RandomSelection

A flexible .NET library for truly random selection with weighted entries. Perfect for raffles, lottery drawings, contests, and any scenario where you need fair, cryptographically-secure random selection.

Overview

The inspiration for this library came from observing bias in prize drawings where participants won multiple times while others were consistently overlooked. RandomSelection is designed to ensure fair, unbiased selection by:

• Accepting weighted entries - Give some participants higher chances than others without complex probability calculations
• Using cryptographic randomness - Utilizes System.Security.Cryptography for high-quality random numbers instead of weak pseudo-random generators
• Shuffling the selection pool - Randomizes the entire pool before selection to eliminate positional bias
• Supporting multiple selections - Select one or many winners in a single operation
• Type-safe and flexible - Generic Selector<T> works with any data type

Installation

Install via NuGet:

dotnet add package RandomSelection

Or via Package Manager Console:

Install-Package RandomSelection

Core Concepts

The Selector Class

The Selector<T> class is the main entry point. It's generic and can work with any type of value you want to track. It maintains a pool of items and provides methods to randomly select from them.

// Works with strings
var stringSelector = new Selector<string>();

// Works with custom objects
var objectSelector = new Selector<Employee>();

// Works with any type
var intSelector = new Selector<int>();

Entries (Weights)

Each item can have multiple "entries" in the selection pool. This creates a weighted random selection:

• An item with 1 entry has a baseline chance
• An item with 3 entries is 3x more likely to be selected
• An item with 10 entries is 10x more likely to be selected

Real-world example: In a loyalty rewards drawing where VIP members should win 4x more often:

- Regular member: 1 entry
- Silver member: 2 entries  
- Gold member: 3 entries
- Platinum member: 4 entries

Quick Start

Example 1: Simple Random Selection

Select one item from a group:

using BNolan.RandomSelection;

var selector = new Selector<string>();
selector.TryAddItem("emp001", "John Doe");
selector.TryAddItem("emp002", "Jane Smith");
selector.TryAddItem("emp003", "Mike Johnson");

// Select 1 random employee
var winners = selector.RandomSelect(1);
Console.WriteLine($"Winner: {winners[0].Value}");
// Output: Winner: Mike Johnson (example)

Example 2: Weighted Random Selection

Give some participants higher chances of winning:

var selector = new Selector<string>();

// Regular members: 1 entry each
selector.TryAddItem("jen", "Jennifer", 1);
selector.TryAddItem("michael", "Michael", 1);
selector.TryAddItem("dave", "David", 1);

// VIP member: 3 entries (3x more likely to win)
selector.TryAddItem("staci", "Staci", 3);

// Select 2 winners from the pool
var winners = selector.RandomSelect(2);

foreach (var winner in winners)
{
    Console.WriteLine($"Winner: {winner.Value}");
}

Example 3: Using Custom Objects

Work with any type of data, not just strings:

using BNolan.RandomSelection;
using BNolan.RandomSelection.Library;

public class Employee
{
    public string Id { get; set; }
    public string Name { get; set; }
    public string Department { get; set; }
    public decimal Salary { get; set; }
}

var selector = new Selector<Employee>();

var alice = new Employee { Id = "001", Name = "Alice", Department = "Engineering", Salary = 120000 };
var bob = new Employee { Id = "002", Name = "Bob", Department = "Sales", Salary = 90000 };

selector.TryAddItem("emp001", alice);
selector.TryAddItem("emp002", bob);

var winner = selector.RandomSelect(1);
Console.WriteLine($"Selected: {winner[0].Value.Name} from {winner[0].Value.Department}");

Example 4: Using the Item Class

For explicit control over all properties, use the Item<T> class:

using BNolan.RandomSelection.Library;

var selector = new Selector<string>();

var item1 = new Item<string>
{
    UniqueId = "ticket001",
    Value = "John Doe",
    Entries = 2
};

var item2 = new Item<string>
{
    UniqueId = "ticket002",
    Value = "Jane Smith",
    Entries = 1
};

selector.TryAddItem(item1);
selector.TryAddItem(item2);

var winner = selector.RandomSelect();
Console.WriteLine($"Winner: {winner[0].Value}");

Example 5: Bulk Operations

Add multiple items efficiently in one operation:

var employees = new List<string> { "Alice", "Bob", "Charlie", "Diana" };

var selector = new Selector<string>();
if (selector.TryAddItems(employees))
{
    Console.WriteLine("All employees added successfully");
    var selected = selector.RandomSelect(2);
}
else
{
    Console.WriteLine("Failed to add employees");
}

API Reference

Adding Items

bool TryAddItem(T uniqueId)

