor of continued research (of a reasonable magnitude relative to that in other areas, as robotic exploration is advancing quickly) since my objections are of a practical nature. This is especially true because often there are unintended applications of such study. It occurred to me this morning that perhaps one of the most significant crossover discoveries might come from one possible technique to mitigate the effects of long term cosmic ray exposure: cure cancer. Even this manned space travel skeptic would have a tough time arguing against that.

Friday, May 1, 2015

Global Disaster Insurance

Radar image of the ground changes due to the 2015 Nepal earthquake.
Rather than the patchwork of often ad hoc disaster assistance systems we have now, really all nations should be part of a global disaster insurance plan. This would provide is certainty as to what degree of response would be mounted so that planning can be done and in theory lead time could be reduced. It would force not only a systematic evaluation of all the risk factors each nation faces but also provide economic incentive to take preventative steps to lessen them. Too often a disaster occurs and obvious, generally inexpensive (relative to the costs of the damage itself) preventive measures come to light. It bears mentioning that a large part of any risk assessment would include the impact of corruption on general preparedness, again providing incentive for solutions to be found before such systems are needed. It is tempting to think this would be useful to only less developed nations, but many other faults, such as Boston's earthquake deficient buildings, could benefit from a sober assessment of the risks and the price of potential remedies, especially placed in a global context.

Note that there is nothing about this proposal that would lessen the opportunity for charitable foreign assistance. In fact, by better highlighting the problem areas and encouraging more efficient mitigation efforts it should vastly improve it.

Sunday, March 29, 2015

The Supreme Consensus


I've posted about how to increase the influence of science in government previously but it was a complete overhaul of our entire democracy, so wanted to explore less drastic alternatives. Modeled on the supreme court, scientists would be appointed to assess the validity of laws, which would be required to cite research. If such research weren't available metrics and studies to gauge the impact of laws would need to be outlined, along with the necessary funding to implement them. An important detail is that this is not necessarily an all or nothing judgement in many cases, as proper risk management requires partial measures be taken in the face of insufficient evidence in order to serve as a hedge. This could be quite far-reaching. For example, a farm bill could be ruled unscientific because it does nothing to address climate change impacts. Inaction is not a way to circumvent reality.

Of course having a small number of experts opens itself up to the same personal failings the Supreme Court suffers from. A better method would be to utilize all those qualified in the relevant fields, possibly weighted by expertise, to provide a broader base. An approach the Supreme Court itself could benefit from and one that was at the core of my previous posting.

Saturday, November 8, 2014

Effective Immortality (Your Results May Vary)

Consider a persistence of personality thought experiment:

You are traveling anonymously (no one knows you are there) in a foreign country when you suddenly acquire an incredibly convenient form of amnesia that wipes out all specific memories other than those forming your beliefs and abilities. Are you still you?

For me, the answer is much more "yes" than "no". When confronted with the same basic high level situations I will tend to react in the same manner which is what matters to me. Put another way, I can envision large changes in my personal details that wouldn't effect my core responses all that much (I can also identify trivial ones that would utterly alter them, but that's besides the point).

This leads me to the conclusion that a clone of myself raised with a rationalist world view and a technical background (clearly there is some genetic propensity given the number of engineers in my family) would effectively be me in all the ways I truly value. The idiosyncratic personal details just don't matter to me very much in the grand scheme of things.

This attitude factors in directly to my decision not to reproduce. I did not expect a miniature version of myself by any means, but I was not prepared to take the chance that my progeny would not be highly intelligent. That being an utterly unfair and irresponsible opinion for a parent (in our society at least) I chose not to become one.

Saturday, October 5, 2013

Scientific Consensus Democracy

97 out of 100 climate experts agree that humans are responsible for
global warming, so why is the ratio of politicians who do so vastly less?
For rather obvious reasons, the government shutdown has me considering representative democracy. I have previously blogged about better compromise methods to avoid gridlock, but that is a minor modification to the existing paradigm rather than a complete redesign. There are three things a democracy needs to determine as far as policy: what our society's priorities are, what specific policies to pursue to attain those, and who should implement them. To limit the scope of this post, I shall only deal with the first two, leaving the third for further entries. I also will not delve into details too deeply, not only for the sake of brevity but because a single person would be hard pressed to work out every aspect of such an enormous undertaking.

