I saw an idea for ranking Overwatch heroes by adding their pick and ban rates, then using each hero's highest score across input types and rank groups. Does that actually measure how oppressive a hero feels?
Does pick rate plus ban rate make sense for ranking oppressive heroes?
Answers
It's a useful starting point, but I'd call it a popularity and ban-pressure score rather than a direct measure of oppression. A hero can be picked often because they're fun or versatile, and banned often because people find them annoying or don't want to deal with a matchup.
Adding the rates also gives a percentage point of picks the same weight as a percentage point of bans. That's a choice, not proof that they mean the same thing.
Reply to this answer
Taking the highest group for every hero is the part that makes me pause. It shows where each hero's combined rate peaks, but the final ranking could compare one hero's all-rank controller score against another's top-rank mouse and keyboard score.
I'd show the four group scores alongside the peak. That way you can see whether a hero stands out broadly or only in one setting. A small group could also produce a noisy peak, so sample sizes would help.
Reply to this answer
I'd keep picks and bans separate at first. A high-pick, low-ban hero tells a different story from a low-pick, high-ban hero, even if their totals match. Seeing both numbers makes the ranking much easier to discuss.
Reply to this answer
The numbers won't tell you which part of a kit is causing frustration. Use the ranking to pick heroes worth looking at, then compare their abilities and the counterplay available. That's a different question from whether they're objectively strong.
Reply to this answer
Add your answer
Answers are reviewed before publication.
DOSmouseTrail_5L:
So the peak is useful for spotting a problem in a particular group, but not necessarily for saying who's most oppressive overall?
P8airBlock_4M:
Exactly. Label it as a peak score and keep the group attached to it.