Beating the stock market is really hard. No, seriously. Most individuals who get into the stock picking game end up losing money, but a new Boston startup hopes to change that, by making it easier for anyone to test out a theory before putting any money on the line.

Based in Boston, Quantopian is building a community of self-identified “quants,” the math geeks turned lords of finance that have become one of Wall Street’s most valuable commodities. Today, Quantopian estimates there are about 10,000 quants at work on Wall Street crafting elaborate trading algorithms, a number the company believes is limited by the extremely high barriers to entry. Without the backing of a hedge fund or years spent securing data and programming a custom solution to back-test algorithms, it’s previously been tremendously difficult to try out stock picking algorithms. That’s what Quantopian hopes to change.

“We want to make being a quant easy,” said Dan Dunn, VP Product & Community. “I haven’t found anyone yet who has been able to do it in less than six months.”

(The Quantopian team includes a few HubSpot veterans, including Dunn and CTO Jean Bredeche.)

Quantopian lets anyone write a trading algorithm using just a little bit of Python code, and then back-test it against 10 years of minute-to-minute stock market data. Of course, that an algorithm would have made you money in the past is no guarantee of future success, but it’s a valuable way of sorting good ideas from bad.

Though the community only launched last August, Quantopian already has thousands of users, mostly drawn from the STEM fields. “We find that there are a lot of people who are interested in writing algorithms,” Dunn told me. Scientists, developers, and engineers who use mathematical modeling in their work and who have some basic programming skills are curious to apply their talents to the market, and Quantopian gives them an easy way to test hunches. (Users own their algorithms and they are private unless the user opts otherwise.)

For now, Quantopian doesn’t allow users to make any actual trades, just back-test algorithms. But in the future it hopes to support trading, and take a cut of the transations. It also hopes to charge as it adds data sources like options and futures.

The company has raised more than $2 million from GETCO and Spark Capital. “Spark loved the idea of the community,” said Dunn. “They know the idea of a community very well from Stack Exchange.”

In a blog post about his decision to invest in the company, Spark’s Andrew Parker expressed his belief in the power of vertical communities:

Niche, vertical social networks are fascinating because they are such internet-native businesses.  The internet is the perfect place to have that “Wow, I thought I was alone in my obsession with ________, but there’s thousands others like me!” moment.

Over email, Parker reminded me that Bloomberg generates billions in annual revenue from just a few hundred thousand users. In other words, the economics of this kind of financial tool are nothing like that of most “consumer” startups.

I closed my interview by asking the Quantopian team how they’d respond to a cynic’s suggestion that their users are bound to mostly lose money, as individual investors often do in the increasingly institutionalized market, perhaps armed with a false sense of security provided by having tested an algorithm.

Dunn acknowledged that, of course, some people would lose money. But, he insisted, the tools the company provides will have a positive impact on finance. “We’re making you think about it, we’re making you evaluate it,” he said. “If you have a bad idea, you’re going to identify it and weed it out.”