I shall start with the second requirement as it sheds light on the first. More and more frequently there have been distortions of the state of scientific evidence by our political figures. This is unacceptable on questions of fact, which is why I propose a model based on scientific consensus. Citizens can become certified in various policy areas by passing an exam. They could then rate the feasibility of the aspects they are qualified in for the various policy options. This is a key point: they cannot weigh in on parts they are not verified to be knowledgeable about. The experts could cede their evaluation to another qualified individual, but should be able to do so on an issue by issue basis. The best option would be the one officially pursued. As transition costs would be included, revision of the assessments could occur continuously and only if a new policy ranked significantly better than the current one would an actual policy shift occur. Anyone could suggest a policy and as they garner more feasible appraisals, they would rise in prominence, similar to other online peer evaluated systems.

That brief summary raises many issues. Tests would need to be transparent and provided free of charge. Our education system would of course need to be improved, but that is a good notion in any event. To increase inclusiveness, partial accreditation should be possible, with a partial vote being awarded although not necessarily in a strictly linear fashion. This would also be a good way to handle out of date expertise: it would slowly lose influence over time, requiring periodic re-certification. The frequency of such exams is a detail beyond the scope of this post, but would obviously be related to the rate of advances in the field.

The voting system would have to be online for practical reasons, so security would be crucial. Each expert should be provided with a review period to approve any votes they have ceded to others before they lead to policy changes, with automated trend detection systems possibly provided to call attention to outcomes they likely wouldn't agree with. Additionally, measures of uncertainty in ratings could also be incorporated to accurately reflect incomplete information and also to highlight where more research is required.

With that as a policy selection method, the nation's priorities must also be considered. This is where the role of general opinion can legitimately factor in. I envision something more along the lines of a periodic census, rather than issue by issue involvement, where citizens can choose what matters to them. To force people to make trade offs rather than emphasizing everything, they would have a set number of points they must distribute between various competing priorities. These measures would be the things the expert evaluation bottom line would be phrased in terms of, along with an assessment of the existing state of those goals, as a high priority that has already been well satisfied may wind up with less influence than a lower one that is as yet poorly attained. Most likely, automated methods of striking the necessary compromises would be the most efficient way to identify the best set of policy solutions.

In the name of conciseness, I have deliberately omitted details and no doubt failed to consider many more. My larger point is that there are other ways to approach governance in a democratic society; ones that leverage newer technology and the mechanisms of our most successful institutions.

Saturday, March 23, 2013

Skin in the Game

Picking just one quote for this post was not easy!
The post hoc rationalizations accompanying the tenth anniversary of the invasion of Iraq got me thinking about ways to discourage conflict except in the most serious instances. Given the necessary (for classified source reasons) information gulf between our leaders and the general public, any such disincentive must occur at that level. I propose the following: a chance of death. Once the conflict is over, those in our government that voted for war should have to go through a lottery in which they may be picked for execution. The likelihood should be linked to the costs of the war (both human and financial) and perhaps any disparity on that score from what the public was told to anticipate. Obviously, such metrics would need to be very careful considered, and the exact details are beyond the scope of this post, which assumes for the sake of discussion that such a formula is possible. This measure would have a chilling effect on entering into war lightly and discourage optimistic portrayals of the sacrifice being considered. It seems entirely reasonable to expect those asking others to potentially lay down down their lives to risk their own.

