The idea that a video game or a fancy test could, in one fell swoop, better and more cheaply match people to their jobs is a seductive one. But it’s also a bit trickier than finding a way to recreate Jason Giambi’s offensive output at a fraction of his salary. Administering a test or a game remotely has the potential to pull in a lot more job candidates who might not otherwise have been considered — people who didn’t go to elite schools, for example — but it could also screen out entire classes­ of workers. Millennials might be totally cool with playing a video game set in a sushi restaurant as part of a job application, but some older workers might balk. (Then again, even I, a millennial, got a little sweaty-palmed when I sat down to play Wasabi Waiter.)

There’s also the matter of the quality of the tools themselves. Do they really measure the things we want them to measure? A lot of these start-ups promote projects they’ve done with individual companies that show strong correlations between some test question or observable behavior and various success metrics. Knack has worked with Shell, for example, and a lot of medical employers; and Evolv with Xerox and Harte-Hanks. But in an economy where work proc­esses and the requisite skills change very quickly, it’s not clear that yesterday’s correlation is a helpful predictor of whether an applicant will be a high performer tomorrow, said John Sullivan, an H.R. consultant and management professor at San Francisco State University. In the meantime, you might worry that job applicants can figure out how to game the system. Or that you’re just cloning the workers you already have, rather than bringing in new perspectives.

The very act of quantifying certain characteristics may also give a false sense of precision that leads to overweighting the things quantified. “You are what you measure,” warned Dan Ariely, an economist at Duke. Ariely thinks the hiring process should become more data-driven, but acknowledges that applying sabermetrics, or the sort of empiricism that helped turn Billy Beane’s Oakland A’s into a winner, to the recruiting process can still create distortions. Online dating sites, for instance, often overemphasize measurable and sortable attributes (like height and income) at the expense of other ineffable ones that might be more useful or relevant. “If I want to hire people for my basketball team, it’s easy to tell who’s seven feet tall and who can shoot the ball really well,” explained Alvin Roth, a Stanford economist who shared a Nobel Prize for his work on market design and recently joined Knack’s board. “The hard thing is figuring out when they’re on the court, how will the rest of the team do?”

Human beings still beat computers at detecting these sorts of soft skills, like empathy. But the people-analytics industry is hot on their tail. Right now, the gaps and glitches in these various vetting systems may encourage companies to subject their potential hires, like Daron, to even more opaque tests than a mystery-illness patient on “House.” But soon enough, perhaps, some firm might convincingly prove that a single, simple algorithm can replace the good old-fashioned grill session.

That might sound a little scary, but automating and expediting the hiring process, if done successfully, offers huge benefits. Three-quarters of the nation’s wealth is in the form of human capital; the talent and training of workers, in other words, offers far more value to the overall economy than anything else. Better allocation and investment could provide huge returns for everyone, which is particularly important at a time when 10 million Americans can’t find work and extended unemployment can deteriorate their skills. The faster we can place America’s workers in jobs they’re good at — even if they need to play a video game to get there — the better.