I can't remember where I found the idea or if there is any scientific evidence behind it but I like it, increasing your rate of failure is the best way to learn

I can't remember where I found the idea or if there is any scientific evidence behind it but I like it, increasing your rate of failure is the best way to learn something. If it took a computer 30 minutes to compile code, compared to a few seconds for a small program today, you could only run a program 8 times a day and debug your mistakes. So I would say, do not worry about time, but find a method of learning where failure is swift :)