🤖 Sonya Mann  🎀Verified account

Er, survivorship bias isn't a thing when you're trying to measure the traits that make survival most likely.

Like, the survivorship bias is working *for* you here. That's the point of looking at such a dataset. (Granted, the tweets of this format are all a bunch of crap, but the reasoning behind looking at the folks who survive and what traits they have is sound.)

The point of survivorship bias is that you cannot see people who have the same traits but don't succeed, or don't succeed extraordinarily.https://twitter.com/enkiv2/status/1006842047511322624 …

That's right! Which means you'd want to compare the prevalence of those traits among "survivors" to the prevalence in the general population, if you have a feel for the prior distribution.

Yes, but that's not survivorship bias working for you, that's remembering to account for survivorship bias.

Survivorship bias is hand-picking the data you want for you in this example. Your survivors are a dataset biased towards the features you're trying to discover, if that makes sense. There's a great example on the Wikipedia page: https://en.wikipedia.org/wiki/Survivorship_bias#In_the_military …

The prior distribution in that scenario can be assumed to be random regular spacing (which might be a little wrong!), and the survivorship bias provides us with a set of features (bullet locations) that we're looking for (non-fatal penetration points).

I think what you're actually concerned with in the "successful people" scenario is that the survivorship bias won't differ significantly from the prior distribution, and people will be encouraged to exemplify the ordinary, "average" traits.

hahaha sonya went viral

eh, barely

Wow bigshot over here thinks a few k ain't enuff oh ok

@sannewman what I was saying about the mistake good people make thinking bad people are effective.

elaborate?

We were talking about how good people misjudge the effectiveness of bad, rulebreaking people -- we overvalue the ones who make it through, ignoring that most bad/sociopathic people are miserable failures.

ah yes, this is a good point! although IIRC sociopaths are overrepresented among CEOs relative to the rest of the population? who knows though, maybe that's fake like all of social science apparently

@sarthakgh win

I think it’s important to be aware of these biases but too many people (GaryVee voice) use it as an excuse to not do things that probabilistically increase the odds of getting success. I refer you back to my tweet :https://twitter.com/sarthakgh/status/1006915082113069057?s=21 …

that's totally true! I just see so much "follow these 8 tips for ultimate success!" and it drives me crazy

literally Success Cargo-Culting

the most honest one would be "I took a bunch of fucking drugs" and is also the only one they never write.

have you met our lord and producitivyt savor A M P H E T A M I N E S

What? I went to NYU and nobody took those to help nobody nobody at all