Actually, 220 - age is a rule of thumb based on one guy’s perusal of a bunch of regression equations. It’s sort of like an informal, qualitative meta-analysis. Hence, I treated it as if it were a linear regression in my post.

Standard error of the estimate applies to our uncertainty about the mean max HR given age. It doesn’t apply (I think) to the difference between someone’s max HR and the mean max HR given their age. Thus, I still think the term prediction error is more accurate. That said, this is splitting hairs as far as the Trainerroad crowd is concerned. All the OP needs to know is that they are still alive. If you are basing your training zones off max HR (and you should not be doing this, you should use FTP or lactate threshold HR instead) and if you know absolutely nothing about your own max HR, then a formula like 220-age is a starting point.