Adds an item with just a unique ID. The value defaults to the uniqueId, and entries default to 1.

bool success = selector.TryAddItem("participant1");

Parameter	Type	Description
uniqueId	T	Unique identifier (case-insensitive)
Returns	bool	true if added, false if ID already exists

---

bool TryAddItem(string uniqueId, T value)

Adds an item with a unique ID and associated value. Entries default to 1.

bool success = selector.TryAddItem("emp001", "John Doe");
bool success = selector.TryAddItem("ticket123", myCustomObject);

Parameter	Type	Description
uniqueId	string	Unique identifier (case-insensitive)
value	T	Associated value of any type
Returns	bool	true if added, false if ID already exists

---

bool TryAddItem(string uniqueId, T value, int entries)

Adds an item with a unique ID, value, and number of entries (weight).

// Regular member: 1 entry
selector.TryAddItem("user001", "Regular User", 1);

// VIP member: 5x more likely to be selected
selector.TryAddItem("user002", "VIP Member", 5);

Parameter	Type	Description
uniqueId	string	Unique identifier (case-insensitive)
value	T	Associated value of any type
entries	int	Number of entries/weight (must be > 0)
Returns	bool	true if added, false if ID already exists
Throws	ArgumentException	If entries < 1

---

bool TryAddItem(Item<T> item)

Adds an Item<T> object directly for maximum control.

var item = new Item<string> 
{ 
    UniqueId = "id1", 
    Value = "My Value", 
    Entries = 3 
};

if (selector.TryAddItem(item))
{
    Console.WriteLine("Item added successfully");
}

Parameter	Type	Description
item	Item	Item object with UniqueId, Value, and Entries
Returns	bool	true if added, false if ID already exists
Throws	ArgumentNullException	If item is null or UniqueId is null/empty
Throws	ArgumentException	If Entries < 1

---

bool TryAddItems(List<T> items)

Adds multiple items at once. This is an all-or-nothing operation: if any item already exists, no items are added.

var employees = new List<string> { "Alice", "Bob", "Charlie" };

if (selector.TryAddItems(employees))
{
    Console.WriteLine("All employees added");
}
else
{
    Console.WriteLine("Failed: at least one employee already exists");
}

Parameter	Type	Description
items	List	Items to add
Returns	bool	true if all items added, false if any already exist
Throws	ArgumentNullException	If items list is null

---

Selection Methods

List<Item<T>> RandomSelect(int numToSelect = 1)

Randomly selects items from the pool. Each item can only be selected once per call.

// Select 1 winner (default)
var winner = selector.RandomSelect();

// Select 3 winners
var topThree = selector.RandomSelect(3);

foreach (var item in topThree)
{
    Console.WriteLine($"ID: {item.UniqueId}, Value: {item.Value}");
}

Parameter	Type	Description
numToSelect	int	Number of items to select (default: 1)
Returns	List<Item>	List of selected Item objects

Return Structure:

List<Item<T>>
{
    new Item<T> { UniqueId = "id1", Value = yourValue, Entries = 1 },
    new Item<T> { UniqueId = "id2", Value = yourValue, Entries = 3 }
}

---

Utility Methods

List<string> GenerateList()

Creates the internal pool list with items repeated according to their entry count. Useful for debugging or understanding how the pool is constructed.

selector.TryAddItem("id1", "value1", 2);
selector.TryAddItem("id2", "value2", 1);

var pool = selector.GenerateList();
// Result: ["id1", "id1", "id2"]

Console.WriteLine($"Pool size: {pool.Count}");

| Returns | List | List of unique IDs, each repeated by its entry count |

---

List<string> RandomizeList(List<string> items)

Shuffles a list using cryptographic randomness. Can be used independently of selection.

var original = new List<string> { "a", "b", "c", "d", "e" };
var shuffled = selector.RandomizeList(original);

// original remains unchanged
// shuffled contains a random permutation of the items

Parameter	Type	Description
items	List	List to shuffle
Returns	List	New shuffled list (original is not modified)

---

int GenerateRandomIndex(int upperLimit)

Generates a random index between 0 and upperLimit (exclusive). Uses cryptographic randomness and can be used independently.

int randomIndex = selector.GenerateRandomIndex(100);
// Returns a random number from 0 to 99

int diceRoll = selector.GenerateRandomIndex(6);
// Simulates a die roll (0-5)

Parameter	Type	Description
upperLimit	int	Upper bound (exclusive)
Returns	int	Random integer from 0 to upperLimit - 1
Throws	ArgumentOutOfRangeException	If upperLimit < 1

---

Error Handling

All add methods validate input and throw appropriate exceptions. Always handle these when accepting user input:

try
{
    selector.TryAddItem((Item<string>)null);  // ArgumentNullException
}
catch (ArgumentNullException ex)
{
    Console.WriteLine($"Validation error: {ex.ParamName}");
}

try
{
    selector.TryAddItem("id1", "value", 0);  // ArgumentException
}
catch (ArgumentException ex)
{
    Console.WriteLine($"Invalid argument: {ex.ParamName}");
}

Common Exceptions:

• ArgumentNullException - When item or UniqueId is null/empty
• ArgumentException - When entries < 1
• ArgumentOutOfRangeException - When GenerateRandomIndex called with upperLimit < 1

---

Why Cryptographic Randomness?

