Skip to content

Auto Suggestion

The Auto Suggestion window runs the genetic algorithm to automatically find optimized balance parameter combinations. The result is a set of balance snapshots representing the best configurations found.

The window is opened via the right navigation bar.

Requirement: The Unity project must implement IGeneticAlgorithmFitnessEvaluator so the algorithm can evaluate candidates. More details under Auto Suggestion (Unity Plugin).


Pre-selecting Parameters

Before opening the window, the parameters to be optimized should be selected in the Parameters view. The window only shows these parameters and allows setting a value range (Min/Max) for each one, within which the algorithm searches.


Settings

Context Settings

Field Description
Snapshot Identifier Name for this optimization run. The best configurations will be saved under this name as balance files.
Base Balance Snapshot Starting configuration from which the algorithm begins.
Unity Scene Unity scene in which the simulations are executed.

Execution Settings

Field Description
Method Optimization algorithm. Currently only the genetic algorithm is supported.
Population Size Number of candidates per generation. Larger populations explore the search space more broadly but increase runtime.
Number of Generations How many generations the algorithm runs. More generations allow deeper optimization at the cost of longer total runtime.
Iterations per Individual How many times each candidate is simulated. Multiple iterations average out random variation and lead to more reliable fitness values.
Choose Top How many of the best candidates are saved as separate balance files at the end.
Speed Multiplier Acceleration factor per simulation run (1× to 20×).
Max Simulation Time [s] Maximum duration per simulation run in seconds. See Simulation.

Parameter Settings

A minimum and maximum can be defined for each pre-selected parameter. The genetic algorithm only generates values within these bounds.


Starting the Optimization

Clicking Start Auto-Suggestion validates all inputs and starts the process. Progress is shown in the status bar of the main window (Auto Suggestion: Generation X / Y).

At the end, the best configurations are automatically saved as balance snapshots in the project and appear in the snapshot selector in the status bar.