(Ferenstein Wire)—A technical glitch during New Year’s Eve turned out to be an information treasure trove for Uber. Last New Year’s Eve in NYC, the surge pricing algorithms failed and didn’t allow Uber to incentivize more drivers to get on the road with higher pay, just as partygoers rushed to call drivers; only 1 in 4 of users who wanted a driver actually got one.

Uber will no doubt use this as evidence against New York state officials who are eyeing regulation of the company’s controversial dynamic pricing model. Some officials are not fans of the fact that Uber prices can be unpredictable or can be many multiples of a traditional cab during peak hours.

But, the flip side of not allowing surge pricing is that Uber becomes as unreliable as the taxi companies on New Year’s, when it’s very difficult to find a ride. Surge pricing has allowed Uber to provide reliable rides during peak demand, and, in San Francisco on New Year’s, there’s been a noticeable drop in drunk driving arrests.

From a data perspective, the report authored by Uber’s data team and the University of Chicago’s Chris Nosko, is also a wonderful lesson in real-life randomization. These kinds of events are goldmines for economists. A big company like Uber would never intentionally turn off a feature that makes them so much money. As a result, data scientists never really know the effect that a particular feature has.

Theoretically, Uber could have been wildly overestimating the amount they should be charging consumers. Perhaps no surge was needed. Without an example of where surge pricing isn’t in effect during high demand, no one can reasonably argue against that critique. Now we know that surging does, in fact, matter.

Data scientists refer to this real life event as “quasi-randomization,” when something happens in the world that allows us to study it like it was created in a laboratory. In essence, NYC became a giant testing ground when the glitch shut off surge pricing. The effect was quite dramatic.

It’s a smart lesson for both social scientists and business leaders: Uber took what, at first blush, seemed like a major malfunction and leveraged the data to learn something valuable (which their policy team can then use against aggressive regulators).

On a final note, with this data in mind, the case against surge pricing becomes all the more clear. If you think that surge pricing should be illegal (or capped), then you essentially believe that public services should be a lottery. Instead of letting people who are willing and able to afford a driver pay more, a law would artificially cap the price and permit the service to a fraction of citizens who are lucky enough to press a button at the right time.

Possibly, this is what regulators will argue for, but the new report does force them to be honest about their reasoning.

Curious readers and data nerds can read more about the study methodology here.