Author Topic: Doing the math on why everyone needs a gun  (Read 4727 times)

Offline mountainmoma

  • Survival Demonstrator
  • *******
  • Posts: 4728
  • Karma: 223
  • suburban homesteader
Doing the math on why everyone needs a gun
« on: May 01, 2018, 03:24:36 PM »
https://medium.com/s/story/the-surprisingly-solid-mathematical-case-of-the-tin-foil-hat-gun-prepper-15fce7d10437

On the probability of civil disturbances compared to, say, 30 year flood probability in a flood plain, and given that, how owning a rifle is not crazy

Offline Alan Georges

  • Survival Demonstrator
  • *******
  • Posts: 4593
  • Karma: 210
  • Still trying to reason with hurricane season.
Re: Doing the math on why everyone needs a gun
« Reply #1 on: May 01, 2018, 08:02:28 PM »
Good stuff, MM.  As the author makes clear in the article, this applies to guns and so much more.  He even starts off with a flood probability example.  That's brilliant, because not only is it a non-controversial way to walk people through the ideas, it makes the article not just about guns but about preparedness in general.

Best article I've read all week.  Thanks MM.

Offline David in MN

  • Survival Demonstrator
  • *******
  • Posts: 2780
  • Karma: 213
Re: Doing the math on why everyone needs a gun
« Reply #2 on: May 01, 2018, 08:42:08 PM »
Even if I accepted the predictable math on war (I don't) Europe is past due for WWIII by about 50 years. Using these statistics Russia and Japan are crazy overdue. We could argue about the probability of a French/English war being so overdue it's past 100%. Especially if you include wars in the American Colonies.

Bad math is never helpful. We could go 10,000 years without another war on this country. And having my shotgun isn't preparing me for a future civil war if we have one. I don't know the future but basing all my decisions on n=2 is a statistical loser.

The Revolutionary War and the Civil War were not statistically repeatable events. Both had nothing to do with the other. The fact that they occurred a few decades apart isn't a pattern.

I advocate all people own a means to defend their property. But using statistics in a bad way doesn't help. I could tell you that in 1969 3 men walked on the moon so due to population growth in 2018 75 people will walk on the moon. I'm making up the numbers but the point is salient.

Offline Alan Georges

  • Survival Demonstrator
  • *******
  • Posts: 4593
  • Karma: 210
  • Still trying to reason with hurricane season.
Re: Doing the math on why everyone needs a gun
« Reply #3 on: May 01, 2018, 09:42:04 PM »
This is perfectly good math.  Yes, the Revolutionary War and Civil War had nothing to do with each other.  That is precisely the point – they were uncorrelated events.  There is no pattern.  From this standpoint it is also completely meaningless to talk about another French and English war being "overdue."  All these calculations can say is "based on prior events, here is the probability of something happening at least once in the next N years."  Reading any more into the conclusion, as if it somehow foreshadows some dire event, is foolish.  The author of that article never hints at such claims.  He merely lays out the math and applies it to the probability of something going badly wrong over the course of an average human lifetime.

This is how risk estimation is done.  Don't believe me, or even the article's author.  Go ask your insurance company.  Ask down at the local college's math department.  This is a very basic probability calculation for rare, uncorrelated events.

I could tell you that in 1969 3 men walked on the moon so due to population growth in 2018 75 people will walk on the moon. I'm making up the numbers but the point is salient.
No, it really isn't.  Linear extrapolation plays no part in the calculation being discussed.
« Last Edit: May 01, 2018, 09:55:03 PM by Alan Georges »

Offline David in MN

  • Survival Demonstrator
  • *******
  • Posts: 2780
  • Karma: 213
Re: Doing the math on why everyone needs a gun
« Reply #4 on: May 02, 2018, 06:48:54 AM »
Ugh. I have to do math and it's early.

He could have reported the data as the probable time between civil wars, fit it to the Gaussian, and reported a confidence interval (which would be uselessly large with n=2).

But, let's use his bad math. I'm going to take the liberty to expand his concept. I would include gun ownership as a defense in the Whiskey Rebellion, the Bonus March, and Jim Crow. I'm also going to propose that knowledge of a gun is a life saving skill during WWI, WWII, Korea, and Vietnam. So I've moved from 2 to 9. Let's reassess the maths.

