Adaptive Urdu Typing Game That Trains Your Weakest Keys First

The problem: Urdu typing is a job skill nobody teaches well

Urdu typing is needed in many jobs: newsrooms, publishing, court and government offices, schools, NGOs, content writing and social media teams. Many people who need it learn by hunting for keys, then plateau at a slow speed with frequent errors. Some teams still depend on a single fast typist, and everyone else waits for them.

Learning is harder than in English for practical reasons. There are several keyboard layouts in use, including phonetic layouts that map Urdu letters to similar-sounding English keys and the standard layout that ships with Windows. Many letters look alike, some share keys with a modifier, and the same letter changes shape when joined, so beginners struggle to check what they typed.

What already exists and the gap

• Keybr is a well-known adaptive typing trainer. It starts with a few common letters and only unlocks new keys when your speed and accuracy on the current set are good enough, and it generates practice words that focus on your weakest letters.
• Monkeytype, Typing.com and similar sites offer typing tests and lessons, mostly designed around Latin-script keyboards.
• Urdu typing tutors exist as simple practice pages and old desktop programs, usually with fixed lessons and little feedback.

The gap is an adaptive trainer built for Urdu from the start: real Urdu words, correct Nastaliq rendering, support for the layouts people actually use, and feedback about the confusions specific to Urdu, wrapped in a game that keeps people practising.

Inspiration

Keybr's approach is the inspiration: each key has a confidence score from your recent speed and accuracy, new keys are introduced only when the current set is mastered, and the generated text over-represents weak keys. Your project applies the same idea to Urdu, where the unit of difficulty is not only a key but also letter pairs and modifier combinations, and measures whether adaptive practice beats fixed lessons for Urdu learners.

What you build

• Learners choose their keyboard layout, take a short placement test, then practise in short levels. Each level targets their current weak keys and pairs.
• Teachers and trainers (stretch goal) run a classroom mode and see which keys a class struggles with.
• Job seekers use the timed challenge to measure their words per minute and accuracy.

The technically interesting part

Even as a beginner project, there is real engineering here:

• Modelling skill per key. Track speed and error rate per key with a moving average, and decide when a key counts as learned. Extend this to letter pairs, because many errors happen at specific transitions.
• Confusions. Record which wrong key was pressed for each target. A confusion matrix shows patterns such as similar-looking letters or neighbouring keys, which you can then target.
• Generating good practice text. Build a word list from openly licensed Urdu text, then select words that contain the target keys while staying readable. Avoid nonsense strings, which do not transfer to real typing.
• Rendering and input. Right-to-left input, joined letter forms and diacritics must display correctly. Compare the typed sequence with the target at the character level, not the visual level.
• Measuring words per minute fairly. Define a standard word length for Urdu so results are comparable across texts.

How to evaluate it

• Learning gain: ten or more learners practise for two weeks; half use adaptive lessons and half fixed lessons. Compare speed and accuracy before and after.
• Targeting: check that weak keys at the start improve more than strong keys.
• Text quality: ask two Urdu readers to rate a sample of generated practice text for naturalness.
• Rendering: test on Chrome, Firefox and an Android browser with a set of tricky letter combinations.

Suggested plan

• Weeks 1 to 2: study Keybr and Urdu keyboard layouts, collect openly licensed Urdu text.
• Weeks 3 to 6: typing engine, layouts, rendering and per-key statistics.
• Weeks 7 to 9: adaptive lesson generator and practice text selection.
• Weeks 10 to 12: game layer, timed challenge and local progress storage.
• Weeks 13 to 15: the two-week study, write-up and demo.

Path to a product

Training centres, vocational programmes and employers who hire Urdu typists need a way to train and test people. A free web app with paid classroom and certification features is a realistic model, and the same engine can support other right-to-left languages. The project shows employers that you can build careful, measured software for a real audience.

Scope warning

Start with one layout and get it right before adding the second. Do not build user accounts or a backend in the first version; local storage is enough. The adaptive engine is the heart of the project, so build it before the game layer, not after.

Learn what you need on Sefism

• Frontend development roadmap: input handling, rendering and local storage.
• Programming foundations roadmap: the data structures behind per-key statistics.
• Game development roadmap: levels, streaks and keeping players motivated.
• JavaScript.info: keyboard events and the browser APIs you will use.
• What projects should I build in my first year of university?: why small, measured projects like this one stand out.

Get help with this idea

The risk is a typing game that looks fine but does not adapt in any meaningful way. Tauseef can help you design the skill model, choose the layouts and plan the two-week study. If you are new here, create a free account first.

• Discuss this idea with a mentor
• Send a query to Tauseef
• Book a 1:1 session