How the Value Engine Works
A clear, high-level explanation of how The Comparator calculates Value Scores ($0–100$) based on performance, specs, and real market prices.
How the Value Engine Works
Unlike traditional hardware sites that rely on static editorial scores, The Comparator calculates a live Value Score ($0 - 100$) for every graphics card, processor, SSD, and RAM kit.
1. The Core Concept
The Value Engine compares what you get (performance, VRAM, warranty) against what you pay (price, power draw, electricity cost).
$$\text{Value Score} = \frac{\text{Performance \& Specs (Good Attributes)}}{\text{Price \& TDP (Cost Attributes)}}$$- Desirable Attributes (Numerator):
- Raster Performance: Benchmark gaming performance index (RTX 4070 = 1000 points baseline).
- Ray Tracing Index: Performance in RT-heavy titles.
- AI & Compute Index: Tensor TFLOPS for Machine Learning and AI workloads.
- VRAM Capacity: Dedicated memory (8GB, 12GB, 16GB+).
- Warranty & Recency: Manufacturer warranty years and release year factor.
- Cost Attributes (Denominator):
- Total Price: Store retail price + shipping costs.
- TCO Electricity (Optional): Estimated 3-year power cost based on component TDP and $0.15/kWh default rate.
2. Dynamic Score Normalization ($0 - 100$)
The Value Engine normalizes raw scores across all currently visible cards in your filter view:
- 🟢 Score 100: The absolute best value deal in your current search.
- 🟡 Score 70-89: Great deal with solid price-to-performance ratio.
- 🔴 Score < 50: Overpriced relative to available market alternatives.
3. Presets & Weight Adjustment
Adjusting the weight sliders on the calculator changes how much importance the algorithm places on each attribute.
- Moving Price to 100% prioritizes cheap budget deals.
- Moving VRAM or Ray Tracing up prioritizes high-end 4K cards.
- Selecting a Preset (e.g. 1440p Sweet Spot, AI / LLM, Everyday) instantly configures the optimal weights for that workload.