I come up with an 88% chance (using the given formula) that knowing how to use a gun will save your life. That's odd. I doubt 88% of us have fired a gun in self preservation. Let's look further. I'm 37 (not old) so there is a.... 63% chance I have already defended myself with a gun.

And while it doesn't feel linear, it is. It assumes equal probability of every day since the first colony. That's just not true. Even if we had a war today there's a much higher lag between Civil War 2 and Civil War 1 than between Civil War 1 and the Revolution. The frequency is decreasing. It's a bad linear model even if it doesn't feel like it.

Maybe I'm being a jackass harping on the data (was my job for a while) but doing sloppy math just gives your opponent fuel.

Offline Smurf Hunter

  • Survival Veteran
  • ********
  • Posts: 7172
  • Karma: 334
Re: Doing the math on why everyone needs a gun
« Reply #5 on: May 02, 2018, 10:10:24 AM »
Maybe I'm being a jackass harping on the data (was my job for a while) but doing sloppy math just gives your opponent fuel.

Perhaps your mistake is attempting "better" math when it's not a math problem at all.
To me this is at least as much a logical fallacy, similar to "survivor bias".

e.g. I am 41 years old, have never lost a gun fight.  The hypothesis is my choice of self defense ammunition contributed to this outcome.
If we found a million people with my same conditions and results, the statistics could trick you into believing something that was not there.


Offline mountainmoma

  • Survival Demonstrator
  • *******
  • Posts: 4728
  • Karma: 223
  • suburban homesteader
Re: Doing the math on why everyone needs a gun
« Reply #6 on: May 02, 2018, 10:15:14 AM »
You have to do the self defense seperate from war, I think, but it is still quite large.

Lets take some arbitrary new york times article quoting the NRA as saying 2x10^6 times a year ( I believe this includes you have it and they run from the house), so 350x10^6 into 2x10^6 is 2/350, so something between 6 and 7 % each year. Using the math that comes out to be..... a large statistic over your lifetime. And, yet, as you say, many of us never will pull out let alone fire one at another.

Your question is why ?

I'm not as mathmatical as my kids, so here is how one of them saw this right away  - instead of flood plain and gun use, I was explaining how the societal status quo is not honest with young people about birth control statistics, because they will say it is, let's say between 95-98% effective, so a teen hears this and takes best case, 98% then thinks a 2% chance. I would have to be pretty unlucky to be in that 2 out of 100. So, I explain, that it is that percent each year, so 2-5 people this year, another 2-5 next year, etc..... so over 10 years, that would mean 20-50 people ( because I am too lazy to do his fancy math). Math teen saw right way that "hey mom, this wouldnt neccessarily be different people, could be a smaller group getting pregnant multiple times" Yep, this is true ( but you could be in that group...)

So, gun self defense can be that some people ( so some high risk jobs, where they live, bad neighbors, etc....) have multiple instances and others none. This is much more likely to be the case than the birth control example.

So, the fact that a majority of people have not pulled put a gun does not invalidate the statistics at all

we do these types of statistics on the probability of a super large earthquake, or Yellowstone blowing.  Oregon and California take these rising percentage risks very seriously, there are quite serious preps and plans in place, and yet, none of us has had this happen in our lifetime, and most likely wont. But, the risk is increasing the longer we go since the last large one. Do you think their statistics are invalid just because we have not personally experienced it ?
« Last Edit: May 02, 2018, 10:27:19 AM by mountainmoma »

Offline mountainmoma

  • Survival Demonstrator
  • *******
  • Posts: 4728
  • Karma: 223
  • suburban homesteader
Re: Doing the math on why everyone needs a gun
« Reply #7 on: May 02, 2018, 10:22:30 AM »
If this math is not valid, then why is it valid to use for flood plain calculations for insurance and not for other places?