Thursday, February 28, 2013

A Ticking Time Bomb is Not Justification

I recently saw new interview clips of Dick Cheney and he once again tried to justify torture with the “ticking time bomb” scenario: when you know a suspect has information that can prevent an imminent, serious threat. The theory is that this possibility is so compelling that we should allow torture to be legally permissible. My view of this line of reasoning is that if the circumstances are so dire, so terrible, that the person charged with extracting the information shouldn't care that torture is outlawed and that they will face legal consequences. If they still have concerns about such matters, then the event they're trying to prevent isn't sufficient justification. We need the illegality to act as a barrier to such behavior, especially since torture's effectiveness is not without doubt and we would increase the likelihood of our captured troops facing similar mistreatment. The argument is to prevent the unthinkable by allowing the unthinkable, and the unthinkable should never be sanctioned by law.

Sunday, January 13, 2013

Well Regulated Militia

Flawed though the test may be, the takeaway is valid.


I have previously blogged about crowd shootings and even touched on the topic of this post, but feel it warrants expansion. While I am choosing for the purposes of brevity to ignore increased regulation of access and storage of firearms and also improving our mental health system, I wish to make it clear that this is not because I discount them. When tackling complex issues, one should consider all the available tools. What I am choosing to focus upon is the frequently mentioned notion that having more armed individuals will somehow reduce deaths from gun violence overall. This concept is usually accepted uncritically by gun advocates and rejected as unrealistic (based on statistics showing a small number of justifiable shootings relative to accidental ones) by proponents of greater gun control; what I wish to consider is what it would take for this to actually be a credible notion.

What seems to be frequently absent from such conversations is proper emphasis on the need for training; just carrying a weapon will not make you a competent combatant. The ability to keep calm and accurately engage an armed opponent, let alone several, is extremely demanding, necessitating a high degree of training. That is true even in a home defense scenario, and if one considers the context of public defense, the difficulty is magnified many times. Your opponent may well have superior firepower (and other equipment) to what you can reasonably have with you all the time and the safety of bystanders is vital. Indeed, there will be some situations where no responsible armed option by a lone individual is possible for fear of injuring others. Gun owners who seriously intend to act in such a capacity need to have training on a par with police and it needs to be ongoing since such skills are perishable. This would be a significant commitment, which would seem at odds with the number of individuals required so that their presence would significantly add response capacity; if they can’t respond appreciably quicker than the police, it makes little difference. So we'd need a large number of highly trained, certified, armed citizens for it to make much of an impact. This sounds very much like the "well regulated militia" the second amendment references (save for the for the fact that they would usually be acting alone) and would represent a great departure from the status quo of often casual gun ownership. One could argue that dedicated security guards might fulfill this role, but that gets away from the idea of armed citizenry and thus is beyond the scope of this post.

It would be easy to interpret my observations as an outright dismissal of the entire prospect, but that is not the case: while all options should be considered, there is no value in unrealistic schemes. The ratio of good actors to bad would have to be very high for it to be a significant net positive, which makes improvements in the other areas I mentioned in my introduction vital for any such possibility, as it takes virtually no training to attack unarmed civilians and accidents do happen. Human lives are not like a Wikipedia article: you cannot simply undo the damage of a malicious few with a well intentioned many (especially if they’re not qualified), and even Wikipedia must impose barriers to access from time to time.

Sunday, December 2, 2012

Ills of the Nuclear Family


The nuclear family is a very recent development and a comparatively isolated model of child rearing. The rise in single parent households, lessened extended family involvement, and increased mobility have compounded this deficiency, leading to some unfortunate consequences. This post examines the shortcomings of the stereotypical nuclear family paradigm, as that is overwhelmingly what is held up as desirable in this country. Note I am not asserting that family structure alone is responsible for the following issues, merely that it is often a contributing factor. Of course in practice there are an almost unlimited number of variations on the nuclear family ideal, but delving into them all would be a massive undertaking. Similarly, I will not attempt to suggest possible remedies, nor explore how existing structures might, and often do, mitigate some of the limitations. Those will be possible subjects for the future.

  • Crowded childbearing window: The best years for bearing children are also critical ones for obtaining an education and beginning a career. The demanding nuclear family system makes balancing these three difficult and has contributed to the drastic increase in maternal age and associated health problems, as well as declining birth rates amongst the better educated. The fact that having a child in your teens is such a huge barrier to success when biologically it is an entirely suitable time frame highlights the current disconnect.
     
