The Streak Secret Squad Busters Doesn't Want You to Know
When Squad Busters removed win streaks, it revealed a far more fundamental truth that game economy designers misunderstand about loss aversion. Loss aversion is misreported as losses hurt more than equivalent gains. But there's nothing behavioral about this. This is simply a function of neoclassical economics and a diminishing marginal utility curve wherein each step forward along the bowed-in curve yields less utility than an equivalent step backward, again as a result of its bowed-in nature rather than any sort of weird fundamental human bias.
The "behavioral" critique is more subtle: utility is shaped by where outcomes land relative to an internal benchmark, not by their absolute size. Two identical end states can feel very different depending on whether the player experienced them as a gain or a fall from expectation. Consider match-3 under two scenarios:
A: Player never builds a streak. They have 11 attempts, winning 10 in a row and losing the 11th. B: Player builds a streak. They have 11 attempts, winning 10 in a row (activating Super Light Ball) and losing the 11th.
Nothing about the final state differs! Yet the second path feels worse. Not because the player lost more in absolute terms, but because the session included a drop below a reference point the game itself had just established. The streak redefined what "normal" progress felt like, and breaking it made later, otherwise-neutral outcomes feel like perceived losses. Identical outcomes can generate different utility once expectations are formed and then violated.
Returning to Squad Busters, remember their win streaks were functionally capped (after 10 wins, players received +3 "taps" or chances to upgrade the loot box), but the number on the badge continued to increment (some had thousands of wins!). The feature was removed, with Supercell citing that players felt anxiety about losing their win streak and stopped playing altogether, as noted in this Squad Busters update.
From a neoclassical perspective, nothing was being lost when a streak broke beyond the foregone future taps, which were capped anyway. From a reference-dependent perspective, a lot changed. Breaking the streak meant falling below an internal benchmark the game itself had encouraged players to form. Play after the break was not evaluated against a neutral baseline; it was evaluated against a recently experienced, higher standard of normal.
This is the same mechanism as the match-3 example: the path includes a rise to a reference point followed by a drop. Once that drop exists, ordinary outcomes are experienced as losses.
Seen this way, Supercell's explanation that players felt "anxiety" about losing streaks is not evidence of irrationality or fear-based loss aversion. Stopping play was not avoidance of loss; it was avoidance of reference violation.