Remember, this percent likelyhood for civil disturbance has so far hit certain areas much harder than here, we have so far not had alot of war on our home turf. SO, even without the fancy math, is that normal ? How long do societies generally go without one ( ie., what is typical stability) ?  Obviously, some societal organizing is more or less stable than others. If ours is thought to be stable, why ? Have the parts that have kept us stable been changing ? Rather like, has the river course and the weather changed to give us more or less flood risk ?
« Last Edit: May 02, 2018, 10:31:12 AM by mountainmoma »

Offline Smurf Hunter

  • Survival Veteran
  • ********
  • Posts: 7172
  • Karma: 334
Re: Doing the math on why everyone needs a gun
« Reply #8 on: May 02, 2018, 11:35:29 AM »
If this math is not valid, then why is it valid to use for flood plain calculations for insurance and not for other places?

What causes floods?  Storms with lots of precipitation.  I'm not a hydrologist, but I figure it generally comes to amount of rain over a duration of time.
That's a single, tidy metric.  cm/hour or whatever they might use.

Presumably weather stations can provide daily, monthly and annual rolling averages for such things.  There's tons of data, and it's pretty objective.
We could also track how often a given craps table rolls a 7.

That's the difference.  There's lots of plain old numerical raw data that feeds into the prediction of a "100 year flood".
He wasn't just looking back over news articles in the past century and counting the flood events.







Offline mountainmoma

  • Survival Demonstrator
  • *******
  • Posts: 4728
  • Karma: 223
  • suburban homesteader
Re: Doing the math on why everyone needs a gun
« Reply #9 on: May 02, 2018, 12:16:44 PM »
What causes floods?  Storms with lots of precipitation.  I'm not a hydrologist, but I figure it generally comes to amount of rain over a duration of time.
That's a single, tidy metric.  cm/hour or whatever they might use.

Presumably weather stations can provide daily, monthly and annual rolling averages for such things.  There's tons of data, and it's pretty objective.
We could also track how often a given craps table rolls a 7.

That's the difference.  There's lots of plain old numerical raw data that feeds into the prediction of a "100 year flood".
He wasn't just looking back over news articles in the past century and counting the flood events.

From what he says, they actually just use the data on how often it has happened. Statistics are like that, they have limitations. So, they predict risk using past data, but shifting weather patterns or global climate change can make it actually better or worse than the percentage risk they calculate.

We can use data to calculate overall probability on needing a gun, with the same caveats, some locations are riskier than others, trends can change so it is better or worse than overall past data shows, etc....

To me the important take away from statistics is the fact that a small percentage risk each year does add up to a much larger percentage risk over 10 years or a lifetime. So that someone looking t statistics and saying, there is a 2 percent chance, I am not prepping is short sighted as that is the annual percentage and over time, the overall likelyhood is much, much higher. (And, obviously, some locations and societal surroundings are riskier than others )

For me, I generally love the truism "there are lies, damb lies, and statistics " . However, if you are trying to hold your own with a family member or friend quoting low probability for certain bad events, as they re using annual percent risk ( aggregate risk), I think it could be helpful to show them how the math changes that risk over time. So, this math holds true to have a discussion with someone who is using a statistic, in other words, for the sake of the argument both parties are taking the annual risk as fact

Offline David in MN

  • Survival Demonstrator
  • *******
  • Posts: 2780
  • Karma: 213
Re: Doing the math on why everyone needs a gun
« Reply #10 on: May 02, 2018, 12:33:54 PM »
Well there's lots of data to assess insuring for flood. Are you near a body of water? on a floodplain? How is the house constructed? How often in the past has it happened?

We use the '100 year flood' metric because we have about 100 years of good data through the US. It's often wrongly treated as a worst case scenario. In reality it's just a comparative benchmark of the worst we'vve documented. It doesn't tell you what this storm will look like but it helps you know what you need to prepare for.

Predicting the odd of a war based solely on assuming randomized odds and two previous is not rigorous. Let me put this another way... The author has basically given us a roulette wheel. Each year we spin and if it hits red 32 war was declared. Taleb nicknamed this the ludic fallacy because real life is much more complex to predict and not everything is approximated by a roulette wheel. No statistician would ever make any claim using 2 data points.

