Exponential Averaging 🧮✨

Quick idea: Exponential averaging is like having a magic memory that prefers fresh news but doesn’t totally forget the past. It smooths out jumpy data so your bot can aim better!

---

1. Why Do We Need It? 🤔

• In Robocode your radar sees an enemy robot at slightly different angles each tick (game step).
• Those readings can jump around because the enemy moves or because of tiny measurement errors.
• A jumpy red dot on your compass makes aiming messy – your turret might wiggle wildly.
• Exponential averaging calms the dot down by blending the newest angle with the earlier average, giving you a smooth, easy‑to‑follow path.

---

2. Everyday Analogy 🥤

Imagine stirring chocolate syrup into milk:

• At the start you only see swirls of dark and light.
• Each stir mixes the new syrup streaks with what was already blended.
• After several stirs the milk looks evenly chocolaty – the big swirls disappear.

Exponential averaging is like those stirs: each new swirl (new angle) gets mixed in, and the milk (the average) gets smoother every time.

---

3. The Magic Formula ✨

newAverage = (1 – α) × oldAverage + α × newValue

• α (alpha) is the smoothing factor (a number between 0 and 1).

  • α = 0.1 – slow change, very smooth (good when data is noisy).
  • α = 0.5 – faster change, follows sudden moves quickly.

• oldAverage is your previous smooth value.

• newValue is the latest reading from the radar.

---

4. Step‑by‑Step Example 🧮

Tick	New Angle (°)	Old Avg (°)	α	New Avg (°)
1	30	–	0.3	30.0
2	40	30.0	0.3	33.0
3	35	33.0	0.3	33.6
4	60	33.6	0.3	41.5

Notice how the average moves toward the new angle but never jumps all the way there.

---

5. Tiny Java Snippet for Your Bot 🖥️

// Put this inside your PlayerBot class
private double smoothedAngle = 0;      // start at 0 or first scan
private static final double ALPHA = 0.2; // tweak 0.05–0.5 until it feels right
 
private void updateSmoothedAngle(double newAngle) {
    smoothedAngle = (1 - ALPHA) * smoothedAngle + ALPHA * newAngle;
}
 
// Use smoothedAngle instead of raw radar angle to draw your red dot

Where to call it: in your onScannedBot(ScannedBotEvent e) handler right after you read e.getDirection() (or bearing).

---

6. Choosing a Good Alpha 🎚️

• Small α (0.05–0.15) – smoother dot, but reacts slowly if the enemy zig‑zags.
• Medium α (0.2–0.3) – good balance for most targets.
• Large α (0.4–0.6) – follows quick turns, but the dot may shake more.

Try this: Add a GUI slider so kids can move α during the match and see the difference live!

---

7. Checkbox Connection ☑️

Your UI idea: “Exponential Averaging” checkbox

• Checked → Use smoothedAngle for the red dot.
• Unchecked → Use the raw lastScannedAngle so the dot shows exactly what the radar just saw.

---

8. Mini‑Challenges 🏆

1. Slider Test: Let α slide from 0.05 to 0.6 and watch how the dot behaves.
2. Dodge Bot: Fight a bot that swerves a lot. Which α keeps your aim most accurate?

Have fun experimenting and may your turret find its mark! 🎯