Here is the story, that I want to share.
When I've been making a Goblinobol, when activity hit test servers, and we started to gather some stats from it, something strange discovered. Blue team had 10% more winrate then intended. That make me double check all mechanics, and then triple check it again. Nothing was found. Then I've started to watch videos taken from the matches by QA.
Here I must explain some mechanics, though my story would make sense. All mechanics is made through the ball, it has certain amounts of auras, each aura does its thing. There is an aura, when it "sees" a player in it, it "attaches" ball to the player, which it sees first in server tick, and disable itself, and enable other mechanics, though the player can interact with the ball. But what will happen, if two players will enter the "attach a ball to a player" aura in the same server tick?
Well, I thought, it will be random, they all there, in same tick, aura can not prefer one player to another. Or did it?
On the video, each time two players from the different teams run to the ball in the same time, every time, it was player of Blue team who gets the ball. (The team called Red and Blue in mechanic). The tests confirmed it, when two players are in the same aura in the same server tick, aura prefers the player from the Blue team each time.
Because Blue is higher in the alphabet.
I had to add additional "randomize" there, problem was solved. winrate hit "magic" number of 49.9%/49.9% (there were some draws), and the activity went to live.
And I got a good story to tell to colleagues.