More trading strategies, or "there is one born every minute"

In this post I discussed a model for how some hedge funds make money.A couple of weeks ago I was chatting about this with an acquaintance who runs a mid-size university endowment, and he said: NO! That's not how it is done! I was prepared to hear that hedge fund managers are caring nurturers, and would never do such a thing, but instead he continued: What you do is this: Let's say you have $25 million in start-up capital. You divide it into five piles of $5 and start up five funds with variations of your strategy (which might consist of throwing darts at the bloomberg terminal). At the end of a year or two, by sheer luck, one of the funds will have done really well, some will have done ok, and some will have tanked horribly. At this point, you shut down the underperforming funds, and publish the results of the overachieving fund (in, for example, the standard databases). The stellar performance will attract suckers investors, and then you continue running that one fund, but now you hug the benchmark closely. What happens then is that your recent results don't look so stellar, but your results from inception look great. It seems almost too easy (and does not work with sophisticated investors like my friend, but, sadly these are in the minority).