Blind Melon’s “No Rain” rated extremely low on novelty in Berg’s research. Dreamy, guitar-driven soft rock wasn’t exactly innovative in 1992. According to Berg, this was the sort of song that was very likely to become a one-hit wonder: It rose to fame because of a quirky music video, not because the song itself stood out for its uniqueness. After that hit, the band struggled to distinguish their sound from everything else that was going on in music.

By contrast, Twain’s breakout hit rated high on novelty in Berg’s research. She was pioneering a new pop-country crossover genre that was bold for her time but would later inspire a generation of artists, like Taylor Swift. “Twain is a great fit for the model, because her blending of pop and country was so original before she had her breakout,” Berg told me. After her second album, he said, her novelty, which had previously been an artistic risk, helped her retain listeners. She could experiment within the kingdom of country-pop without much competition from other artists, and this allowed her to dominate the charts for the next decade.

Berg’s research also found that musical variety (as opposed to novelty) was useful for artists before they broke out. But down the line, variety wasn’t very useful, possibly because audience expectations are set by initial hits. “After the first hit, the research showed that it was good for artists to focus on what I call relatedness, or similarity of music,” he said. Nobody wants Bruce Springsteen to make a rap album.