  • Identity loss:
    At the moment it can be hard, especially for women, to both have a family and achieve. Many adults are forced to abandon much of their own pursuits and ability to make new social contacts, which can result in feelings of isolation and frustration. Even when their children are grown, the effects are still felt, such as the popularized "empty nest syndrome" and lost opportunities generally. By sharing the burden of child rearing, these issues could be lessened.
     
  • Little diversification of parental failings: Children are exposed a great deal to the flaws of their parents. Greater care redundancy softens the impact of personality disorders, income inequality, and more reliably brings neglect or abuse to light. This is the same principle behind the common practice of diversification when investing, so that a single bad stock won't ruin your portfolio.
     
  • Personality conflicts:
    Our media drives us to expect an almost magical closeness with our progeny. If that fails to materialize, disappointment and parental attempts to foist unreasonable expectations on their children can occur. Significant exposure to a greater number of adults would allow them more opportunity to find those they have things in common with. It would also help diminish the practice of infantilizing young adults.
  • Reduced stability:

    The narrow scope of the modern family unit exaggerates issues surrounding parental rights, divorce (or marriage for that matter), death of a parent, or adoption. A broader base would ensure greater continuity, softening the impacts of these disruptions.
  • Inefficiency:
    The nuclear family often neglects gains that can be made through economies of scale, increasing use of resources and labor. Housing costs, electricity, transportation, child care, cleaning, maintenance, food (both purchasing cost and effort of preparation), and clothing are some areas affected, with young parents being the most impacted. Health care also bears mentioning, as our employer based models currently focus on the nuclear family as their basic unit of coverage. Some of this waste is also impacted by our cultural trend towards sole ownership of items even if they are infrequently used; a practice reinforced by the isolated model the nuclear family ideal espouses.
     
  • Marginalization of the elderly: The mobile nature of individuals today can often result in parents being isolated later in life if their children relocate. A less punishing, more interconnected system would allow more personal development during parenthood as well as providing more opportunities for cross generational bonds (not necessarily just with relatives) to be made. The simultaneous problems of marginalized seniors and a lack of affordable day care are especially telling.

Sunday, November 18, 2012

A Measure of Success

538's estimate of state probabilities.
Came up with a way to measure 538's success in predicting the election other than a simple boolean for each state. All predictions are weighted by their certainty. A 50/50 estimate would thus count for nothing (it's worth pointing out that those calling the election a tossup were doing this...), whereas a prediction of 100% would yield a weight of 1.0. Correct predictions are positive, incorrect negative (538 got all 50 states right, plus DC). This weight is then multiplied by a state's electoral votes, summed, and then normalized on the range of the worst score (100% certainty on all, but getting them all wrong: -538) to the best (100% certainty, all correct: 538). Doing this gives 538 a score of .9485. Note that I treated states that award their electoral votes in a split manner as all or nothing for simplicity.

Baseline estimate of state probabilities
To provide more insight, a naive model was used to provide a baseline: all swing states estimated at 50/50 and all non swing states at 100%. This gives a score of 0.8977. This means that 538's estimate gets us 49.66% of the way from the baseline to a perfect forecast. Unfortunately, the data necessary to apply this measure to other forecasts wasn't easily available (and by "easily", I mean "with the amount of effort I was willing to put into this blog post"), but I grabbed some from a couple other sites.

PEC's last estimate map before the election.
Princeton Election Consortium's last posted data was in the form of a color map, so I extracted their estimates by examining the colors and they were clearly not that precise. Applying my measure, they got a score of 0.9429, and were 44.19% of the way from the baseline to a perfect forecast.

Simon Jackman's swing state estimate data.
Simon Jackman over at The Huffington Post had this estimate of swing states probabilities the day before the election. I couldn't find the non swing state opens, so filled in 100% for them, which almost certainly inflated his overall score. It worked out to 0.9544, and was 55.38% of the way from the baseline to perfect.

