Mathematics of life

[kamakura]

Brilliant pupil's 'logical' suicide

A BRILLIANT schoolboy shot himself in the head after carefully calculating the benefits of life and deciding it was not worth living, an inquest was told yesterday.

www.independent.co.uk

You summarise the major factors of lives (health, goal, career, family, friendship, etc) into a few stochastic factor equations (x1, x2, x3...).
Assume equal-weighting and independence, so you simply multiply them into a complicated life equation in f(x1, x2...).
Each factor equation is exposed to randomness, so you can run a Monte Carlo simulation for each factor.
You do that for the entire life equation.

The worth of life is the Reimann-sum of the life equation.
If it becomes negative (life equation below x-axis), it is better to end the life to bring the worth back to 0.

 

[NotMeant2B]

I'm curious, there isn't any mathematics referenced in the article, where did you retrieve them from? Or are they of your own?

 

[HadEnough1974]

I'm curious, there isn't any mathematics referenced in the article, where did you retrieve them from? Or are they of your own?

I'm curious about that as well.

 

[kamakura]

It is from me. I am reasonably knowledgeable in terms of mathematics and statistical modelling.

1. You identify major factors to capture your life worth.
2. Represent each major as a stochastic differential equation (F1, F2, F3...Fi), each with random factors embedded, so they should look like a Brownian motion
3. Use your current state as initial condition for each equation.
4. Assume independence and equity-weighting, so you can obtain a Life function L(F1, F2, F3...Fi) by multiplying all F.
5. Run monte carlo simulation for each Fi, so you obtain a simulated run of life function L.
6. Integrate L for the Reimann sum for the life worth of that simulation
7. Simulate it many times to obtain the distribution of your life worth.
8. Look at how positive/negative the Reimann sums are, to decide if you should neutralise it by voluntary methods.

If your life is already negative, you should either end it or increase the volatility of life first if such means available.

 

[HadEnough1974]

It is from me. I am reasonably knowledgeable in terms of mathematics and statistical modelling.

1. You identify major factors to capture your life worth.
2. Represent each major as a stochastic differential equation (F1, F2, F3...Fi), each with random factors embedded, so they should look like a Brownian motion
3. Use your current state as initial condition for each equation.
4. Assume independence and equity-weighting, so you can obtain a Life function L(F1, F2, F3...Fi) by multiplying all F.
5. Run monte carlo simulation for each Fi, so you obtain a simulated run of life function L.
6. Integrate L for the Reimann sum for the life worth of that simulation
7. Simulate it many times to obtain the distribution of your life worth.
8. Look at how positive/negative the Reimann sums are, to decide if you should neutralise it by voluntary methods.

If your life is already negative, you should either end it or increase the volatility of life first if such means available.

That's so interesting. I'll send you a pm.

 

[kamakura]

If your life is already substantially negative and you are fine with neutralization, it implies that you have no more downside, so you can try to increase the worth of life by first taking big risks. If that doesn't work out, you just neutralise life. Kinda like if you are to die anyways, place a big roulette bet first, there is nothing to lose.

 

[Soul]

@kamakura, can you show us an example of how you'd calculate the value of someone's life? I'll volunteer as a subject if you want!

 

[Deleted member 1465]

It is from me. I am reasonably knowledgeable in terms of mathematics and statistical modelling.

1. You identify major factors to capture your life worth.
2. Represent each major as a stochastic differential equation (F1, F2, F3...Fi), each with random factors embedded, so they should look like a Brownian motion
3. Use your current state as initial condition for each equation.
4. Assume independence and equity-weighting, so you can obtain a Life function L(F1, F2, F3...Fi) by multiplying all F.
5. Run monte carlo simulation for each Fi, so you obtain a simulated run of life function L.
6. Integrate L for the Reimann sum for the life worth of that simulation
7. Simulate it many times to obtain the distribution of your life worth.
8. Look at how positive/negative the Reimann sums are, to decide if you should neutralise it by voluntary methods.

If your life is already negative, you should either end it or increase the volatility of life first if such means available.

Really interesting. Utterly beyond me and my ancient A level in maths and stats.

 

[kamakura]

you do it in matlab or octave, impossible by hand

 

[NotMeant2B]

It is from me. I am reasonably knowledgeable in terms of mathematics and statistical modelling.

1. You identify major factors to capture your life worth.
2. Represent each major as a stochastic differential equation (F1, F2, F3...Fi), each with random factors embedded, so they should look like a Brownian motion
3. Use your current state as initial condition for each equation.
4. Assume independence and equity-weighting, so you can obtain a Life function L(F1, F2, F3...Fi) by multiplying all F.
5. Run monte carlo simulation for each Fi, so you obtain a simulated run of life function L.
6. Integrate L for the Reimann sum for the life worth of that simulation
7. Simulate it many times to obtain the distribution of your life worth.
8. Look at how positive/negative the Reimann sums are, to decide if you should neutralise it by voluntary methods.

