Subscore

DiscoveryTesting methodology

Discovery measures how easy it is to find a character you actually want to talk to.

A big character library is not very useful when you need to scroll through hundreds of profiles to find someone you like. We test the filters, categories, search tools, and general browsing experience.

4 evidence groups
4 scored tests

Discovery is organized into 4 evidence groups. Each group contains one scored test: Filters, Categories, Search, and Browsing. The four tests count equally toward the final Discovery score.

How the Discovery score is calculated

Every scored test gets a score from 0 to 10.

All four tests count equally. We combine their scores to calculate the final Discovery score. Discovery makes up 33% of the Characters score.

25%Filters
25%Categories
25%Search
25%Browsing

Weighted evidence groups (combined 100%)Discovery score33% of Characters score

View exact calculation

Each test gets a score from 0 to 10. We multiply every score by 25% and then add the points together.

Discovery makes up 33% of the Characters score. Characters makes up 10% of the overall performance score.

Scored tests and weights

Scored test	How much it counts	See scoring
Filters	25.00%	View
Categories	25.00%	View
Search	25.00%	View
Browsing	25.00%	View
Total	100%

How the score is calculated

Each test gets a score from 0–10
Test score × how much it counts
Example: 8 × 25.00% = 2.00 points
We do this for every test
We add all the points together
Final Discovery score
Counts for 33% of Characters

Example calculation

We multiply each test score by how much it counts. We then add all the points together.

Scored test	Test score	How much it counts	Calculation	Points added
Filters	8.00	25.00%	8.00 × 25.00%	2.00
Categories	8.00	25.00%	8.00 × 25.00%	2.00
Search	10.00	25.00%	10.00 × 25.00%	2.50
Browsing	8.00	25.00%	8.00 × 25.00%	2.00
Final Discovery score	—	100%	Add all points	8.50/10

Special cases

Not Applicable

If a test does not apply, we remove it and spread its weight across the remaining tests.

Unknown

If we cannot verify a result, the test receives a score of 0.

Manual adjustment

In rare cases, we may adjust a score when the calculated result is clearly unfair. We always record the reason.

Why Discovery matters

A huge character library can quickly become annoying when there is no easy way to explore it.

You might know that you want an anime character, a specific personality, or a certain type of roleplay. Without useful filters, categories, or search, you may need to open dozens of profiles before finding someone suitable.

This becomes an even bigger problem when the app keeps adding new characters. More choice should make the app better, not make it harder to find what you want.

Good Discovery tools save time and help you find characters you might otherwise miss.

What good Discovery looks like

A high Discovery score means the library is easy to explore, the filters and categories are useful, and search works for both character names and general keywords. A lower score means you need to scroll through a large number of profiles, the search tool barely works, or the available filters do not help narrow the results.

Main benchmarks

GoodTypicalWeak

Useful filters

7+4–60–3

Useful categories

7+4–60–3

Names and keywords workSome searches failNo useful search

Minimum useful standard

Filters: Actually narrow the results
Categories: Lead to different groups of characters
Search: Finds existing names and general keywords
Browsing: Most tasks can be completed smoothly

Common red flags

Search only works when you enter an exact character name
Filters are available but barely change the results
Several categories show almost the same characters
A huge library with no useful way to narrow it down

How we test Discovery

We use a paid account and open the full ready-made character library available during testing.

First, we count every useful filter and category. Sorting options such as Newest and Most Popular are not counted as filters.

We then test the search tool using three existing character names and three general keywords.

Finally, we complete the 10 fixed browsing tasks from our testing guide. We record which tasks are easy, which take too many steps, and which cannot be completed.

We use the normal tools available to users. We do not use admin access, direct profile links, or outside search engines.

Limitations and edge cases

Character libraries change regularly. Platforms may add new filters, remove categories, or update their search tool after we finish testing.

Discovery tools can also differ between devices. A filter may be available on desktop but missing from the mobile app.

Some platforms personalize their library based on your previous activity. This means two users may see different characters or categories.

Our results reflect the paid account, device, and settings used on the recorded test date.

Evidence groups

Discovery has 4 evidence groups made up of 4 scored tests.

Filters measure how many useful controls help you narrow the character library.

A good set of filters lets you quickly remove characters that do not match what you are looking for.

Filters

25%of Filters (25% total)

The number of useful filters available in the character library.

How we test

We open the character library and count every filter that meaningfully narrows the available characters.

What counts

Gender filters
Visual-style filters
Personality filters
Relationship or scenario filters
Ethnicity filters
Other controls that narrow the character results

What does not count

Sorting by newest or most popular
Buttons that only change the order of the results
Two filters that do the same thing
Filters that do not actually change the results

Result shown

7 useful filters

Scoring

More useful filters means a higher score. We use the ranges below.

Useful filters	Score
0 or fewer	0/10
1–3	4/10
4–6	6/10
7–10	8/10
11+	10/10

A result of 7 useful filters scores 8/10.

Evidence group 2 of 4

Categories measure how many useful groups help users browse different types of characters.

Categories are different from filters. A category normally opens a ready-made group, while a filter lets you narrow the current results.

Search

25%of Search (25% total)

Whether the character search works for existing names and general keywords.

How we test

We search for three existing character names and three general keywords. The keywords describe a character type, personality, style, or scenario rather than an exact profile name.

What counts

A search tool inside the character library
Search results for existing character names
Search results for general keywords
Partial-name search when it returns the correct character

What does not count

The browser’s page-search tool
A site-wide search that does not search the character library
Filters without a search box
Search suggestions that cannot be opened

Result shown

Yes — all 3 names and 3 keywords worked

Scoring

We use the result categories below.

Search result	Score
Yes — all six searches worked	10/10
Limited — some searches failed or only names worked	5/10
No useful character search	0/10
Unknown	0/10

In this example, all six searches worked, so Search scores 10/10.

Evidence group 4 of 4

Browsing measures how easy the character library feels to use as a whole.

Filters, categories, and search may all exist, but the library can still feel confusing when buttons are hard to find, results reset unexpectedly, or profiles take too many steps to open.

Browsing

25%of Browsing (25% total)

How easy it is to find and open suitable characters during normal use.

How we test

We complete the 10 fixed browsing tasks in our testing guide. We record whether each task works smoothly, takes unnecessary extra steps, feels confusing, or cannot be completed.

What counts

Browsing through the normal character library
Using available filters, categories, and search
Opening profiles and returning to the results
Changing or clearing browsing choices

What does not count

Direct links to a character profile
Admin tools
Browser search
Outside search engines
Features unavailable to normal users

Result shown

Easy overall — most tasks completed smoothly

Scoring

Browsing receives a score from 0 to 10 based on the complete test.

More completed tasks and fewer confusing moments lead to a higher score. We also save notes explaining what worked well and what caused problems. In this example, Browsing scores 8/10.

Back to Characters