I am using Ben's data, though only in general terms. While I have done and published statistical research in an unrelated area (social sciences) in the past so am at least aware of, if not competent in, how to do so, I simply haven't the time right now to try to build on Ben's numbers, much as I'd like to. I'm too busy ranting about ICANN in general. :)|
So, I loosely interpret Ben's data to show that whenever a new, open, more or less undifferentiated TLD (including some ccTLDs, and one could also include new.net, IOD's .web, RealNames, Internationalized Domain Names, and a host of others) hits the market, it is immediately swarmed by speculators, cybersquatters, and defensive trademark registrants. While I can't quantify that, I think you'd be very hard-pressed to quantify otherwise. I have heard, for example, that as many as 50% of the registered IDN names are held by speculators. I do know that when Veri$ign opened IDN for registrations its server was hammered enough to take it offline for some hours. Now each time there is some new TLD, or even potential TLD, or even ersatz TLD, comes on the market, this happens again and again. I don't think my reasoning has to be grounded in good science to allege that we have a pattern here. Now whether this trend when pointed out by Ben rises to the level of significance or is substantial enough to pass some non-defined line before conclusions can be drawn is, to a degree, not the important question. Even if it doesn't rise to the level of scientific validity, it certainly passes my reasonable man test.
I agree with you that a properly done study to see if there is in fact any general demand for new, short, memorable addresses has not been done SFAIK, and would be extremely useful, almost a necessity, in focusing this debate. I don't think Ben is going so far as to claim that that is what he has done, though I admit that his conclusions perhaps come somewhat too close to that. I don't doubt that there is some, perhaps considerable, demand out there, I'd prefer such a choice myself, but if the level of demand is less than that of speculators and defensive registrants, or if one or both of the latter two are less but still get there first to those new, short, memorable addresses, then such a study would tell us almost nothing about how to correct that problem, it would only better show the extent of it.
Things are getting quiet in here so I think I'll move over to Peter Deutsch's submission, as I think that is germaine to much of what we are talking about. -g