This issue has its own thread now:
We're working on a new stat to help convey the reach of your posts here
Please call People Helped something else or drop it: that name is disingenuous.
You've explained the way this is calculated, and I've read it:
We also added a new "people helped" counter based on the views your questions and answers have gotten. This is essentially a sum of views of your questions, and answers that fall within the following criteria: accepted, have a score of 5 or more, have more than 20% of total score of all answers on a question or are within the top 3 answers by score. No deleted posts are counted.
The problem is, we all know full well that view counts are not remotely correlated to people helped. It pretends all of the following scenarios, which would increase the view count, also signify a new person helped, even though that isn't the case (this list is not exhaustive):
- Someone who happened across the question, determined it wasn't what they were looking for, and left.
- People who were visiting to read, already knew everything written in an answer, agreed with it and considered it helpful, and upvoted that answer.
- Repeat visitors, who aren't looking for an answer to help them. (In 2010, people investigating the views count mechanism found that the same visitor could increase the view count multiple times, if they simply spaced out their visits. If this is still current, I could count for 5 "people helped" myself on a question just revisiting my own answers to edit them further.)
- People visiting to perform moderation, or check that things are okay. (This accounts for most times I've added to a view count)
The people who visited and were helped by an answer is only a portion of views. It's also an indeterminable portion of the views. (That means we can't emulate such a number accurately either.)
My answer to Betrayed by my GM; how can I respond?Betrayed by my GM; how can I respond? is one of my highest-scored, most viewed answers. I'm confident I made a difference to some people, but I don't believe it's accurate to say the views (6595) capture that number. This was an extreme and toxic social situation most viewers will never have to deal with (thankfully). Fewer than that could have learned something, or maybe through a ripple effect a lot more than that number did. But I don't know how many were helped, really, and neither do any of us.
The honest thing to do would be to admit we don't know how many people we help. Views are views. That statistic is more accurately labelled "views on questions you've answered which meet certain criteria" or something else, but "people helped" is not an accurate label. I'm not sure what such a statistic would mean, exactly, but if you display it, please do not label it "people helped". Raff has suggested in comments the possibility of calling it "total readership", which is one possibility much nearer the mark.