some cases of logic that's not immediate common sense

Status
Not open for further replies.
Shannow

Shannow

Joined
Dec 12, 2002
Messages
43,965
Location
'Stralia
Couple of examples of proper use of logic have popped into my psyche over the last decade, the first of which has helped me a lot since coming across it...the latter not so much but will help my kids (when they listen to it eventually).

First one was Abraham Wald, on armor protection of bombers.

Bombers during WWII were returning back the base, often shot to bits...a lot didn't make it.

Bomber command started planning to install armor...and the LOGICal place to put it was to protect those places that the planes were usually shot to bits...so they started doing that.

Wald ?

He looked at it differently.

Assuming a normal distribution of bullets in the average shot up plane, the places that there's WEREN'T bullet holes were likely those places that didn't allow survival and return on the plane...the places with the holes, and successful return didn't need protecting.

Picture below for example...You can see why the original "mistake" gets made.

How has it helped me ?

Holes (no pun) in the available data are often a good place to look

bullet hole logic.webp
 
Other one, a bit more difficult to apply is by Author Seth Godin (I posted a vid of a speech a couple weeks ago).

In his book "linchpin", he makes an example of fields to chose if you want to be outstanding.

"Don't choose bowling"....

his logic is that there's a limit to the perfect score. No-one can beat that score, and all of the best in the world just do it more often, which is a poor statistic to compare legendary status versus simply very very good.

Example he uses as the statistically most outstanding athlete who ever lived is Cricketer Don Bradman....compared to the best in the world of all time, he stands well beyond best, and in a league of his own.

bradman linchpin.webp
 
Originally Posted by das_peikko
What do the red dots represent?


As I read it, the red dots are the relative distribution of damage on returning bombers - where they were finding the most damage when surveying battle-damaged aircraft. So, the temptation was to armour those areas. Pretty sure that's a Lockheed Hudson, a major RAF type in the early part of the war.

The real issue was where the damage was less common from the damage survey:

The engines (loss of propulsion - fatal)
The cockpit (loss of crew - fatal)
The narrow area of the aft fuselage before the empennage, aft of the upper ball turret (structural failure), and
The narrow area of the wing forward of the outer 1/3 of the flaps (structural failure)

Hudsons with damage in these areas weren't returning, and thus not showing up in the study of battle damage.
 
Last edited:
I read about the bomber armor issue in another book, I think by Richard Overy , where he cited the same research. I think similar principles were applied to the U-boat war in WW2 and how best the Allies could counter the threat: no single tactic was responsible, but rather a combination of tactics including increased airborne patrols, Leigh leights , airborne radar, convoy and sonar. Once the cracking of Enigma was revealed, the popular conclusion was that it was code breaking and sonar that defeated the u-boat campaign. However this recent book reveals the huge amount of "operations research" that combined these methods into a winning strategy , and downlayed the code breaking. I highly recommend both.

Black May U-boat war
https://www.amazon.com/gp/product/B003YCOOYY/ref=ppx_yo_dt_b_d_asin_title_o02?ie=UTF8&psc=1
The Bombing War
https://www.amazon.com/gp/product/B00C9Q24NE/ref=ppx_yo_dt_b_d_asin_title_o03?ie=UTF8&psc=1
 
I thought the thread title was about some BITOG threads.
grin2.gif
 
I've heard similar conclusions about Army helmets in WW1. Guys were surviving to show other injuries once their heads were protected.
 
Good observation. One of the most important things with statistics is to ask what isn't there, and why.

In the airbag example, a statistical analysis might observe an increase in crash related injuries, failing to account that it doesn't figure in that a lot of those people should be dead. A person in more than one serious accident over there lifetime continues to contribute to the injury statistics, where they maybe should have never contributed at all.

There's other things that can slip through the cracks with numbers. The BSA (burn surface area) diagram comes to mind. Using the "one Palm, one percent" method, the BSA has the human body split up into percentages by body part. This process leaves the groin circled off from the other body parts and rated at 1%.

Statistically, a person with a burnt forearm has greater injury than a person burned solely in the crotch, but I think we know better.
lol.gif
 
Originally Posted by eljefino
I've heard similar conclusions about Army helmets in WW1. Guys were surviving to show other injuries once their heads were protected.


Interesting reversal of that story...we were more exposed to British TV as kids, and there were regular sightings of stoic british workers riding bikes to work, in their suits with little clips to stop their trouser legs getting caught in the chain (I used to have a set when I rode to school in uniform).

Per New Scientist in the '90s

Britain brought in mandatory helmet laws, which reduced head injuries to an extent greater than they predicted.

An emergency doctor over time realised that there were less cycling injuries all up, broken legs arms, ribs etc. and started analysing why this anomaly occurred.

Lower participation...people getting to the office with helmet hair discouraged them from riding.

Researchers ran with the lack of participation and ran a series of interviews, and the exercise from the cycling wasn't being replaced...extrapolation indicted that more people would die from Cardiovascular disease than were saved from head injuries.

Not a fan of people creating numbers and percentages on future projections like that but clearly there were less participants, and they weren't replacing that exercise.
 
Like all the old motorcyclists being killed....it's because there are less younger riders. It's a wonder they haven't used those statistics to ban us old guys from riding motorcycles.
 
Originally Posted by Shannow

In his book "linchpin", he makes an example of fields to chose if you want to be outstanding.

"Don't choose bowling"....

his logic is that there's a limit to the perfect score. No-one can beat that score, and all of the best in the world just do it more often, which is a poor statistic to compare legendary status versus simply very very good.

Let's try another view...

If you want to be outstanding, then you would choose bowling over football, soccer, etc. Because there are large numbers of people who bowl, but there are far fewer slots available to do it professionally compared to the number of professional sports team slots available for football, soccer, etc, players.

So even the worst professional bowler is still more outstanding in his chosen field than a professional football, soccer, etc, player is in his.
 
Here's one related to oil. Using USA speed limits and driving habits.

There are speed limits on roads. Many people say everyone will always do 10 or 20 mph more than the limit no matter what it's set at, so it has to be set lower to account for that. But that is false.

If a road's speed limit is 55mph and the average actual speed is 65mph, much of that is because people driving on that road think 65mph is still a safe speed. If the speed limit were raised to 100mph, the average speed would not be 110mph due to self-preservation (due to various reasons).

Since the average speed is 65mph even with a 55mph speed limit, the best way to prevent accidents is to raise the speed limit up from 55mph and set a minimum speed. This reduces the speed differential between vehicles, which is a bigger cause of accidents than higher speed.
 
The problem we have here is people driving under the speed limit....cars stack up behind, and then take chances with passing.
 
Status
Not open for further replies.

Similar threads

V
Cut protective clothing underneath protective gear
Replies
12
Views
440
Bonz
Astro14
United 108 IAD-MUC Engine failure at 1,000 feet on takeoff
2 3 4
Replies
66
Views
2K
edyvw
B
Case study: Chemical cleaning stuck rings to eliminate oil burning
2
Replies
22
Views
3K
Patman
supton
F9F vs MiG-15s
Replies
8
Views
3K
Hohn
wwillson
  • Sticky
Pennzoil - Low-speed pre-ignition (LSPI) Q&A - Answers
Replies
6
Views
2K
garageman402
Back
Top Bottom