Is booze for programmers like spinach for Popeye? The Ballmer Peak, a reference to Steve Ballmer of Microsoft, holds that imbibing alcohol improves cognitive ability, up to a point–a variation of the Yerkes-Dodson law, which shows a correlation between arousal and performance. Anecdotal evidence suggests that optimal programming occurs about two beers in. But what sayeth science?

recent study at the University of Illinois tested the creative problem solving ability of a group of men who were given vodka cranberry and snacks and asked to solve brain teasers. The results were starkly different for the tispy group, which had a blood alcohol concentration level of 0.075, versus the control group:

Astonishingly, those in the drinking group averaged nine correct questions to the six answers correct by the non-drinking group. It also took drunk men 11.5 seconds to answer a question, whereas non-drunk men needed 15.2 seconds to think. Both groups had comparable results on a similar exam before the alcohol consumption began.

The study notes that the Ballmer Peak effect was present for creative problem solving but not for working memory. So drink up, coders. Just stay off Twitter.