If your life is already negative, you should either end it or increase the volatility of life first if such means available.

Got it. I'm not specialized in the subjects involved, but I'm somehow familiarized with stochastic processes to understand what is going on. Saying this, I have some questions which I'd like to clear up, if you are willing to answer of course, as well as some things I'd like to add:

1. I think there are two ways to approach the first step in the methodology:
   1. Individual-based factor selection, in which every individual picks the factors which they consider most important in their life. This would provide more accurate results to every subject, but could be prone to variability depending on the current mood or the sensitivity of said subject.
   2. Population-based factor selection, in which we take psychological and sociological studies that compile the most relevant contributors to human happiness. This would mitigate the subjectivity of the previous approach, but it will also confine every test subject to the same values, and would be less meaningful to any particular subject.
2. Assuming every variable has an uniform distribution would not be interpreted as equality of opportunities for every test subject? and even so, how could we possibly determine the interval of said distribution?
3. It would be necessary to implement a metric in which the initial conditions can be determined from the current state (with respect to the factors) of the subject, even if universal (applicable for anyone) or individual (meaningful only to the current subject).
4. Is independence a sensible assumption for all Fi? Perhaps a more careful analysis would be adequate?

I think that it's reasonable to think of life as a chaotic system, which will not pose as much as a problem to the stochastic part of the study, but it will certainly affect the deterministic part (since by definition, a chaotic system is one that is highly sensitive to the initial conditions), so we can only draw results locally, with more certainty the closer we get to them. This would give the study a character that is more similar to forecasting of the weather or the stock market. Following these arguments, it leads me to conclude that the transitory nature of the study may provide unreliable results for long term decisions as it is suicide. And this is a boomer for me too, since that makes it sound like the old loathed statement "permanent solution to a temporary problem".

However, I want to add that mathematics are not necessary to make a decision of that kind. We as humans are sensitive agents of our environment, and we respond in accordance to its stimuli. If the environment rewards us, we are happy; if it punish us, we are sad. After reaching certain maturity level, we've gained enough awereness of the environment to know when we have control and how much we consent on " keep playing the game". For me, suicide will always be a human right, simply because it concerns our liberty to choose a dignified life or death.

I'll be looking forward to your response or any comment you wish to include. And I know this is not the platform to address this kind of discussions but I think it is interesting discussing nonetheless, even if in a shallow level.

 

[Thereisnothing]

I there an an easy way for maths dunces like myself to do this and find the score?

 

[TAW122]

I believe I've read this story somewhere many years ago. While it is sad and tragic for someone so young to go especially if they are physically healthy, I still respect their decision. At the end of the day, he was suffering enough and decided that 'life is NOT worth living' so he arrived at that conclusion. In a sense, I'd say that he is ahead of his age group.

 

[zherhk]

I don't think an algorithm is needed to weight pro and cons to realise if it's worth living or not.

 

[Thereisnothing]

I don't think an algorithm is needed to weight pro and cons to realise if it's worth living or not.

You are quite right, we know ourselves and what our lives are like, without Maths getting involved............but it is interesting nonetheless.

 

[zherhk]

Yes. At least this brought the fact that who make the decision to ctb, isn't because one day someone wakes up with that idea, but because there was a logical reasoning behind.

 

[kamakura]

1. Assuming every variable has an uniform distribution would not be interpreted as equality of opportunities for every test subject? and even so, how could we possibly determine the interval of said distribution?
2. Is independence a sensible assumption for all Fi? Perhaps a more careful analysis would be adequate?

Following these arguments, it leads me to conclude that the transitory nature of the study may provide unreliable results for long term decisions as it is suicide. And this is a boomer for me too, since that makes it sound like the old loathed statement "permanent solution to a temporary problem".

The random variables are not uniform, some say the Health factor can be a slowly declining Brownian motion with a downside jump variable modelled by as a Poisson variable. Career, Relationship, Wealth factors can be modelled more or less symmetrically using normal distributions (ie typical Brownian motion).

Independence is simply assumed to derive the Life function as a multiplication of all factor equations. Otherwise, you can't get one.

Every dynamic model is sensitive to initial state, so can be said to be transitory/temporary...but is the best you can get at each point in time without a crystal ball.

 

[mesohappy]

I there an an easy way for maths dunces like myself to do this and find the score?

Reasons to leave>reasons to stay=waiting at the bus stop...Im a math dunce too,dont feel bad.