Rating something on an arbitrary five-star scale is hard - what happens when you find something better? Do you have to readjust your old ratings?
Pairwise comparisons avoid the ambiguity of absolute ratings and create a dynamic ranking that adjusts as you add more comparisons. Just ask yourself "which of these two notes wins?" and the rest is calculated for you.
_Other comparison arena UI options are available in Settings - shown here is **right-split**, useful if you'd like to watch your Base update as your ranks change, but I personally use **new window/popout**_. Mobile devices are supported by an additional single-note layout type for phones with smaller screens.
- The plugin opens **two notes side-by-side** and you choose a winner.
- The plugin (optionally) writes stats like **Rank** (1 = best in cohort), **Rating** , or **Wins** to frontmatter (property names are configurable).
- Your Bases can then sort or filter by those properties, so you can rank notes by subjective criteria without having to decide what star rating a note "deserves".
Saved cohorts can be renamed and reconfigured in Settings. If the folder or Base that defines a cohort is later moved or renamed, the plugin will prompt you to point it to the new location to safely migrate your saved ratings.
Two notes open side-by-side in Reading mode for you to compare. Use the arrow keys on your keyboard or the buttons on the session bar to choose a winner.
Press Escape or run **"Pairwise Glicko Ranking: End current session"**. If you've enabled a Rank property for this cohort, the plugin recomputes ranks across the cohort and writes them to frontmatter.
Optionally, the plugin can generate a **post-session report** - a Markdown note summarising the session (biggest gains, losses, surprises, leaderboard snapshot, match log, and more). Reports use a template with `{{glicko:...}}` placeholders that you can customise, or you can use the built-in default. Run the command **"Generate example report template"** to create a starting template you can edit. Report generation is configurable per-cohort and globally in Settings.
You can then use the values computed by the plugin however you want. For example, enable the Rank property and then sort a Base by it to see your notes in ranked order, or use the Glicko rating property to filter for your best notes with a rating above a certain threshold.
- Post-session reports: enable/disable, report folder, file name template, and custom report template path
- Cohorts section: rename a cohort and change its frontmatter, scroll, and report overrides. The plugin can preview and perform bulk updates (write, rename, remove) across the cohort.
## Mobile support
The plugin is fully functional on mobile devices, with support for all desktop layouts on tablets (except popout/new window), and an additional single-leaf layout for phones that allows comparison between two notes using a switch button.
If you've made it this far: start a session, pick a small cohort, and do a dozen comparisons. You'll be surprised how quickly a meaningful order appears.
\*Disclaimer: I have not used this plugin to end any friendships. If you try it out, please let me know how it goes! Although I accept no responsibility for any fallout. :)