|
So everyone thinks they're hiring the top 0.5%, even though the same 199 losers are spamming their resumes to every job posting. Yeah, when you think about it, this whole "we only hire the top 0.5%" condescension really amounts to a basic statistical reasoning fallacy. To construct an alternate hypothesis: the "real" threshold for ranked competence[1] at "elite" companies might, just might, be more like "the top 5%". But we can safely assume that 95% of the available talent pool in this category will be employed (and not actively seeking), compared with say 70% in the bottom, or "sub-elite" category, at any given time.[2] So for example in a "town" of 1000 putative developers, with skill levels ranging from god-like to "good enough" to mediocre to utterly bogus and self-deluding, let's say 50 (or 5%) of these are truly "elite", and 950 are "non-elite". Further, of the elite category, 47.5 will be employed (and not actively looking), and only 2.5 will be in the job market -- versus, out of the 950 non-elite developers, we might have 665 (or 70%) employed, versus 285 (or 30%) on the market and actively looking. Given this over-simplified (but arguably "less wrong") hypothesis, and assuming that a given opening draws applicants whose fitness ranking is evenly distributed across the pool of in-the-market candidates, it's easy to see that for any given opening, we'll have a ratio of 2.5:285 for elite:non-elite developers -- or a paltry 0.869%(!) of the candidate pool (2.5/287.5) will appear to be from the "elite" category. That is, for every 1 "elite" resume in your inbox, you'll be seeing 114 "non-elite" resumes. And since tech geeks are, by and large, an insecure, envious lot, they like to brag and fudge a bit -- so that 114 easily becomes 199 in their minds (or the 1 a 0.6, whichever). Which gives you the infamous "we only hire the top 0.5%" urban legend right there. Even though they're all really just angling over the... top 5% (ranked over the general pool of putative developers). And since hot-shots like Joel Spolsky, not to mention nearly all of the investment bank / hedge fund types, and well, probably half or more of everyone in the monthly "Who's hiring?" thread just keeps repeating it, and repeating it, and repeating it... well, shucks, it must be true. [1] As such a thing as generic competence exists can be objectively measured, etc, etc, but that's a separate issue. [2] Which corresponds -- conservatively, for the purposes of this argument -- to another widely espoused, but never quantitatively substantiated belief out in startupland: that "all" of the good developers are "always" employed and "almost never" on the job market longer than 2 days. |