The Input System

Unity has two input systems, and most tutorials you find use the old one. Knowing which you are looking at saves a lot of confusion, because their code does not mix.

Old vs new

The mental model

The new system separates what the player wants to do from which button they pressed.

• Action — an intent, like "Jump" or "Move".
• Binding — a control that triggers it. One action can have many.
• Action Map — a group of actions for one context, like Gameplay or Menu.
• Action Asset — the file holding all of it.

Your code asks "is the player moving?" and never asks "is W held?". Add a gamepad stick as a second binding and the same code supports controllers with no changes.

Setting it up

1. 1
   Install the package

   Window → Package Manager → Input System. Unity restarts the editor afterwards.

2. 2
   Create an Input Actions asset

   Right-click in the Project window: Create → Input Actions. Double-click to open the editor.

3. 3
   Add a map and some actions

   A Gameplay map with Move (Value, Vector2) and Jump (Button) covers most games.

4. 4
   Bind controls

   Give Move a 2D Vector composite for WASD plus the gamepad left stick. Give Jump Space plus the south button.

5. 5
   Tick Generate C# Class

   In the asset's Inspector. That produces a typed wrapper you can use directly.

Reading input

using UnityEngine;
using UnityEngine.InputSystem;

public class PlayerInputHandler : MonoBehaviour
{
    private PlayerControls controls;   // the generated class
    private Vector2 moveInput;

    private void Awake()
    {
        controls = new PlayerControls();

        // Events: never miss a press, even between frames
        controls.Gameplay.Jump.performed += OnJump;
    }

    private void OnEnable()  => controls.Gameplay.Enable();
    private void OnDisable() => controls.Gameplay.Disable();

    private void OnDestroy()
    {
        controls.Gameplay.Jump.performed -= OnJump;
        controls.Dispose();
    }

    private void Update()
    {
        // Polling: right for continuous values like movement
        moveInput = controls.Gameplay.Move.ReadValue<Vector2>();
    }

    private void OnJump(InputAction.CallbackContext context)
    {
        jumpQueued = true;
    }
}

Switching contexts

Action maps make pausing trivial. Disable Gameplay, enable UI, and movement keys stop driving the player while menu navigation starts working — no isPaused checks scattered through your code.

public void Pause()
{
    controls.Gameplay.Disable();
    controls.UI.Enable();
    Time.timeScale = 0f;
}

public void Resume()
{
    controls.UI.Disable();
    controls.Gameplay.Enable();
    Time.timeScale = 1f;
}

Rebinding

public void StartRebind(InputAction action, int bindingIndex)
{
    action.Disable();   // required while rebinding

    action.PerformInteractiveRebinding(bindingIndex)
        .WithControlsExcluding("Mouse")
        .OnMatchWaitForAnother(0.1f)
        .OnComplete(op =>
        {
            op.Dispose();
            action.Enable();
            SaveBindings(action);
        })
        .Start();
}

private void SaveBindings(InputAction action)
{
    PlayerPrefs.SetString("bindings", action.actionMap.asset.SaveBindingOverridesAsJson());
}

Touch

Touch is just another device. An on-screen stick can bind to the same Move action as WASD, so your movement code never learns it is running on a phone.

What to take away

• Check which input system a tutorial uses before following it.
• Actions describe intent; bindings describe buttons.
• Call Enable() or you get no input at all.
• Poll continuous values, use callbacks for discrete presses.
• Action maps make pause and menu contexts clean.
• Rebinding is built in — use it.