A study of more than 6,000 athletes published in PLOS One found that world-class athletes played a variety of sports before choosing one. As reporter David Epstein writes in his 2019 book, Range: Why Generalists Triumph in a Specialized World, “Among athletes who go on to become elite, early sampling across sports and delayed specialization is by far the most common path to the top.”
Can confirm anecdotally, athletes in my childhood circles that played well in college or went pro, all played football and baseball until late into high school.
Then there’s this. A study published in Creativity Research Journal found that Nobel Prize-winning scientists were nearly three times more likely than their non-prize-winning peers to have outside interests: hobbies like singing, acting, and creative writing, or “craft” hobbies like painting, woodworking, and glassblowing.
Main piece of extra advice is to not pick a hobby that's too similar to the main area you are achieving in.
If your hobby is too similar to what you do—if your leisure activity involves demands and skills similar to those of your profession—then the study found feelings of self-efficacy suffered. The same is true for the level of “seriousness” of the hobby: both how difficult it is and, more important, how important it feels to not only do, but to your sense of identity. The more “serious” the hobby, the more compelled you feel to engage in it, and the more your identity is tied to skill or achievement.
If I had time, I'd look closer at the first mentioned study:
A study of more than 5,000 entrepreneurs in a 15-year period published in Academy of Management Journal found the startups of people who kept their day jobs were 33 percent less likely to fail than those who went all in.
I have a strong sense there's a selection bias. People who kept their jobs and had their startups fail probably aren't going to report having started a startup