On Farkle Tournament Scoring Systems

Discussion on BBS

After the above proposal I have decided to write up a clear outline of some tournament structures and their differences.

3-2-1 System

Advantages

• Reduces the point gap between 1st and 3rd place, making it easier to catch up
• Still provides clear differentiation between top players

Disadvantages:

• Might not sufficiently reward high scores
• Could lead to more frequent ties in the overall tournament results

3-2-1-1 System (for top four places)

Advantages:

• More players have a chance to score

Disadvantages:

• Reduces the reward for first place
• With 10 players, nearly half the field scores points each day, which might not create enough separation

5-3-1 for top half (5 for 1st, 3 for 2nd, 1 for the rest of the top half)

Advantages:

• Creates a good balance between rewarding top performers and keeping more players engaged
• More players can consistently get some points and therefore not fall behind as easily

Disadvantages:

• Might not provide enough distinction between 3rd and lower place finishers
• Could still lead to runaway leaders if the same players consistently finish in the top two

Now, let's explore some alternative systems:

Logarithmic Point System

Award points based on the logarithm of the player's daily rank:

1st: 10 points

2nd: 7 points

3rd: 5 points

4th: 4 points

5th: 3 points

6th-10th: 2 points

Advantages:

• Rewards first place while still keeping lower scoring players in the running

Disadvantages:

• Slightly more complex to explain
• Might still allow for significant point gaps over time

Dynamic Point Allocation

The total points available each day are fixed (let's say 20 points), but they're distributed based on the spread of scores:

If scores are close (within 10% of each other), points are distributed more evenly (e.g., 6-5-4-3-2)

If there's a clear winner (>20% ahead of second place), more points go to the top (e.g., 10-5-3-2-0)

Advantages:

• Adapts to daily performance spread
• Rewards exceptional scores while keeping things competitive on close days

Disadvantages:

• More complex to implement and explain
• Less predictable for players

Here's an example dynamic point allocation formula implemented in python:

import numpy as np

def calculate_points(scores, total_points=20, min_points=1):
    n = len(scores)
    sorted_scores = sorted(scores, reverse=True)

    # Calculate the spread factor
    max_score = max(scores)
    min_score = min(scores)
    spread_factor = (max_score - min_score) / max_score if max_score > 0 else 0

    # Calculate base points
    base_points = np.linspace(total_points, min_points, n)

    # Apply spread factor to create more separation when scores are spread out
    adjusted_points = base_points ** (1 + spread_factor)

    # Normalize to ensure total points sum to the desired total
    normalized_points = (adjusted_points / adjusted_points.sum()) * total_points

    # Round points to integers
    final_points = np.round(normalized_points).astype(int)

    # Adjust to ensure we hit exactly total_points
    while final_points.sum() != total_points:
        if final_points.sum() < total_points:
            final_points[np.argmax(normalized_points - final_points)] += 1
        else:
            final_points[np.argmin(normalized_points - final_points)] -= 1

    return dict(zip(sorted_scores, final_points))

# Example usage
daily_scores = [8150, 6500, 5500, 5150, 5100, 4500, 4250, 3850, 3500, 3250]
point_allocation = calculate_points(daily_scores)
for score, points in point_allocation.items():
    print(f"Score: {score}, Points: {points}")