Hey PH! I'm C. W. Chen, indie developer. Quick story behind this 👇 A friend was convinced his DualSense was drifting. He "tested" it in three different games over a weekend and got three different answers — because every game applies its own software deadzone on top of the raw stick values. The controller never changed; the software lens did. That's the problem GearLint works on: read the raw numbers straight from the browser (Gamepad / KeyboardEvent / PointerEvent APIs), and end every test with a plain verdict — fix it, replace it, or return it — with the threshold shown, not hidden. Three things I deliberately did differently: 1. Raw numbers over pass/fail badges. 0.02–0.04 of resting offset is normal electrical noise; drift is a value that stays above ~0.05 and creeps upward. The test tells you which one you're looking at. 2. Honest about limits. A browser reads what the OS reports, not the potentiometer itself — the site says so on every page. It's a behavioral measurement, and for fix/replace/return decisions that's enough. 3. Copyable reports. Result cards have a one-click copy that formats the reading for a warranty or return claim. The open-box test most people skip is now a 30-second screenshot. Everything runs locally — no uploads, no accounts, no telemetry. It's free and stays free; the site carries ads, that's the whole business model. What I'd love feedback on: - The drift verdict thresholds (0.08 / 0.15) — do they match what you've seen on real hardware? - The mouse double-click grouping logic — does it match what flaky switches actually do? - Anything missing that would make the "evidence" usable in a real RMA fight? Ask me anything about the build. Stack is React + Vite, prerendered, static on Vercel.

GearLint

Is your controller drifting, or is it just the game?

Is your controller drifting, or is it just the game?