For all their court-mandated blogging and self-promotion, venture capitalists keep the details of their decision-making to themselves. But newbie VC Ryan Sarver, Twitter's former director of platform, broke ranks about his first month on the job as partner at Redpoint Ventures.

In a widely-shared post on Medium, Sarver said what really dropped his jaw was not, as we suspected, unsubstantiated valuations and sly M&A manipulation, but rather how little VCs know before investing:

Placing Bets

I knew trying to pick the startups that will have huge wins would be hard, but I didn't know it would be this hard. I've been astonished by the relatively small amount of data we have on which to base our investment decisions.

Here's how it goes: you meet with founding teams a handful of times at most. You get a smattering of stats and financial predictions that may or may not be based on reasonable assumptions. You listen to a narrative about where consumer or corporate behavior is expected to trend. You do whatever due diligence you can. And then, you decide on whether or not you want to work closely, through the highs and lows, with someone for the next seven-plus years of your lives. That's right. Seven years in on average how long it takes to really know if an idea has worked the way you hoped it would.

So if investors have a paucity of data and don't spend that much time talking to the founders, how do they figure out where to send the check? Hello again, good ole pattern-matching.

This is where you hear people talk about pattern matching. You get very few dots in space and you need to quickly be able to match those to patterns you have seen before in order to be confident in pushing your chips in. The more you see, hopefully the better you become, but that clearly isn't absolutely true or the most experienced venture capitalists would statistically be the most successful.

I think the reality is that while pattern matching is critical, you also can't become too locked into a specific set of patterns as they too change over time. Consumer behavior has changed, platforms have changed, the cost of doing a startup has changed, the profile of an ideal founder has changed. So if you stay locked into the patterns of old you'll miss the next Evan Williams or Drew Houston. Best to balance staying agile with a solid knowledge base to have the best chance of discovering The Next Big Idea™.

Considering how well being open-minded to the best idea or entrepreneur has worked out for them so far, they might wanna tweak the pattern.

To contact the author of this post, please email nitasha@gawker.com.