Offline mountainmoma

  • Survival Demonstrator
  • *******
  • Posts: 4728
  • Karma: 223
  • suburban homesteader
Re: Doing the math on why everyone needs a gun
« Reply #11 on: May 02, 2018, 02:45:21 PM »
I dont think it is a claim on when a war will happen, I think it ia a way to see that small likelyhoods of a given event each year add up to much larger likelyhoods over time

I like what AlanGeorges said upthread
Quote
This is perfectly good math.  Yes, the Revolutionary War and Civil War had nothing to do with each other.  That is precisely the point – they were uncorrelated events.  There is no pattern.  From this standpoint it is also completely meaningless to talk about another French and English war being "overdue."  All these calculations can say is "based on prior events, here is the probability of something happening at least once in the next N years."  Reading any more into the conclusion, as if it somehow foreshadows some dire event, is foolish.  The author of that article never hints at such claims.  He merely lays out the math and applies it to the probability of something going badly wrong over the course of an average human lifetime.

This is how risk estimation is done.

Offline Carl

  • Mr HamTastic!
  • Forum Veteran
  • *********
  • Posts: 13105
  • Karma: 716
  • COW?...No ,I haven't seen your cow.
Re: Doing the math on why everyone needs a gun
« Reply #12 on: May 02, 2018, 04:51:31 PM »
  So ...birth control AND a shotgun can be many times more effective?

Offline Alan Georges

  • Survival Demonstrator
  • *******
  • Posts: 4593
  • Karma: 210
  • Still trying to reason with hurricane season.
Re: Doing the math on why everyone needs a gun
« Reply #13 on: May 02, 2018, 06:14:39 PM »
Hoo-boy.  Nobody's paying me tuition to tutor them in probability methods, so I'll just say to go back and re-read beginning at the paragraph starting with "We don’t buy houses".  It's a very clear explanation of the math.  Pay particular attention to the parts about this being a non-cumulative sequence, and a little farther down, about how the probabilities for "not-flooding" several years in a row multiply through.  If that doesn't work for you it is unlikely that anything I can say here will bring you around.
« Last Edit: May 02, 2018, 06:25:21 PM by Alan Georges »

Offline LvsChant

  • Resident Master Mudder
  • Global Moderator
  • Survival Veteran
  • ******
  • Posts: 7817
  • Karma: 614
Re: Doing the math on why everyone needs a gun
« Reply #14 on: May 03, 2018, 07:19:24 AM »
haha! I really love this article. Thanks for sharing...

Offline Smurf Hunter

  • Survival Veteran
  • ********
  • Posts: 7172
  • Karma: 334
Re: Doing the math on why everyone needs a gun
« Reply #15 on: May 03, 2018, 09:27:47 AM »
Hoo-boy.  Nobody's paying me tuition to tutor them in probability methods, so I'll just say to go back and re-read beginning at the paragraph starting with "We don’t buy houses".  It's a very clear explanation of the math.  Pay particular attention to the parts about this being a non-cumulative sequence, and a little farther down, about how the probabilities for "not-flooding" several years in a row multiply through.  If that doesn't work for you it is unlikely that anything I can say here will bring you around.

Stuff happens:


Offline Carl

  • Mr HamTastic!
  • Forum Veteran
  • *********
  • Posts: 13105
  • Karma: 716
  • COW?...No ,I haven't seen your cow.
Re: Doing the math on why everyone needs a gun
« Reply #16 on: May 03, 2018, 09:40:24 AM »
 1 Individual -  gun = greater than average odds of becoming a statistic

 1 individual + 1 GUN = greater than average chance of survival

  Why is it that the people who want citizens disarmed all have armed security?

Offline LvsChant

  • Resident Master Mudder
  • Global Moderator
  • Survival Veteran
  • ******
  • Posts: 7817
  • Karma: 614
Re: Doing the math on why everyone needs a gun
« Reply #17 on: May 03, 2018, 05:25:53 PM »
1 Individual -  gun = greater than average odds of becoming a statistic

 1 individual + 1 GUN = greater than average chance of survival

  Why is it that the people who want citizens disarmed all have armed security?

Well... not all. Some of them really believe that they can depend on law enforcement to take care of them... their credulity is rather pitiable. It isn't so much that I mean to disparage law enforcement, btw. It's just that they can't be everywhere at once and there are relatively few of them...