If anyone has other data (or better versions of the data I did use) I'd be happy to include it in this post. Here is the spreadsheet I used to compute the scores.

Bonus: I edited these maps that have been making the rounds since the election that scale state size according to electoral votes and population respectively to have purple shading that reflects the percentage of Obama and Romney votes. They don't really warrant a post of their own, especially since I didn't create the original maps they're based on.

Size proportional to electoral votes (1 square = 1 vote)
Size proportional to population (original image didn't include HI or AK)



Wednesday, November 7, 2012

DIY Shingle Lift

The lift is carefully aligned at an angle of that-looks-about-right degrees.

Needed to get 4,200 lbs of shingles onto my friend's roof and the nearest rental place with a power ladder was very far, it was expensive, and we lacked any way to transport it (and an appropriate ladder). For $30 of lumber and hardware I turned my electric hoist and some scrap wood into a lift. It took one hour to design and two to build. This is less time than renting would've taken, given the two separate round trips required.

I used two sixteen foot 2x6s for the rails, and ordered an additional twelve foot piece (for the carriage and key pieces I didn't trust making out of the scrap available), all of which were delivered along with the shingles from the hardware store. When the project is done, they'll be cut in half and saved for future projects. We already had decking screws to assemble it, so all that was needed was a pulley (a 640 lb one is far more than the 75 lb bundles require but it was inexpensive and I can use it for future hoist projects), a u-bolt to affix the cable, and a cable stop.

My hoist can handle 440 lbs in single line mode and goes reasonably fast. I mounted it on the bottom so that the controls would be accessible to the operator on the ground without having to lengthen the control cable, so its weight would be at ground level, and also so the extra cable length and flexibility of the pulley at the top would even out the slight position shifts as the cable is unwound from different spots on the drum. The width of the lift was determined by the need to put the spool in the center and also provide enough room for ventilation. One way to reduce it would've been to mount the hoist sideways. To protect it from rain, a contractor duty garbage bag was placed over it and was weighted down. A hurricane came through before we were done, so the bag was additionally stapled to the frame. It didn't suffer any water damage, despite the need to work the bag around the cable (this junction was taped).

I designed the carriage to go low enough that the shingles could be easily loaded and also so that the bundles go over the top of the lift, sliding off right into your arms. The lift is set at a height above the roof to make this convenient. The platform is a piece of scrap from a demolished deck, which is why it has that odd angle. I made the carriage 1/4" larger than the frame to account for small variations in width. By simply resting a board at the back, the lift can also raise tools or other materials. Large marks drawn in sharpie allow the operator on the ground to know when to stop. Ideally a limit switch would be used, but it wasn't worth it for a device with such a short lifespan. Similarly, furniture gliders were considered but then rejected as the hoist is sufficiently powerful to overcome friction.



Small legs ensure the carriage doesn't collide with the edge of the roof and also increase the contact area versus simply leaning it. It is surprisingly easy to lift and move, should I need to.



The lift worked very well, although we did run up against the hoist's 25% duty cycle rating, especially as the cooling fan is attached to its motor and does not run unless it does. I plan to add a continuously running fan from an old power supply, but not having one on hand, I rigged a vacuum cleaner blowing air to increase its cooling.

Saturday, October 20, 2012

Programming Languages Should be Unnatural


I have always thought that the notion of making programming languages more like natural ones is absurd. Without a truly formidable amount of artificial intelligence, practical natural-like programming languages amount to eschewing symbols for words, which more adds inconvenience than clarity (I'm looking at you, Applescript). If anything, I feel natural languages should be made more like artificial ones than the other way around. In very technical disciplines, this is actually done after a fashion: they adopt a precise specialized jargon to help combat the inherent ambiguities and limited vocabularies of context dependent natural language.

