What is going on with that KDE chart? From the code, it seems like the X-axis is hour-of-submission, and density is determined by average points per DOW/hour combo.<p>So all the chart shows is that the greatest proportion of points from votes on the weekend come from submissions a few hours later than the proportion of weekday votes, which doesn't answer the question "is it better to submit on a weekday or a weekend?"<p>> <i>Time-related variables explain less than 1% of variation—meaning, unsurprisingly, that the other 99% depend on something else. </i><p>Where do you get the 1% figure from? That's not shown anywhere in the code. You could calculate a R^2 from a linear regression, sure, but the data violates all the rules that make linear regression valid.