This library uses System.Security.Cryptography instead of System.Random for several important reasons:

Feature	System.Random	Cryptographic RNG
Predictable	Yes (if seed known)	No - truly random
Suitable for fairness	No	Yes ?
Passes statistical tests	No	Yes ?
Seed bias	Yes	No ?
Reproducible	Yes (sometimes a problem)	No

For raffles and contests where fairness matters, cryptographic randomness is essential.

---

Performance Considerations

• First selection: Building the pool is O(n) where n = total number of entries. This happens once per RandomSelect() call.
• Shuffling: Uses Fisher-Yates algorithm, O(n) complexity
• Cryptographic calls: Each random index requires at least one cryptographic RNG call
• Repeated selections: Reuse the same Selector instance for multiple selections to avoid rebuilding the pool

Optimization Tip: For large selections, it's more efficient to call RandomSelect(10) once rather than RandomSelect(1) ten times.

---

Common Use Cases

Raffle/Lottery Drawing

var selector = new Selector<string>();

// Add raffle participants
selector.TryAddItem("ticket001", "John Smith");
selector.TryAddItem("ticket002", "Jane Doe");
selector.TryAddItem("ticket003", "Bob Johnson");
selector.TryAddItem("ticket004", "Alice Williams");

// Draw 3 winners
var winners = selector.RandomSelect(3);

foreach (var winner in winners)
{
    Console.WriteLine($"Winner: {winner.Value}");
}

Contest with VIP Privileges

var selector = new Selector<string>();

// Regular participants (1x chance)
selector.TryAddItem("regular001", "Regular User 1", 1);
selector.TryAddItem("regular002", "Regular User 2", 1);

// VIP gets enhanced odds
selector.TryAddItem("vip001", "Premium User", 5);

var winner = selector.RandomSelect(1);
Console.WriteLine($"Contest winner: {winner[0].Value}");

Team Assignment Rotation

var employees = new List<(string id, string name, string role)>
{
    ("emp001", "Alice Johnson", "Developer"),
    ("emp002", "Bob Smith", "Designer"),
    ("emp003", "Charlie Brown", "QA"),
    ("emp004", "Diana Prince", "Manager")
};

var selector = new Selector<(string, string)>();
foreach (var emp in employees)
{
    selector.TryAddItem(emp.id, (emp.name, emp.role));
}

var assignedEmployee = selector.RandomSelect(1)[0];
Console.WriteLine($"On-call: {assignedEmployee.Value.Item1} ({assignedEmployee.Value.Item2})");

Lottery with Weighted Chances

public class LotteryPlayer
{
    public string Name { get; set; }
    public int TicketsPurchased { get; set; }
}

var selector = new Selector<LotteryPlayer>();

selector.TryAddItem("player001", new LotteryPlayer { Name = "Alice", TicketsPurchased = 5 }, 5);
selector.TryAddItem("player002", new LotteryPlayer { Name = "Bob", TicketsPurchased = 3 }, 3);
selector.TryAddItem("player003", new LotteryPlayer { Name = "Charlie", TicketsPurchased = 10 }, 10);

// Winners weighted by ticket count
var winner = selector.RandomSelect(1);
Console.WriteLine($"Jackpot winner: {winner[0].Value.Name}");

---

Features at a Glance

? Weighted selection - Control probability with entry counts
? Cryptographically secure - Uses System.Security.Cryptography
? Generic support - Works with any data type (T)
? Bulk operations - Add and select multiple items efficiently
? No duplicates - Each selection returns unique winners
? Comprehensive validation - Clear error messages
? Thoroughly tested - Includes statistical randomness tests
? Easy API - Intuitive methods with sensible defaults

---

Testing

The library includes comprehensive unit tests covering:

• Basic functionality (adding items, selection)
• Error handling and validation
• Statistical randomness verification
• Edge cases (single item, all items, duplicates)
• Distribution uniformity tests

Run tests with:

dotnet test

---

License

MIT License - See repository for details