Your calculation is indeed probably quite a bit off. As a health scientist, part of my work is to detect correlations in data that look messy to the naked eye. So, I’m right now applying my tools to figuring out how the earnings algorithm works. I can only do this with my own numbers because I don’t have access to other writers’ stats. And the problem is that, given the number of predictor variables (e.g. member views, reads, claps read-to-view ratio etc.), I need more articles (as for each article, its earnings is the outcome variable). And I still have too few of them (I just started on Medium a few months ago). But I’m working on it. I aim to arrive at an equation that predicts earnings reasonably correctly. When I have that algorithm, we can see what weight the individual predictor variables have. And, voila, you’ll have your strategy for what to aim for. One caveat, though: I suspect the algorithm will never make 100% correct predictions because FoM’s views will count differently from mere partners’ views. Also, I guess that the algorithm penalizes certain scam behaviors, like clapping without having spent enough reading time. But once I have the algorithm, I’ll put it in a spreadsheet and attach it to a post. So everybody can use it. And if several other writers contribute their data, we can get better still.