Stop guessing where every action belongs.
BindIQ compares your current layout with your keyboard, mouse and reach preferences. It produces three explainable plans—never a mysterious “best bind.”
Describe the desk, not an imaginary average player.
This is an editable starter model, not an official game configuration.
Your local project library
Build the layout BindIQ will evaluate.
Add custom actions and relationships
Create a real action model for a game or mode that is not covered by the starter profile.
Teach the model what your hand can actually reach.
Select a band, then click keys or press physical keys. Calibration changes the scoring model; it does not diagnose ergonomics.
Browser-observed chord check
Select 2–6 assigned inputs and hold them together. BindIQ reports only what this browser received; it cannot certify the keyboard hardware.
Choose the decision you want BindIQ to protect.
Minimal change
Preserve learned binds unless a clear conflict or large cost justifies a move.
Competitive flow
Prioritize simultaneous actions, high-priority reach and rapid sequences.
Comfort first
Reduce stretch and repeated same-finger load, accepting more changes.
Three answers, with every trade-off exposed.
Why this changed
Compare every action across all three plans
| Action | Current | Minimal change | Competitive flow | Comfort first |
|---|
Learnability trial
A short browser-only recall test for changed binds. It measures observed response time and accuracy, not skill or in-game performance.
From a messy bind set to a decision you can test
BindIQ is most useful as a controlled three-pass workflow, not as a one-click promise.
Find the real collision
Map every action in its actual context. A reload/use collision in gameplay matters; the same input in mutually exclusive menus may not.
Protect muscle memory
Lock movement, jump or accessibility-critical binds before optimization. The engine must work around what you refuse to move.
Trial one plan
Compare the three strategies, run the recall check, export the old layout and validate the chosen plan in a training mode.
What BindIQ calculates—and what it cannot know
The engine uses physical key codes, a transparent keyboard coordinate model, action priority, same-context collisions, declared simultaneous/sequence relationships, your reach bands and the cost of changing learned binds. It does not measure your hands, reaction time, skill or a game’s hidden input processing.
Deterministic
The same project and strategy return the same plan. There is no generative AI or remote inference.
Hard constraints first
Locked binds, unavailable buttons and avoided inputs are handled before preference scoring. Every imported field is bounded and validated.
Verify in game
A high model fit is not a performance guarantee. Test one plan at a time in training and keep an export of the previous layout.
The public weight table documents how each strategy values relearning cost, modeled reach, same-finger load, action relationships, avoided inputs and conflicts. Scores compare plans only inside this model.
Questions before you move a learned bind
Does BindIQ change my game settings?
No. It cannot read or write installed game files. Apply recommendations manually in the game.
Is the highest score automatically best?
No. Scores compare plans inside this declared model. Comfort, accessibility and learned motor patterns still require an in-game trial.
Why are there three plans?
Keybind design has competing goals. Showing multiple strategies makes the trade-off visible instead of hiding it in one number.
