News for hackers. Stuff that matters.

A comment on The Kelly criterion: How to size bets (2019).

Re:The Kelly criterion: How to size bets (2019) by ultra_nick on Tuesday February 08, 2022 @07:54PM (#30267162)
Would this work for calculating the odds of startup success?

Runway of 500k, a 5% chance, 100:1 odds, so don't bet more than ~$22k on each bet/pivot.

[ Parent | on HN ]
Re:The Kelly criterion: How to size bets (2019) by cortesoft on Wednesday February 09, 2022 @01:52AM (#30269740)
The only number you know there is the 500k… everything else is just a guess.
[ Parent | on HN ]
Re:The Kelly criterion: How to size bets (2019) by rohitb91 on Tuesday February 08, 2022 @09:17PM (#30267849)
It is definitely one way to approach startup investment and why VC funds spread their bets out on so many startups. Almost all go to 0, some go 100x.
[ Parent | on HN ]
Re:The Kelly criterion: How to size bets (2019) by shakezula on Tuesday February 08, 2022 @07:58PM (#30267193)
I think it would be a good heuristic if you were able to be that precise about expenses, but I think most startups aren’t.
[ Parent | on HN ]
Re:The Kelly criterion: How to size bets (2019) by paulgb on Tuesday February 08, 2022 @08:08PM (#30267263)
(Author here) I agree, but I think there is some applicability of the Kelly Criterion to startups (at least qualitativey) in explaining why an entrepreneur who is all-in on a company may be more inclined to take an exit than their investors, who have spread their bets.
[ Parent | on HN ]
Re:The Kelly criterion: How to size bets (2019) by kqr on Wednesday February 09, 2022 @01:12AM (#30269522)
The Kelly criterion is a U-shaped optimisation problem, so it is rather tolerant of imprecise estimations!
[ Parent | on HN ]