You’re right, and I didn’t mean to imply that all charts should start at Y=0. If you follow the argument all the way to the end you’ll see the greater issue here is not so much the y-axis choice but rather the date range. I believe the data was (naively) manipulated, leading to the wrong conclusions and ultimately a bad story. You’ll see in the final chart (with truncated y-axis) I can use the same data source to tell the complete opposite story.