One More Step

“Fitting the curve” is when you take the results of an exam and change them to fit what is “normal”. Through squashing, stretching, and standardising unexpected shapes. If a group has an average, almost no one is average. A Standard Deviation is the average amount each result is different from average. Again, almost no one is the average amount different from average. “Normally”, about 68% of results sit between one Standard Deviation below and one Standard Deviation above the average. 95% sit between two below, and two above. 99.7% between three down, three up. That we do this for exams results shows that the goal is not education. It is sorting. Most education is designed to find your role in the world. This is the exact opposite of what we should be doing, and limits everyone to their starting point. Like Samwise Gamgee, the step you need to take is to realise the story is not about you. Education should be about understanding, and solving, the problems in the world. And financing the things that aren’t problems, but need money. At some point, you need to take the vital step that snaps the connection to the constraints you don’t find helpful.

Normal Results

Cause and Effect

I am distrustful of cause and effect. I think nature is too. That is partly why our ability to disagree and interpret things differently is so powerful. There is consciousness in the sense of awareness of what is going on now. Then, there is consciousness in the sense of connecting those nows in a meaningful way.

Two big obstacles to those stories, are Spurious Correlations and Confounding Variables.

Spurious is like a combination of spew, furious and hilarious. It is the fake news of statistics. Two things appear connected or correlated. That is not because they are. It is because we are pattern seekers in an incredibly random world. Two things can dance together just because. We will add meaning even though it is not there.

Tyler Vigen collects these kind of amusing dances that point out that 'correlation does not equal causation'. The number of people who drowned by falling into a pool, is correlated with the number of films Nicolas Cage appeared in. 

A Confounding Variable is something that confuses everything. It may seem that we have figured out the cause of something, but there are other things involved. The 'outside influence' affects both the cause and effect. It is incredibly difficult to isolate something so that nothing else interferes. 

Food is a great example. Anyone wanting to eat healthily will get millions of competing theories. It is very hard to stop eating everything else, eat just a single ingredient, then observe its effect. Then repeat with each combination of two, etc.. Each of us is different. Instead, we come up with sweeping 'cause and effect' stories that we generalise to others because it worked for us. We think. Till it doesn't. Then sorry. The biggest Confounding Variable is the story we bring with us. The prejudice we already have that needs to be unwound to see clearly.

You can be distrustful of cause and effect, and still use the stories you learn. Just add more than a sprinkle of humility. The world is complicated and random. Everything is a pattern if you want it to be. We can thank our imaginations for that.

Being Right

Predicting something correctly only gives the illusion of 'being right'. The world is far more uncertain than that. Nate Silver of fivethirtyeight.com gives roughly equal chances to Hillary Clinton (a) Winning Solidly, (b) Winning by a large margin, and (c) a close call where Trump could win. This means he thinks Clinton is a strong favourite, but he is far more nervous than most people thinking Clinton being favourite means she will win. Daniel Kahneman points out in 'Thinking Fast and Slow' that even statisticians aren't intuitive statisticians. We are far too willing to believe things without sufficient evidence. We overestimate our control over the world. We underestimate randomness. We don't think in 'alternate possible paths'. Even if Clinton wins, I feel very uncomfortable that he got this close. I think the world will go on either way, but Brexit and the US election make it very clear that we have some soul searching to do.

Lottery v Election

I am busy reading 'The Democracy Project' by David Graeber. One thing it turns out I have been completely wrong on is thinking the key to Democracy is elections. The vesting of power in a representative. It turns out that is more of a core concept behind Republicanism. I have always had a bit of an issue with elections for two reasons. The first is that I have never felt completely competent in putting my cross in the box. I have always felt like I am about to write an exam I haven't studied hard enough for. I haven't done my homework. The second problem is that I don't actually think having done your homework helps, it just provides the illusion you have voted with your head rather than your prejudice.

When it came to the British Elections earlier this year, it would have been my first opportunity to vote as a British Citizen. Two problems. First, I was down under pet-sitting, and second, my vote didn't matter at all. Putney, where I had lived, was safely blue. I diligently read the manifestos anyway and came away feeling even more despondent. Even if I did my homework, what was I looking for? The ideology that best represented my inherent beliefs? I have never met someone who represents my beliefs. I don't even always represent my beliefs or best interest. The writer who seems to say the least I disagree with is Steven Pinker, and that makes me feel very uncomfortable. Fortunately, another writer whose work (rather than his manner) I enjoy, Nassim Taleb, thinks Pinker is very wrong on something pretty fundamental. So the conflict leaves sufficient doubt to keep me happy.

Graeber makes interesting points regarding of the usefulness of lottery rather than elections. What you want is people who represent the general view, not the people who are best at convincing people that they represent the general view. Then you just get the dog show that is reality TV electioneering. It reminded me of a lesson in one of my early statistics classes at university.

We often don't tell the truth because we don't want people to judge us poorly. We are worried that if they know what we are thinking, they will walk away. So we couch what we say with what we think they want to hear. This is a problem for people looking for the truth. Statistics Class taught a neat trick using the 'Law of Large Numbers'. Everyone in the class took a coin out of their pocket and flicked it. If it was heads, they answered the question 'Have you ever used marijuana?'. If it was tails, they answered 'Are you born in the first half of the year?'.

The second question should be something very neutral that has an answer that splits the group in a known way. So if roughly half the people are born in the first half of the year, you can assume that roughly half of the people will be answering that question. If the group is big enough, you can get an answer that is close to the truth since the person answering the question knows that their identity has been stripped. This is similar to my quest to get people to talk about the things that really matter to them. One of the reasons we don't open up is we are scared of people's responses. We want our identity stripped from our stories. We only tell the stuff that is really worrying us to people we feel secure will stick around.

Instead of elections, we could hold a lottery. Only two things matter, a agreed basic level of competence and a willingness to be selected. If the group is big enough, just like a survey that is big enough... it will represent the people.  It would also save a lot of bother in terms of feeling incompetent and faking that your vote is more considered than it is. We could also do away with party politics, and lobbying, and instead build a bigger tribe.

Risk

The study of risk is the study of decision and indecision. The range of possible qualitative and quantitative consequences of change. 'Decision Theory', an area focused on by Daniel Kahneman, looks at the psychology, economics, statistics, philosophy, sociology and history of choice. I believe happiness is closely related to choice. A feeling of empowerment and acceptance of the random beauty of the world. I don't think the world is malicious. It is the experimental trial and error that creates this beauty. Our appreciation and consciousness creates meaning. A static world would protect the way things are, but remove the ability to dream.

See:

'Thinking Fast and Slow' by Daniel Kahneman

'Fooled by Randomness' by Nassim Taleb

'Black Swan' by Nassim Taleb

'Risk' by John Adams
'The Signal and the Noise' Nate Silver