Comment by krisoft

> Which part of my statement you find difficult to understand?

I understand your whole statement perfectly. It is just wrong. My understanding is not the problem here.

We are not comparing them to other samples. We say that out of the currently existing X they are the best.

Imagine a town with 3 bakeries. Lets call them A, B, and C. Bakery A gets shut down by the health deparment and B goes bankrupt. Then we can, rightfully and without survivorship bias, call C the best run bakery of the town. Because if you get shut down by the health department, or you go bankrupt then by definition you are not the best run bakery. (Obviously it is not a high praise with that kind of competitions, but they still are the best run bakery.)

Staying in the business is not some incidental part of “being the best run bakery”. It is a core component of it.

Imagine a marathon with 100 runners. Henry runs the fastest time, and 25 others do not finish. Some got lost, some had medical issues during the race. Is it survivorship bias to call Henry the fastest competitor in that race? Of course not. You need to finish the race to be even considered to be the fastest. Just because there are others who didn’t make it, doesn’t make him somehow not the fastest. Definietly doesn’t make calling him the fastest “survivorship bias”.

Finishing the race is a core component of “being the fastest finisher”.

Similarly in the case of wikipedia. If other similar sites stopped operating then they by definition did not “successfully navigated that minefield”. Their bakery is shut and they did not finish their marathon. That is the very definition of “not succesfully navigating that minefield”.

This is how rationalwiki defines survivorship bias: “Survivorship bias is a cognitive bias that occurs when focusing on entities that made it past a selection process, while overlooking those that didn't.”

We are not overlooking the failed attempts here. We are considering them.

Bakery A and B is worse run than C. And we know that because they got shut down.

The runners who did not finish the marathon are not faster than Henry. And we know that because they haven’t finished the marathon.

The abandoned community edited websites are worse at “successfully navigating that minefield” than the ones which are still operating. We know that because they are no longer operating. They were not overlooked.

What you are missing is that the “selection process” here is not some independent, and unrelated thing. The selection process is, at least in part, is what we are talking about. You cannot be considered the best run bakery unless you are running a bakery. You cannot be considered the fastest racer unless you finished the race. And your community edited website cannot be the one who most succesfully navigates a minefield unless you are navigating the minefield at all.

Please let me know if any of the above is unclear. Happy to go into details.

kurtreed2 7 hours ago

It is indeed a survivorship bias since we have no good other sample in the form of competitor to compare to, like how Pepsi is to Coca-Cola. Which part of my statement you find difficult to understand?

Reply View 1 reply

krisoft 23 minutes ago

> Which part of my statement you find difficult to understand?
I understand your whole statement perfectly. It is just wrong. My understanding is not the problem here.
We are not comparing them to other samples. We say that out of the currently existing X they are the best.
Imagine a town with 3 bakeries. Lets call them A, B, and C. Bakery A gets shut down by the health deparment and B goes bankrupt. Then we can, rightfully and without survivorship bias, call C the best run bakery of the town. Because if you get shut down by the health department, or you go bankrupt then by definition you are not the best run bakery. (Obviously it is not a high praise with that kind of competitions, but they still are the best run bakery.)
Staying in the business is not some incidental part of “being the best run bakery”. It is a core component of it.
Imagine a marathon with 100 runners. Henry runs the fastest time, and 25 others do not finish. Some got lost, some had medical issues during the race. Is it survivorship bias to call Henry the fastest competitor in that race? Of course not. You need to finish the race to be even considered to be the fastest. Just because there are others who didn’t make it, doesn’t make him somehow not the fastest. Definietly doesn’t make calling him the fastest “survivorship bias”.
Finishing the race is a core component of “being the fastest finisher”.
Similarly in the case of wikipedia. If other similar sites stopped operating then they by definition did not “successfully navigated that minefield”. Their bakery is shut and they did not finish their marathon. That is the very definition of “not succesfully navigating that minefield”.
This is how rationalwiki defines survivorship bias: “Survivorship bias is a cognitive bias that occurs when focusing on entities that made it past a selection process, while overlooking those that didn't.”
We are not overlooking the failed attempts here. We are considering them.
Bakery A and B is worse run than C. And we know that because they got shut down.
The runners who did not finish the marathon are not faster than Henry. And we know that because they haven’t finished the marathon.
The abandoned community edited websites are worse at “successfully navigating that minefield” than the ones which are still operating. We know that because they are no longer operating. They were not overlooked.
What you are missing is that the “selection process” here is not some independent, and unrelated thing. The selection process is, at least in part, is what we are talking about. You cannot be considered the best run bakery unless you are running a bakery. You cannot be considered the fastest racer unless you finished the race. And your community edited website cannot be the one who most succesfully navigates a minefield unless you are navigating the minefield at all.
Please let me know if any of the above is unclear. Happy to go into details.

Reply View | 0 replies