Merge Units

Merge same-level units to upgrade, leveraging exponential growth for long-term objectives; the core lies in recipes, output curves, and full-cell handling.

What is it

Drag two identical pieces together and they merge into a higher-tier one — 2048, Suika Game, and Merge Dragons are all built on this.

How it plays

You stack matching pieces and watch them pop into a bigger one; the grid keeps filling up while you juggle space and chase the next tier — simple, but impossible to put down.

Why learn it

The pull of exponential growth is nowhere clearer than here — if you want to learn how to give players a long-term goal, merging is the best specimen to dissect.

Merge play is a self-reinforcing growth engine: same-level merging converts low-frequency actions into deterministic tier jumps, and an exponential output curve keeps "the next tier" forever more valuable than "this one." The hard part is not the merge rule — it is three loop-closing details: a full grid must always have an escape that never deadlocks, the feed rhythm must leave a thinking budget, and the level cap decides how far the pull reaches. The "just one more merge" addiction comes from the curve shape, not the art.

Why This Mechanic Matters
For Players

Merging feeds the twin instincts of collecting and growing stronger: low-tier units are visible savings, and every merge is a certain upgrade — the goal chain is always clear, and the next tier is always beckoning.

For the Ecosystem

Merge is a dual winner of retention and monetization in casual games: the goal chain natively supports content expansion (a new tier IS new content), and grid pressure interlocks cleanly with stamina and booster monetization — proven repeatedly from Merge Dragons to Suika Game.

Transferable Craft

Learn merging and you learn exponential pull-curve design — escalating output, cap placement, feed pacing — a numeric structure that holds just as well in idle games and card collection.

Play Behavior
Merge Units
Gameplay Features
Complexity
Beginner → Advanced
Learning Goals
  • Beginner: understand the minimal loop — feed → same-level merge → level up → higher output — and reproduce one merge and one full-grid handling in the demo
  • Intermediate: write the full merge-rule contract (same-level judgment, recipe table, full-grid policy) and tune the cost/output curve to control the growth rhythm
  • Advanced: design pity and anti-farming mechanics, and verify board health with full-grid rate, average merge interval, and tier distribution
Core Patterns

② Instant Playable Demo

Merge Units Demo
Loading game…
Rule Hints
Rule: Acquire units → Merge same-level units → Upgrade to higher-level units → Increase output → Trigger full-cell handling when a cell is full.
Lab Guide: treat the demo as an experiment
Pressure Valve Experiment
Do: Squeeze "Spawn Interval (pressure valve)" from 6s down to 1s
Observe: The interval is the board-inflation valve: at 1s, full-grid pressure spikes and you are forced into constant merging; at 6s you get real planning time. Spawn interval sets the thinking budget — the most direct cost-side tuning knob
Level Cap & Exponential Pull
Do: Raise "Max Level Cap" from 4 to 7 and compare how long the top tier takes to reach
Observe: Output grows exponentially per level, so each +1 cap doubles the pull distance to "the next level." Too low and the late game loses its goal; too high and the goal feels unreachable — where you cut the exponential curve decides the game’s lifespan
Full-Grid Watch
Do: With a 1s interval and a level cap of 4, refuse to merge until the board jams
Observe: On a full grid the system warns and blocks placement (or force-consumes low-tier units to free space). Without full-grid handling the board deadlocks — it is the line between minimal rules and a playable product

③ Behavior System Breakdown

3.1 Problems Solved
  • Transform low-frequency actions into clear long-term goals via merge-and-upgrade.
  • Maintain continuous pull toward the next level with an exponential growth curve.
  • Ensure the board remains playable through full-cell or unmergeable handling.
3.2 Minimal Ruleset
  • Same-level merge rule: at least two units of the same level merge into one unit of level +1; if the target cell is already full after merging, trigger full-cell handling.
  • Output growth: each unit's output increases with level, starting at 1 for the lowest level and up to 10 for the highest, forming an exponential curve.
  • Full-cell handling: when a target cell is full and cannot merge, the system alerts and blocks placement; if the player opts to force placement, the system automatically consumes low-level units to free space.
  • Recipe restriction: same-level merging only applies when the recipe table (e.g., 2→1, 3→1) is satisfied; units not listed in the recipe cannot merge.
3.3 System Loop (optional)
Input (place/drag) → Evaluation (mergeable?) → Feedback (merge upgrade) → State change (output increase / space release) → New board state → New input.
Design Notes
The pull comes from the exponential curve, not the reward count: doubling output per tier means the next tier is always worth chasing — but too steep and the late game feels unreachable, too flat and the goal dilutes. Where you cap the curve is the switch for the game’s lifespan.
Low-tier units are worth more than high-tier ones (counterintuitive): early output rides on a broad base of cheap units, and rushing tiers starves your own production — good merge design lets players discover "build the base first, then climb" on their own.
A full grid is not a punishment; it is a pacing valve: force-consuming low-tier units to free space looks like a penalty but is actually the fuse that prevents board deadlock — it turns "space management" from a background rule into an explicit every-move decision.

④ Common Combos & Variants

Merge Units + Idle
Offline resource accumulation for merging upon return, reducing daily active pressure.
Merge Units + Numbers
Exponential output curves amplify growth satisfaction, requiring synchronized consumption design to prevent inflation.
Merge Units + Levels
Use level objectives to guide merge direction, preventing aimlessness.
⑤ Advanced Design & Algorithms (collapsed)Click to expand
Design Warnings
  • A steep output curve can make later levels unattainable, while a shallow curve loses momentum; balancing the exponential growth rate is a core tuning point.
  • Missing full-cell handling can freeze the board; forcing consumption of low-level units keeps flow but diminishes their value.
  • High output from low-level units can provide abundant early resources, sometimes more important than high-level units—counterintuitive, so early resource allocation must be considered.
  • Random drops lacking a safety net amplify frustration; a guarantee and anti-spam mechanism are needed.
Algorithm Examples
  • Merge determination: same-level matching, recipe table, chain merge settlement.
  • Growth curve: exponential/segmented cost and output design, guarantee and anti-spam.

⑥ Related Case Articles