Abraham Wald is a brilliant mathematician that lived in Hungary before World War II. Being a Jew, he was discriminated against, and when the Nazis took over, he emigrated to the US and quickly joined the new "Department of War Math", where he and other scientists are asked to help solve math problems that is related to war. And one of them is about bomber survivability.
 |
| English: Boeing Y1B-17 in flight (Photo credit: Wikipedia) |
Obviously, one cannot armor the bomber enough to make it bulletproof, and make it still fly and carry bombs. So where should the armor be added? And how much? Where should the trade-off be done between bomb payload and armor? That's where Wald and his colleagues come in.
The story goes that the scientists and assistants flew to Washington, to be briefed by the US Army Air Corp generals and their staff, where they explained the problem, along with representatives from Boeing (the builder) who explains the structure of the aircraft and explains the problems. Then there's data from the USAAC where they present the data they gathered from the bombers that survived, where they are hit and patched, so on and so forth. And they tried armoring the parts that keep getting shot up, but it's not helping the aircraft to come back safely.
Allegedly Wald listened to some more of this, then stood up, told the generals that they are looking at it ALL WRONG. The parts that got shot up on the bombers that came back are the parts that should NOT be armored.
Wait, what?!
Really! He's right! Think about it for a minute...
Do you see why yet? If not, let me explain.
Remember, we are NOT seeing the bombers that had been shot down. We're only seeing the bombers that *did* make it home. Did the bombers that did NOT come back did something wrong? Or is it just luck?
We can only judge the vulnerabilities on the bombers and see where armor can be improved, i.e. add armor in the spots that will make the most difference. Except, the only data we have are on the ones that *did* come back. And the instinctual reaction is to COPY THEM. Most simply assume that came back must be doing *something* right. The factor of luck is ignored. We thought we can COPY LUCK.
So why the parts that *were* shot up were the parts that do NOT need armor? Because those bombers made it back WITHOUT needing armor in those places. Assuming that every part of the plane can be hit and will be hit (i.e. uniform hit distribution) it's clear that OTHER parts of the plane that got hit caused the plane to crash whereas the damage on the plane that came back are NOT as critical!
This sounds counterintuitive, but only if you go by first instinct, which is why it's considered a cognitive bias, that today is known as "survivorship bias", in that you think the survivors did something special to survive when it's often just blind luck.
And this has far more applications than you think. Successful businesses or individuals often are simply 'at the right place at the right time with the right people (and willingness to fulfill a need)". It's not as if they have some special insight or special business process or whatever. But because they survived and thrived, and other companies or individuals did not, we tend to think of the survivors as having done something special. (and they go on seminars and write books and we hear about them even more)
When you think about it, the *same* happens in network marketing. You are constantly seeing the "winners" (read: survivors) being paraded in front of you as some sort of sterling example of what should be done, and they insist that they didn't do anything special... that if you do the SAME THING they did, you can also be successful like them. But that's basically admitting that their success is purely a fluke, a matter of luck, or a roll of the dice, doesn't it?
You don't ever hear about the people who failed at network marketing, do you? Your leaders are trying to ENFORCE survivorship bias on you if they tell you to "avoid negativity", "ignore the critics", "they just don't understand us", "those who failed are lazy", and so on and so forth. They think their success is depending on their business skill when they also claim that no skills are needed; just follow the system. Isn't that admission that it's all luck, and their claim on what is the cause of their success completely BOGUS?
Just today, I encountered two separate individuals who used the same argument: anyone who failed at _____ must be lazy / not growing their business (read: recruiting).
One wrote: "I simply meant the people that join and do not follow through with what it takes to make a success out of the business outnumber those that do."
The other wrote: "Those who are ... complaining about not make money are not making money because they are spending their time ... complaining instead of working at what they learned to grow their business."
Or imagine this conversation:
Recruit: "The company says my chance of success is 1 in 100, or worse, according to the income disclosure statement. What if I'm the other 99"?
Upline: "If you work hard you will succeed."
Recruit: "If 99 of us will fail anyway, what's the difference between them and the success? Surely SOME of them worked hard too."
Upline: "No, because they failed. They can't have worked as hard as people who succeeded."
That's a self-justifying argument. They refused to acknowledge the factor of luck. To them, the system is an idiotproof success copy machine. Work hard and you shall be rewarded, when there's no proof of such. In fact, this particular company's own income disclosure stated very clearly that 99% of "active" participants receive less than 35000 a year. Of these, 74% receive less than $600 a year. Hardly a definition of success, is it, out of roughly 150000 participants?
Think about that... Next time you see one of those "winners"... and why they are the exception, not the rule, despite their claims.
Reference:
http://lifehacker.com/how-survivorship-bias-skews-your-perceptions-and-decisi-1564223180
http://youarenotsosmart.com/2014/04/24/yanss-podcast-022-how-we-miss-what-is-missing-and-what-to-do-about-it-with-statistician-megan-price/
Building a business is not like buying a lottery, it does not depends on odds and chances.
ReplyDeleteThis applied if you are building a traditional or MLM business.
In MLM business (and any other business), you get paid if you move a product/service to a customer or the team that you build doing the same. You don't get paid to go to power up meeting, to stare at the phone and afraid of making calls. Heck, you don't even get paid by talking to people about your product/business.
You only get paid when there is a business transaction. But all those activities above might leads you to that goals.
Once you understand this, you must figure out if the product deliver values to the customer (will the customer buy the product if it is not marketed thru MLM?). You must also figure out the profit that you are going to make if transaction took place, will it make it worthwhile for your resources invested?
That's how you evaluate the business.
The problem is often in MLM, quite a bit of the compensation depends on recruiting, even though it's not explicitly stated as such (because that'd be illegal).
DeleteIf you are selling the same stuff as 10000 other affiliates, what value are *you* adding to the product to make the customers want to buy it from you, rather than your upline / downline / sales group? You don't even have geographic exclusivity any more with advent of internet.
So I don't disagree with you, Darren, but I think most MLM noobs never thought about the part about "adding value". NOR DO THEY look at the odds. Yet if they don't "improve" their odds, then they are indeed playing the lottery, and the more they recruit, the more likely they'll hit a jackpot.
THAT is the dangerous mentality I'm trying to explain. AND that sort of recruitment driven mentality is what turns a normal network marketing group into a pyramid scheme. You no longer try to sell... you are recruiting, and you've convinced yourself that you're selling the opportunity. That is a dangerous delusion.
"'survivorship bias', in that you think the survivors did something special to survive when it's often just blind luck."
ReplyDeleteYou're mixing two different topics. The planes that made it back DID do something right -- they didn't get shot up in the places that would bring them down.
The fact that pure statistical chance determines where they got shot is irrelevant to the mitigation strategy. Whether it's luck or not, armoring the areas which didn't get shot up, as recommended by Wald, would STILL be the better strategy.
Survivorship bias would be: If the generals didn't armor ANY part of the planes, because they assumed it had nothing to do with armor, and everything to do with the skill of the pilots.
You're missing the point completely.
DeleteYou thought:
>> The planes that made it back DID do something right -- they didn't get shot up in the places that would bring them down.
They didn't do anything. They simply survived a die roll.
You're doing exactly what I wrote: "you think the survivors did something special to survive when it's often just blind luck."