Feature Proposal: Input Method–Aware Personal Bests & Statistics

[MKishoreDev]

I've been using Monkeytype since around 2021, although I only recently created an account because I wanted to start tracking my progress more seriously instead of typing as a guest.

First of all, thank you for building such an amazing project. I've tried many typing websites over the years, but Monkeytype has always been the one I return to because of its clean design, performance, and detailed statistics.

While tracking my own progress, I noticed something that I think could make statistics even more meaningful without changing the core experience.

---

💡 Proposal

Separate Personal Bests and Statistics by input method rather than combining all sessions into a single record.

For example:

• ⌨️ Physical Keyboard
• 📱 Touchscreen
• 🌍 Overall (optional)

---

Why?

Many users—including myself—practice on both desktop and mobile depending on where they are.

Although the measured metric (WPM) remains the same, typing on a physical keyboard and typing on a touchscreen are fundamentally different input methods. They involve different ergonomics, muscle memory, and typing behavior.

Currently, all sessions contribute to the same personal bests and statistics.

Because of that, it's difficult to answer questions like:

• What's my fastest keyboard score?
• How much have I improved on mobile?
• Am I progressing equally on both input methods?

Separating these statistics would provide a clearer picture of personal progress without changing gameplay or competition.

---

Example

Instead of showing only:

Personal Best

60s — 103 WPM

it could optionally become:

⌨️ Keyboard

15s — 118 WPM
60s — 103 WPM

📱 Touchscreen

15s — 84 WPM
60s — 76 WPM

🌍 Overall

118 WPM

Or simply provide a filter:

Statistics

• All
• Keyboard
• Touchscreen

---

Benefits

✅ Better long-term progress tracking

✅ More meaningful personal bests

✅ Easier to compare improvement across different input methods

✅ Doesn't affect existing leaderboards or competitions

✅ Completely optional for users who don't switch devices

---

Possible Detection

I understand that perfectly identifying devices isn't always possible because of hybrid devices, tablets with external keyboards, convertible laptops, and similar setups.

Instead of classifying based purely on the device, sessions could be categorized by the active input method whenever it's reliably available:

• Keyboard
• Touchscreen
• Unknown (fallback)

This would avoid making assumptions while still improving statistics for most users.

---

Mockup

The mockup is only meant to illustrate the concept—not the final UI or design.

---

Why I believe this fits Monkeytype

One of Monkeytype's biggest strengths is how detailed and motivating its statistics are.

This proposal doesn't introduce new game modes or separate communities. Instead, it builds on the existing statistics system by making personal progress more representative for users who regularly switch between keyboards and touchscreens.

---

I'd love to hear what the community thinks and whether there are technical considerations or edge cases I may have overlooked.

Thanks again for creating and maintaining such an amazing project. ❤️

[fehmer]

hi @MKishoreDev,

you can use tags to tag your tests. You could add one for desktop/mobile. Active tags are not synced between devices so you only have to do this once. On the account page under filters, advanced you can filter by tag and see the statistics for desktop/mobile.

[MKishoreDev]

Thanks for pointing that out! I actually wasn't aware of the tag filtering feature.

That definitely works as a manual solution, and I'll probably start using it.

The reason I suggested this as a built-in feature is mostly from a usability perspective. New users (or even long-time users like me) may never discover tags, and it also relies on remembering to create and apply them consistently.

My thought was that if Monkeytype can reliably detect the active input method (or fall back to "Unknown" when it can't), users would automatically get separate statistics without any setup. It would make the feature more discoverable and eliminate the possibility of forgetting to tag a session.

That said, I completely understand there may be technical limitations or reasons why the current tag-based approach is preferred. Thanks for letting me know about it!