Friday, September 14, 2012

Structured Compromise System


If one or more parties in a negotiation is unwilling to compromise, many decision making processes can grind to a halt; the current state of American politics being a prime example. What's needed is a framework that forces uncooperative sides to come to an agreement, while according each their appropriate level of influence. Rather than a series of small votes with an implied give and take linking them, all points should be considered explicitly together. Each party could propose actions to be taken, and would receive points based on their degree of representation which could spend to support or oppose each of these proposals. This process could undergo many iterations as each side determines where best to spend their influence. There would be no partial implementations, instead new, toned down versions of previous suggestions would be added to achieve this effect. For directly contradictory items only the highest scoring of them would go into effect, so the parties would have to split unrelated items up, rather than presenting them as a single action item. A limit to the number of proposals may be required, however, to prevent either side for complicating things intentionally.

By eliminating a series of small gated decisions, the ability to impede action is greatly reduced. Even if closely matched in point totals, differing priorities will make deadlock unlikely in most instances. The final point allocations would provide clear record of the weight each party places on various areas and help voters more easily see if they are being properly represented. This technique could equally be applied to compromises other than in politics and theoretically to any number of participants. Which chores each family member should do around the house, for example.

To prevent attempts to game the system by endlessly changing their point allocations, a tax on changes could be employed. In its most basic form, all changes would be penalized, but a more sophisticated one would allow all parties to alter their positions if they all agreed to do so. If only some wished to change, then the tax could be applied only to those parties. To clarify, in that case everyone would get to reallocate but those not asking for another round wouldn't be penalized. This would discourage huge changes but still allows the participants to come to an accommodation of each others priorities.

This system grew out of one I came up with to help with a software maker's survey problems. They would ask users on their website what new features they would like to see in future versions so they could gauge where to focus their development effort. Somewhat predictably, the users all voted for all the good features, providing little useful direction. I suggested each user get points they could allocate to various features, with the amount received varying by things such as forum participation, useful bug reporting, or number of copies of the software purchased. For reasons I've never quite understood, they weren't interested in the notion.

Thursday, June 7, 2012

How bad can the electoral college be?



I was curious as to how poorly the electoral college could reflect the popular vote, so I whipped up a program to compute it (assuming only two parties). Below you see the map as it would need to be for the red candidate to win (yeah, yeah, I notice I'm showing the Republican candidate as winning unfairly, but that was just random). The best (or should that be worst?) answer involved all the districts of Nebraska and Maine voting together so I didn't need to split their electoral votes in the picture, although the program did explore those options. While there are some odd states voting together, for expediency I'll ignore that detail since it's not impossible demographic or political shifts could lead to such a result. In order to carry the red states with just over 50% in each, only 21.61% of the popular vote would have to be obtained (21.3% if you count the US territories which currently aren't represented in the electoral college). You can see many of the least populated states are chosen, maximizing the error. I knew the electoral college was bad, but this is astounding! To have a candidate defeated who had more than 78% of the vote would simply be unacceptable.


To force a tie you need just a slightly smaller percentage of popular vote: 21.52% (or 21.2% if counting the US territories).


Next I took the election results for each of the past three presidential races (the last three all used 2000 census data) and changed my program to look for the minimum subset of the electoral votes actually carried by the winner that still ensured victory. All three required less than 25% of the popular vote for victory (the 2004 minimum solution required 271 electoral votes, not just 270). We can also approach the inherent unfairness of the electoral college another way, by assuming each state's electoral votes per percentage of population is the same. This makes all states more equal, but the very "all or nothing" nature of the system would still allow victory with (270 / 538) * .5 = 25.09% of the popular vote, since you only need just barely over 50% of each state to win all its electoral votes. For this last calculation I ignored the proportional vote system of Maine and Nebraska for simplicity. The rough bottom line is this: a candidate for president can win 75% of the popular vote and still lose the electoral college.

