Random Promotions and the Gervais Principle | Hacker News Reader

Random Promotions and the Gervais Principle

ribbonfarm.com

19 pointslliles9 comments

It can't work for a simple reason: you can't make promotions stick. I can name a half-dozen of my colleagues who've been promoted to team lead, discovered that it's a shitty job (this will obviously vary between organizations, but in some cases it involves a lot of responsibility with no authority to actually get stuff done) and returned to being "just" a worker. Then the second- and third-choices get a go and eventually you get to someone who wanted the job, and had in fact been angling for it all along, but that they're willing to do it in return what whatever motivates them (e.g. fancier title) doesn't correlate at all with being any good at it.

In a "normal" situation then, youll get that someone whos always wanted the job as the first pick to fill it, wich makes the randomnized promotions all that much desirable.

Also, the stickness of the promotion in your comment, sounds pretty random too. In the end, the proposition of "... using random processes to mitigate the pathological effects of deterministic models is extremely solid..." seems to hold.

I think he's on to something for education, too. There's an inherent sort of non-monotonicity in the admissions process: you want to admit the "best" students in order to make up the "best" class and give them all the "best" college experience, but even if you could somehow omnisciently rank the best 250 or 500 or 1000 applicants and accept from the top of the list, you would lose so much in heterogeneity (which is, I think, part of a good education) that it would no longer make up the "best" class or experience. (Of course you have problems if you admit too many from the low end, too, but that's why you put weights on the random process. :)

If we assume that Universities try to pick students with the highest chance of being "successful" after graduating, maybe testing for the ability to pass "tests" is not a bad thing to do. "Success" is hardly perfectly correlated with language or math ability.

Doesn't being good at resume building, making an impression via various mediums (essays, CVs, interviews..) imply being good at the things required for being "successful" in many of the success-rich arenas (Politics, large companies, the NGO world, etc.)? If someone will work for some charity to suck up to a power-broker to get an internship to put on a resume to get into a college also likely to figure out how to do the equivalent for an executive position 20 years later?

The way I see it, you are presenting these kids with a puzzle. You're not telling them its a puzzle. You're telling them its not a puzzle. Some call it a puzzle & solve it. Some call it a puzzle & cry foul. Some don't call it a puzzle & solve it anyway. Some don't call it a puzzle & don't solve it. Which is most likely to be a judge in 40 years?

*I'm not American & never went through a US style admissions process so I just view it curiously. I imagine that if I had, I would dislike it more.

Does any other hner can point me to other materials related to this series? I really think this series are, at least, a great intellectual exercise.

See http://searchyc.com/submissions/gervais?sort=by_date .

I've read all of the series in ribbonfarm, what I was looking for (in my poorly phrased english) is pointers to other related resources.

Ah. I don't have any decent reading suggestions. But as an unasked-for writing tip, I'd say you meant "like this," not "related to this."

Straight D’s in kindergarten

/cringe

Reply on news.ycombinator.com