Selected feedback loop models
A feedback loop is a rule where a system's output changes its own input. Loops that amplify an advantage are positive feedback; loops that work against a gap, or that slow growth as a resource runs short, are negative feedback. These seven public diagrams from the Machinations Community show loops in games and in one population model.
We write about each one from what its author published: the name, the description and the number of components. Each one is embedded, so you can Play it here, and each one also opens on its own page.
Partisans - Game Loop
Partisans - Game Loop describes the turn cycle of a game about leading an insurrection in an occupied city, in 451 components. A turn starts with the player deploying partisans, weapons and intel; partisans can act again next turn, while weapons and intel are consumed. The underground network then supplies resources on its own, and recruitment of new partisans depends on civilian sympathy, which the player's actions and propaganda move. The occupying force then responds to accumulated animosity, which rises with partisan actions and can gradually fade over time. Each stage feeds the next. Play the embed and watch sympathy, animosity and recruitment move together.
Subway Surfers - Gameplay loop and monetization
Subway Surfers - Gameplay loop and monetization maps the endless-runner loop and the money around it, in 118 components. Keys and coins arrive through direct purchases or through play and sit in pools; converters spend coins on boosters and hoverboards, and ad interactions and purchase options feed the same economy. The feedback element is revival: a register calculates the cost of reviving the player and that cost increases, which the description ties to difficulty and engagement. Gates decide the flow from conditions such as score or revival cost, and drains record what is spent. It is a clear case of a loop that raises the price of continuing as it goes.
Open Subway Surfers - Gameplay loop and monetization on its own page
Population growth
Population growth has the structure of a classic balancing loop between food and people. Farms and the environment produce food at different rates into a central food store, the population eats from that store, and registers calculate food from farming and hunting, population growth and consumption per person. The population's appetite draws the store down, and the store in turn limits how large the population can become. Drains represent food consumption and a population decline when resources are insufficient, which eases the demand again. The description stresses sustainability: change the hunting efficiency or the food each person eats and watch where the balance moves.
Outer Wilds Loop
Outer Wilds Loop models a game built on repetition, in which dying sends the player back through the loop. The diagram tracks progress through stages (Find Coordinates, Find Ash Twin Project, Take the Core, Reach the Vessel With the Core) as Minutes tick by. A Player Chance gate gives a 50% chance of either finding a clue or dying, and a Loops Counter with a register decides when events re-trigger. The author's gates use both deterministic and dice distributions, which is how the model turns risk, reward and repetition into one cycle. The loop ends only when the player reaches the vessel with the core.
Triangle (interplay loops, triangularity)
Triangle models the rock-paper-scissors shape in game balance: three types in a circle, each strong against one and weak against another. Fire, Grass and Water Pokémon pools start with 30 units each, and connections set Fire against Grass, Grass against Water and Water against Fire. Automatic drains act as battle arenas where units are consumed, and dice formulas such as d2 and d4 on the connections keep every matchup uncertain, so an underdog can still win. Because each type both wins and loses against another, no single type is dominant by design. Change the starting 30 units or the dice and see which settings keep all three in play.
Open Triangle (interplay loops, triangularity) on its own page
Negative Feedback - Mortal Kombat 9
Negative Feedback - Mortal Kombat 9 is filed under the design idea of negative feedback, a mechanism that helps the side that is behind. It is inspired by the Super Meter of Mortal Kombat 9, a resource the player spends on special moves. Sources create basic attacks, special moves and defensive actions; pools hold damage, the Super Meter and the outcome of each attack (landed, blocked or missed); dice-driven gates decide those outcomes. Drains and converters spend the meter and turn actions into effects, and registers work out the damage taken and the enhancement of special moves. Open it to see a spendable meter treated as a balancing resource, then compare it with the glossary entry on negative feedback.
Multi Loop - Mortal Kombat 11
Multi Loop - Mortal Kombat 11 models two closed loops of counters inside a fighting game. In close combat, attacks beat grabs, grabs beat blocks and blocks beat attacks; at range, projectiles beat teleports, teleports beat projectile parries and projectile parries beat projectiles. Each player chooses a move from the set, gates and pools compare the choices, and a register works out the result and declares a winner. The model also plays several rounds in sequence, so you can watch how a strategy fares over a match. In a pure cycle like this every move loses to another, which makes the two loops a compact way to teach balance. Compare it with the glossary entries on multi loop and mini loop.
How to model a feedback loop in Machinations
A feedback loop exists when the value of a Pool changes how much flows into or out of it, directly or through other nodes. In Machinations you build that link with a State Connection or with a formula that reads a Pool or a Register. A link that makes the flow shrink as the value grows is a negative loop; one that makes it grow is a positive loop.
Play the diagram and chart the Pool: a negative loop tends to settle toward a level, while a positive loop keeps growing or collapsing. Change one number at a time and Play again to see which loop is in charge.
In the docs: State connections, Registers, Charts, Triggers.
In the glossary: Negative feedback, Slippery slope (positive feedback), Equilibrium, Mini loop, Multi loop, Comeback mechanic.
In the articles and Learn: Game systems: feedback loops and how they help craft player experiences, How to build your game loops in Machinations, Infinite runner game loop.
In solutions: Game economy design, Game economy design and balancing.
Frequently asked questions
What is a feedback loop in game design?
A feedback loop is a rule where a system's output changes its own input. A loop that amplifies an advantage, so the leader pulls further ahead, is positive feedback, sometimes called a slippery slope. A loop that works against the gap, such as help for the trailing side or growth that slows as a resource runs short, is negative feedback.
How can I spot a feedback loop in a Machinations diagram?
Look for a path that returns to where it started: a pool whose value changes, through a state connection or a formula, how much flows into or out of itself, directly or through other nodes. Play the diagram and chart the pool. A negative loop tends to settle toward a level, while a positive loop keeps growing or collapsing.
Is a game loop the same as a feedback loop?
Not exactly. A core game loop is the cycle of actions a player repeats, such as play, earn and upgrade. A feedback loop describes how the outputs of a cycle change its own inputs. Many diagrams show both; the Partisans and Subway Surfers models on this page are examples.
Do feedback loops always make a game fairer?
No. Negative feedback can keep matches close but can also feel unearned, and positive feedback can make wins feel decisive but lets a leader run away. Most designs mix both, and modeling the mix before you build lets you test how strongly each acts.
Can I copy one of these diagrams and change it?
Yes. Every diagram on this page is a public Community diagram. Open its page, sign in and remix it to copy it into your own account, then change the loop without affecting the original.
More selected models
Build your own feedback loop
Draw the loop, Play it and chart the pool to see whether it settles or runs away.