Nice ! For the "when to give up" problem, optimal stopping theory (wikipedia) provides a framework that may help (essentially 37% of total time you are willing to work on the problem - e.g. 2 years for a PhD).
I've known about optimal stopping theory and the 37% heuristic for a long time but I've never once managed to (or felt like) using it in real life. It is possible that I don't understand how to use it. But also possible that in real life you never really have enough data to ever be able to apply optimal stopping theory effectively.
Agree that the theory is too rigid for real-life application. But i find that the rule-of-thumb of 1/3 : 2/3 for calibrate : persist generally applies well to most settings.
Examples : 2yr to find a PhD problem (6yrs for a successful phd), atleast 3 interviews before accepting (assuming a 9 interview journey), 2-3 years in a startup before giving up (8years for a succesful startup). At the 1/3rd point, if you dont see a good path forward, give up and try something else.
We live in the era where only a final reward is celebrated. To reach a goal, there are various dips a person goes through, but the person is recognized for goal achieving and that specific moment, I feel even the process of ups and down should count. Celebrating dips, as dips are signals of us reaching closer to our goal should be introduced.
Nice ! For the "when to give up" problem, optimal stopping theory (wikipedia) provides a framework that may help (essentially 37% of total time you are willing to work on the problem - e.g. 2 years for a PhD).
Thanks Shamik.
I've known about optimal stopping theory and the 37% heuristic for a long time but I've never once managed to (or felt like) using it in real life. It is possible that I don't understand how to use it. But also possible that in real life you never really have enough data to ever be able to apply optimal stopping theory effectively.
Agree that the theory is too rigid for real-life application. But i find that the rule-of-thumb of 1/3 : 2/3 for calibrate : persist generally applies well to most settings.
Examples : 2yr to find a PhD problem (6yrs for a successful phd), atleast 3 interviews before accepting (assuming a 9 interview journey), 2-3 years in a startup before giving up (8years for a succesful startup). At the 1/3rd point, if you dont see a good path forward, give up and try something else.
Good examples. Thanks.
Indeed a great article , I loved reading it.
We live in the era where only a final reward is celebrated. To reach a goal, there are various dips a person goes through, but the person is recognized for goal achieving and that specific moment, I feel even the process of ups and down should count. Celebrating dips, as dips are signals of us reaching closer to our goal should be introduced.
Good article