To go back to an earlier link, I was a little wary of this link declaring that the quants were about to take over venture capital. Here’s the nub of the article, as I see it: “If InLab’s magical model works and math can replace that vaunted gut judgment that to date has been a critical criteria for deciding whether to invest in a startup, then why should VCs be using their gut to pick investments at all?”
The trend towards quantification has been a defining one of the past decade or so, and it’s had some wonderful results (e.g. Google, sports, logistics) and some actively harmful results (e.g. high finance). Which would a quantified venture capital tend towards? I suspect the latter.
Statistics are just another form of language, but this isn’t generally understood (as measured by our actions); we generally give numbers a much higher credibility than words, when in fact a number is just another way of representing a bunch of words. You’d never be able to get away of some of the high finance models if you described what they were saying in natural language; indeed, that’s the source of the vast incredulity that’s greeted the revelations of Wall Street’s tactics during the high finance boom.
Given that, thinking about what modeling actually entails in quantifying venture capital shows why people should be wary of the trend: it will take average data of startups in the past and apply them for the future. The problem is that venture capital is about seeking exceptional companies of the future. The future won’t be like the past and exceptional things are by definition dramatically different than the average. It’s hard to predict either one of these things; both at the same time is very different, and even more difficult by looking at statistics, as the typical statistics may not show the path of the future. Therefore modeling venture capital is probably the wrong way to do it; meanwhile, the gut, while not reliable, can be a great indicator of the shock of the new and truly novel.