Most InstantGameZone editorial research is based on transparent synthesis of official sources, independent measurements, public data, and carefully classified community observations. We do not present external evidence as first-hand testing.
This page applies when InstantGameZone publishes a formal first-party benchmark. Performance numbers are only useful when the test conditions are clear, so formal benchmark records document the hardware, game, render configuration, test date, driver and operating-system context, run count, method, and measured results.
Measured results
Formal benchmark data is treated as measured data, not an estimate. Our current benchmark protocol is version IGZ-BP-0.1. The initial Windows workflow uses CapFrameX 1.9.0 and requires at least three valid measured runs for real production benchmark data.
- Warm-up activity is excluded from measured results.
- We do not select only the best run.
- Average FPS is the arithmetic mean across valid measured runs.
- 1% Low FPS is the arithmetic mean of each valid run’s CapFrameX 1% Low Average value.
- The benchmark configuration stays fixed across the runs being compared.
- Scene and environment conditions are kept as consistent as the test allows.
- Raw capture evidence is retained outside the public WordPress record for audit and review.
When a test changes
Changes that materially affect the meaning or comparability of results require a methodology review rather than silently reinterpreting older measurements. Articles should identify important settings and test conditions so readers can understand what a number represents.
Guides that are not formal benchmarks
Not every settings or troubleshooting article is a formal benchmark. When we make measured performance claims, we aim to make the evidence and conditions clear. Advice based on configuration, documentation, or observed troubleshooting is presented as such rather than being labeled as laboratory-style benchmark data.