2008: winner is in blue
270 electoral votes
24.0566% of popular vote actually required (includes US territories)


2004: winner is in red
271 electoral votes
24.33% of popular vote actually required (includes US territories)


2000: winner is in red
270 electoral votes
24.575% of popular vote actually required (includes US territories)

Sunday, May 13, 2012

There is No Sharp Transition to Personhood



The belief that fertilized eggs qualify as people reflects an intuition that there really is no clear dividing line for the concept of personhood. It is much like the paradox of the heap: if one grain at a time is removed from a pile of sand, at what point does it stop being a pile? There are many ways to approach this question, but I favor the fuzzy logic answer: that even one grain of sand has some degree of "pileness" to it, two grains have twice as much, three grains three times as much, etc. The salient point for this post is that there is no sudden transition, something many people have difficulty contemplating. So, is an egg a person? "Somewhat", while unsatisfying as an answer, is nonetheless accurate. A fertilized egg? More so. A baby? Yet more. People attempt to use the ability to survive outside a mother's body as a sharp transition, but even that is artificial: an infant is still utterly dependent on a large input of care and resources, let alone one born right at the edge of viability. More still is required for them to attain anything approaching their full potential.

The problem with extending the notion of legal personhood to ever earlier stages of human development is that it is an argument from potential. Every egg, every sperm has the potential to become a person, but a policy aimed at bringing as many to term as possible would quickly lay waste to the planet. In a world of limited resources, potential life has a finite worth, which by definition means it cannot automatically trump all other concerns. This seemingly heartless conclusion is part of our own biology: many miscarriages are induced by defects that would not be immediately fatal to the fetus but would yield young so weak as to not be worth the associated opportunity cost or risk to the mother. Exactly what the value of a potential person may be is not only beyond the scope of this post, but is arguably impossible to answer definitively owing to the subjective and deeply personal nature of the many competing factors. My goal is instead to dispel the absolute thinking that seems to frequently accompany such issues. Life may be precious, but, on this planet at least, it is far from priceless.

Sunday, May 6, 2012

The Failure of U.S. Representative Democracy


There are many bemoaning political gridlock in America, questioning how we will overcome the current dysfunction in order to deal with pressing issues. While I could not agree more, my favorite (if that is the appropriate word) example of the failure of representative democracy in this country comes from an earlier decade, illustrating that there have long been serious problems. It is this: not adopting the metric system. You may wonder why I choose such a seemingly minor event, but while other disasters such as the War on Drugs are loaded with complicated issues, the conversion to the metric system was completely unambiguous in its long term benefit relative to the modest costs involved. The fact that other nations successfully switched around the same time makes our attempt appear all the more feeble. It was a moment for our leaders to lead, rather than to cave to irrational discontent. Things have not improved since, unfortunately, and it leaves me pessimistic at a deeper level than even the Citizens United ruling.

Tuesday, April 24, 2012

Attractiveness Components


I was pondering the different ways people can be viewed as attractive, and decided to systematize it. I am considering only static images of a neutral appearance, so no personality cues are incoporated. Using evolutionary pressures as a guide, I came up with the following four:
  • desire to protect children
  • desire for general health
  • desire for the gender we are sexually attracted to
  • desire for peak fecundity
The first component evinces itself in what I refer to as "cuteness". It can sometimes be at odds with the last measure, depending on the one making the judgement, giving rise to the "cute but not sexy" phenomena. The second is the typical beauty measures that we hear so much about: symmetry, proportion, height, clear skin, etc. The third is simply how much your appearance contains gender cues. Note that how this is interpreted by the viewer is what matters, so it says nothing of your absolute attractiveness in its own right. A good example of someone who varies from their stereotypical gender appearance but is still very attractive would be Katharine Hepburn, who has often been referred to as a "handsome" woman. The last is straightforward, mostly being about not being too old or too young, although it can overlap the second to some extent depending on how you frame the two.

The way this could be used in practice would be to have users rate many images whose components have been determined for overall attractiveness. Standard recommendation system techniques could then be used to infer the value each user places on each component, and should then allow estimating the rating of new photos once they've had their components rated. You could omit the requirement that images have their components directly determined, inferring them as well, but that would result in far slower convergence for newly added images.