Survivorship Bias 101

History is written by the victors. But if we exclusively focus on these successes, we allow survivorship bias to distort our understanding of the world.

But what is "survivorship bias" and how does it work?

Here's Survivorship Bias 101!

👇👇👇

1/ First, a few definitions.

Survivorship bias is the logical error of concentrating on survivors (successes) and ignoring casualties (failures).

When we do so, we miss the true "base rates" of survival (the actual probability of success) and arrive at flawed conclusions.
2/ When we completely ignore failures, we lose our ability to correctly identify the differences between successes and failures.

Put simply, exclusively focusing on successes may actually inhibit our ability to identify (and replicate) the actions that led to such success.
3/ Cicero wrote on the topic over 2,000 years ago.

An atheist named Diagoras is shown portraits of people who prayed and were saved from death at sea as proof of God's existence.

Diagoras replies, "I see those who were saved, but where are those painted who were shipwrecked?"
4/ Cicero cuts right to the point - we cannot conclude that prayer led to being saved from the sea if we ignore those who prayed and then drowned.

Survivorship bias can lead to a deeply-flawed understanding of cause-and-effect relationships.

Let's look at some examples.
5/ One of the famous examples of survivorship bias comes from World War II.

The U.S. wanted to add reinforcement armor to specific areas of its planes.

Analysts plotted the bullet holes and damage on returning bombers, deciding the tail, body, and wings needed reinforcement.
6/ But a young statistician named Abraham Wald noted that this would be a tragic mistake.

By only plotting data on the planes that returned, they were systematically missing data on a critical, informative subset - the planes that were damaged and unable to return.
7/ The "seen" planes had sustained damage that was survivable.

The "unseen" planes had sustained damage that was not.

Wald concluded that armor should be added to the unharmed regions of the survivors.

Where the survivors were unharmed is where the planes were most vulnerable.
8/ Based on his observation, the military reinforced the engine and other vulnerable parts, significantly improving the safety of the crews during combat.

Wald had identified the survivorship bias and avoided its wrath.

Where else do we see survivorship bias arise?
9/ In business?

We all love business success stories. We love to read about companies and people who have achieved incredible success and fame.

We scan their backgrounds in desperate search of the actions or features that led to that success.

But these conclusions are flawed.
10/ For every entrepreneur who took out a second mortgage in a last-ditch attempt and succeeded, there may be 10 who did the same and went bankrupt.

For every company that pursued a moonshot idea and is now worth billions, there may be 10 who did the same and collapsed.
11/ In markets?

Analysts may try to calculate the performance of funds or groups of stocks using only the survivors at the end of the period.

By ignoring closed funds or stocks removed from an index, we bias the data and overstate returns.

Flawed data ➡️ Flawed decisions
12/ The point? By ignoring the companies or people who failed, we miss out on valuable data points from those cases.

To actively fight this, we must study the features of both successes and failures.

The "seen" and the "unseen" are equally important in guiding our decisions.
13/ As @nntaleb wrote in his classic, Fooled by Randomness, "Heroes are heroes because they are heroic in behavior, not because they won or lost."

So the next time you read a story of the hero who won, be sure to reflect on all of the heroes who lost, whose stories are not told.
14/ So that was Survivorship Bias 101. As someone who shares many success stories on Twitter, I thought it was an important topic to address.

Special thanks to @austin_rief of @MorningBrew, whose interesting tweet on new founders sparked this thread. https://t.co/ldSVsPP4uX
15/ For more on the topic of survivorship bias, I highly recommend the below resources.

Fooled by Randomness by @nntaleb: https://t.co/fB9ncIMlvE

Great blog from @ShaneAParrish and @FarnamStreet: https://t.co/o1T85OJMl7
16/ And for more educational threads on business, money, finance, and economics, check out my meta-thread below. Turn on post notifications so you never miss one! https://t.co/53UhhfzIcp

More from Sahil Bloom

Short Squeeze 101

If you follow financial markets (or if you watch Billions), you've heard the phrase "short squeeze" used quite frequently.

But what is a "short squeeze" and how does it work?

Here's Short Squeeze 101!

👇👇👇

1/ First, the basics.

The "short" in "short squeeze" refers to the concept of short selling.

The basics are covered in my thread below.

https://t.co/xecJYNCxMs

TL;DR - short selling is a way of betting against a stock - i.e. betting that its price will decline.


2/ "Short interest" is a measure of how heavily an asset is shorted by the market.

It is the total number of shares that have been sold short (borrowed and sold), but have not yet been covered (bought and returned).

It is usually measured as a % of the # of shares outstanding.

3/ A "short squeeze" occurs when a heavily-shorted asset experiences a rapid upward price movement.

When this happens, short sellers may be forced to close their short positions (i.e. buy the stock and return it to the broker), further accelerating the upward price movement.

4/ Let's look at a simple example to show this in action.

We will use Tesla, one of the most heavily-shorted stocks in the world.

Imagine the stock price is $1,000 per share. This seems crazy. Ricky Rational decides to short the stock at this level.

More from For later read

There is some valuable analysis in this report, but on the defense front this report is deeply flawed. There are other sections of value in report but, candidly, I don't think it helps us think through critical question of Taiwan defense issues in clear & well-grounded way. 1/


Normally as it might seem churlish to be so critical, but @cfr is so high-profile & the co-authors so distinguished I think it’s key to be clear. If not, people - including in Beijing - could get the wrong idea & this report could do real harm if influential on defense issues. 2/

BLUF: The defense discussion in this report does not engage at the depth needed to add to this critical debate. Accordingly conclusions in report are ill-founded - & in key parts harmful/misleading, esp that US shldnt be prepared defend Taiwan directly (alongside own efforts). 3/

The root of the problem is that report doesn't engage w the real debate on TWN defense issues or, frankly, the facts as knowable in public. Perhaps the most direct proof of this: The citations. There is nothing in the citations to @DeptofDefense China Military Power Report...4/

Nor to vast majority of leading informed sources on this like Ochmanek, the @RANDCorporation Scorecard, @CNAS, etc. This is esp salient b/c co-authors by their own admission have v little insight into contemporary military issues. & both last served in govt in Bush 43. 5/

You May Also Like

1/12

RT-PCR corona (test) scam

Symptomatic people are tested for one and only one respiratory virus. This means that other acute respiratory infections are reclassified as


2/12

It is tested exquisitely with a hypersensitive non-specific RT-PCR test / Ct >35 (>30 is nonsense, >35 is madness), without considering Ct and clinical context. This means that more acute respiratory infections are reclassified as


3/12

The Drosten RT-PCR test is fabricated in a way that each country and laboratory perform it differently at too high Ct and that the high rate of false positives increases massively due to cross-reaction with other (corona) viruses in the "flu


4/12

Even asymptomatic, previously called healthy, people are tested (en masse) in this way, although there is no epidemiologically relevant asymptomatic transmission. This means that even healthy people are declared as COVID


5/12

Deaths within 28 days after a positive RT-PCR test from whatever cause are designated as deaths WITH COVID. This means that other causes of death are reclassified as