Idea for refining the "effective APM" metric (unit rightclick spam)#165

Only one suggestion per post, okay.

I was just watching a stream by MrPlanner, and we discussed that many players might achieve high eapm numbers not by actually playing fast, but just by spam-clicking the same group of units 5-10 times in half a second instead of letting 1-2 clicks do the same thing.

A more practical way to count how fast a player effectively plays the game could be to ignore some of these spamclicks. This wouldn’t be a perfectly objective metric, but it could be a decent estimate of how fast a player actually is.

Implementation idea: Remember the last group of units that received a rightclick command, and ignore any pure rightclick commands given to that group for the next 0.5 seconds, escaped early if a different type of input is made.

We have discussed this internally fairly recently, but decided against it as the benefit is too small for a pretty large effort. Basically it’s really difficult to accurately tell when players are being purposefully spammy. A quick example is a player that right clicks a unit 5 times to go to the same overall spot out of habit, and a player who does so because of it being necessary due to pathing quirks.

That makes sense, it would indeed be not that reliable of a metric, and would take some effort to implement