There's an infamous 2002 Rolling Stone article about Weezer frontman Rivers Cuomo where the nerdy and unlikely rock star stops the interview to show rock journalist Jenny Eliscu an elaborate spreadsheet where he mathematically breaks down the best pop and rock songs in search of a "magic formula."
As unorthodox as the approach sounds, it's not far from what the immensely talented and sharp Durham-based Bombadil are doing on their new record. The band heads to San Francisco next month to lay down the tracks with venerable engineer/producer John Vanderslice, but they've already been featured at places like FiveThirtyEight for their data-heavy songwriting approach.
According to bandleader Daniel Michalak, the idea actually came from a fan who came to a Bombadil show in Chicago a few years ago.
"We were on tour in Chicago and a good friend up there came to see us, and he noticed a lot of the songs we played involved the topic of love, and he was asking why we were doing that," the singer and multi-instrumentalist recalls. "And we said something like, 'Well, people like those songs better.' He was like, 'Wow, I wonder if there are other things about songs, an aspect of a song, that makes it better and worse.'"
This was not an idle question. The friend, Nasir Bhanpuri, worked as a data scientist for a healthcare systems company and did predictive modeling — and he was keen to apply his knowledge to the band's music.
"I basically broke down every Bombadil song for him, any kind of category you could think of — number of drums, amount of bass, who is singing what, what speed, what key," explains Michalak. "And he used internet streaming data to see which songs were 'best' and why, what aspect of songs makes one more popular than the others. Like, you need to add the drummer's voice, or this songs need to be 10 beats per minute faster. So that was the experiment with a lot of new songs we're recording in September, to try and [apply] his feedback."
While the approach sounds unusual, in some ways it feels par for the course for a band who so willingly embraces the carefree and esoteric as Bombadil. The group started in 2005 as a sort of fun-house pop version of The Avett Brothers, embracing a wry, harmony-drenched approach that felt like a retro mash-up of the Kinks and Simon & Garfunkel, while also featuring echoes of Belle & Sebastian and the Shins. It was humorous and heartfelt, humbly song-oriented, and relentlessly adventurous. It was, above all, distinctive.
The band seemed poised to follow in the Avetts' footsteps but has been sidelined sporadically over the years by rotating band members and an extended period where Michalak was suffering major nerve damage in his hands. While they've never reached the commercial heights of their North Carolina brethren, they have proven to be masterful record makers as much as songwriters, with each of their five full-lengths bumbling with ideas and stretching the limits of their sound.
In a sense, their new algorithmic approach is a conscious reigning in of their creative impulses — in another, it's yet another stretch.
"I think it's very much in line with what we've done in the past, just trying to keep it fresh and different," says Michalak. "And maybe it's just another really weird, quirky way of doing something. That's what our management thought — 'Oh, there's quirky Bombadil doing something weird again.'"
And that may be, but it's hard to deny that the experimentation stems from a fundamental urge to push forward and try new things, something which makes sense for a band so interested in tinkering with arrangements and pop formulas.
"It's been fun, a really outside-the-box creative process for me," admits Michalak. "It's an experiment. We talked about [the challenge] a lot, how if you followed the algorithm, every song would sound the same. So we did pick and choose to make sure it wouldn't sound like that. But from a band's perspective, we've seen that records that sound the same do better than records that sound completely different from song to song. People might actually like having the same voice on every song and like having that kind of continuity. So we actually were trying to make a similar-sounding record. We tried writing similar melodies and in similar keys with similar word choices."
Michalak doesn't really think the band will take the same approach again; the point was just to try something new. And it's clear that the turn toward data-driven popularity was never about cracking some code of commercial success, the way Cuomo's spreadsheet might have looked. It was simply another creative way, or even opportunity, to make a different kind of record.
"It's not something that feels stressful," he says of the band's dedication to left turns. "We just have so many things that we want to do, so many records that I want to do that we haven't done. I've never felt a lack of ideas or direction. The sky is the limit. I think for better or worse, we've given ourselves a blank check for what Bombadil can do."