An SEC document filed Wednesday and reported by VentureBeat reveals 500 Startups is looking to raise money for a third early-stage start-up fund. Just last week, the Mountain View, California start-up accelerator finalized $44.1 million for its second fund. It had previously tried to raise $50 million. 

While this news might add fuel to critics' assertion that it's a "spray and pay" outfit, it's business as usual for Dave McClure, the incubator's sharp-witted founder. 

In a recent interview, McClure addressed the idea of 500 Startups being run like sweatshop. "We think that we are a high-value investment firm," he said. "We invest early, way earlier than most people, at an accelerator and seed stage. The earlier you invest, the higher that attrition rate is because it's compounded. We have a more scientific method of investment." 

Put another way, McClure invests by an algorithm

"Our target is at least a 20 percent or higher success rate to institute levels of capital from series A to B and beyond," he said. "We like to think of ourselves as lab scientists at a petri dish rather than gamblers at a roulette table. In the same way those MIT students gamed Vegas, we're taking that strategy to the world of internet start-ups. It's about predicting whether or not they're fail." 

As the name implies, 500 Startups is also built for high volume. "Most firms do two investments a year," said McClure. "They may talk to between 50 and 100 and do a couple, whereas we're talking to thousands and doing several hundred. We invest in several companies a week, perhaps even one a day. Most are doing that once a quarter." 

But isn't it difficult to keep tabs on so many start-ups? Not for McClure, who's brought on more workers to take on the busywork and suss out the hits. "Genius emerges slowly," he said, and "we have three core metrics that we look at to find it: user growth and retention, revenue growth and economics per customer, and other downstream investment decisions. If we see two of those three things, but preferably all three, we're happy. If we only see one, it's a relatively easy decision to not invest." 

McClure stresses that he isn't looking for "the next Facebook or Twitter." Although he may run into those "because we deal with so many," he feels "disruptive innovation is overrated. There's a shit-ton of problems that need to be solved, which are straightforward. A company like Red Bus in India, which is a basic ticketing company, might not necessarily be disruptive, but there's a much lower stack of problems that need to be solved."