The Logic of Failure (2011) | Hacker News Reader

The Logic of Failure (2011)

tersesystems.com

40 pointstimdaub8 comments

> Delayed Feedback. Human beings tend to assume that an action will yield a response immediately, or not at all. This is the way that we interact with the world on a daily basis, and we can become very confused when there's a significant delay in the system's response.

This is what I find the most difficult, personally. When the feedback actually arrives, it's hard to link it to possible causes located far away in the past. It's also difficult to live with the fact that our actions can have distant consequences (but of course they do).

I was a project manager on web development efforts at IBM when I read that book and "Managing the Design Factory"[0] around the same time. The combination of the two completely changed how I approached managing projects generally.

[0]http://www.amazon.com/Managing-Design-Factory-Donald-Reinert...

Systems Theory; the greatest miss in general education.

just having a better feel for Stock and Flow would be beneficial for everyone who is voting.

It is fascinating to read about all these thinking failure modes - but my own (very tentative) takeaway from that book was that there is no one way of thinking that is appropriate everywhere, that if many people are failing in a particular way in one simulation game - then there is probably another game where their way of thinking would be the winning strategy.

Your comment seems to encapsulate the thinking process found to be effective in the experiment.

E.g. You say: very tentatively ... there is probably another game ...

Such measured language suggests that you'd try to test your hypotheses, given the opportunity.

Well - there was another game described in that book (but not in that linked article) where it was better to act quickly upon any tentative hypothesis instead of gathering data first and doing much testing.

On the other hand, the thinking process found to be effective appears to be more recursive, and the failure thinking modes seemed to be less so: less data driven, less flexible in their handling of feedback, and so on.

I might suggest that this article/research sets out a meta-thinking process, a second-order thinking process if you like (akin to second-order moral principles: another level of abstraction), compared with many other more narrow, less flexible, less recursive and more simply rule-driven than this.

Reply on news.ycombinator.com