Round #652

10 May 2026, 08:30· csharp· qwen2.5-coder-7b-instruct· r13-clean-hints
Complexity
band 0
ZP / UO / OBS
3 / 4 / 0
Journey weight
402
Lightshift ΔL

Prompt

Write a SQL query that returns the top 5 customers by total spend in 2025.

Zero-point index (what the engine surfaced)

ZPI-ZPI-0001 | 2026-05-10T08:30:56Z Write a SQL query that returns the top 5 customers by total spend in 2025. ZERO POINT — 3 entries (structural invariants) ● hsl(27.69, 100%, 50%) async [feature] ● hsl(145.64, 100%, 50%) Task of T [feature] ● hsl(348.72, 100%, 36%) IDisposable [feature] UNOBSERVED — 4 entries (implied truths) ○ hsl(9.23, 100%, 50%) interface [feature] ○ hsl(180.92, 100%, 28%) Database health check [function] ○ hsl(144.00, 100%, 42%) Connection resilience [function] ○ hsl(83.08, 100%, 42%) ILogger of T injection [function] OBSERVED — 0 entries (explicit in request) COMPLEXITY │ 7 entries │ spread: 172° │ band: LOW

Baseline response (no ZPI)

To write an SQL query that returns the top 5 customers by total spend in 2025, we need to assume some table structures. Let's assume you have two tables: 1. `Customers` table with columns: `CustomerID`, `CustomerName`. 2. `Orders` table with columns: `OrderID`, `CustomerID`, `OrderDate`, `TotalAmount`. Here is the SQL query that will give you the top 5 customers by total spend in 2025: ```sql SELECT c.CustomerID, c.CustomerName, SUM(o.TotalAmount) AS TotalSpend FROM Customers c JOIN Orders o ON c.CustomerID = o.CustomerID WHERE YEAR(o.OrderDate) = 2025 GROUP BY c.CustomerID, c.CustomerName ORDER BY TotalSpend DESC LIMIT 5; ``` ### Explanation: 1. **SELECT**: We select the `CustomerID`, `CustomerName`, and the sum of `TotalAmount` as `TotalSpend`. 2. **FROM Customers c**: We start from the `Customers` table and alias it as `c`. 3. **JOIN Orders o ON c.CustomerID = o.CustomerID**: We join the `Orders` table (aliased as `o`) on the condition that the `CustomerID` matches. 4. **WHERE YEAR(o.OrderDate) = 2025**: We filter the orders to include only those from the year 2025. 5. **GROUP BY c.CustomerID, c.CustomerName**: We group the results by `CustomerID` and `CustomerName` so that we can calculate the total spend per customer. 6. **ORDER BY TotalSpend DESC**: We order the results in descending order based on the total spend. 7. **LIMIT 5**: We limit the result to the top 5 customers. Make sure your database system supports the `YEAR()` function and `LIMIT` clause, as these are standard SQL functions but may have slight variations depending on the specific database management system (e.g., MySQL, PostgreSQL, SQL Server).

Enhanced response (with ZPI)

To implement the request, we'll write a SQL query to return the top 5 customers by their total spend in the year 2025. We'll assume there is a database table named `Orders` with columns `CustomerID`, `OrderDate`, and `TotalAmount`. Here's the SQL query: ```sql SELECT CustomerID, SUM(TotalAmount) AS TotalSpend FROM Orders WHERE YEAR(OrderDate) = 2025 GROUP BY CustomerID ORDER BY TotalSpend DESC LIMIT 5; ``` Explanation: 1. **SELECT CustomerID, SUM(TotalAmount) AS TotalSpend**: This selects the `CustomerID` and calculates the total spend for each customer. 2. **FROM Orders**: Specifies the table from which to retrieve data. 3. **WHERE YEAR(OrderDate) = 2025**: Filters the orders to include only those from the year 2025. 4. **GROUP BY CustomerID**: Groups the results by `CustomerID` so that the total spend is calculated for each customer. 5. **ORDER BY TotalSpend DESC**: Orders the results in descending order based on the total spend. 6. **LIMIT 5**: Limits the result to the top 5 customers. This query will return the top 5 customers who spent the most in the year 2025, along with their total spend.

Grader chain

Each grader reads the same baseline + enhanced pair and casts an independent verdict. Disagreement is the framework's own signal — not noise to be averaged away.

V1samba/Meta-Llama-3.3-70B-Instruct
Worse
V4qwen2.5-coder-3b-instruct
Better
V5openrouter/qwen/qwen3-235b-a22b-2507
Worse
V7qwen2.5-coder-7b-instruct
Same
V8openai/gpt-4.1
Worse
V9anthropic/claude-opus-4-7
Worse
V12openai/gpt-4o
Worse

Comments

Our grader said what it said. What do you say? Comment as a guest below.

No comments yet. Be the first to say what you make of this round.

← back to all rounds