In plain words
A feedback loop is any cycle where an outcome feeds back into its own cause. Balancing loops pull things back toward a target. Reinforcing loops push things further in the direction they’re already going, faster and faster.
See it
A thermostat is a balancing loop: too cold, heat on; warm enough, heat off. A snowball rolling downhill is a reinforcing loop: bigger means it picks up more snow, which makes it bigger. Playground marbles for keeps are a reinforcing loop too, which is why one kid ends up owning the shoebox.
Why designers care
Loops decide how a game’s story unfolds. Reinforcing loops make early choices matter and give games an ending, but too much hands the win to whoever got ahead first. Balancing loops, like handicaps or letting the loser go first, keep things close, but too much makes skill feel pointless. Spot which kind you have.
Common mix-up
“Feedback” here isn’t the same as player feedback. One is a loop inside the system. The other is the game telling a player their action worked.
Origin The study of such loops, cybernetics, was named by Norbert Wiener (1948).