How New Characters Break Into the Rankings
A newly published character starts with no history, no favourites and no position, which makes the first week the hardest part of its life. The weekly chart holds 20 places and rebuilds every Thursday, so a new arrival has roughly seven days to show a favourite rate that can compete with cards sitting at 24.3M views and 7.1K favourites.
Key takeaways
- A new card competes on ratio, not volume: 40 early favourites carry the same signal as roughly 137,000 views at the top card's 0.029% conversion.
- The window is short — the list rebuilds every Thursday, so a launch is judged against a seven-day frame.
- A creator with an existing following supplies the first wave of favourites that an unknown name cannot.
- Tag choice decides whether search can find the card at all once the chart moves on.
The cold-start problem in one sentence
The ranking rewards a record of engagement, and a new character has no record. That single sentence explains almost every pattern in early catalogue life, because the weekly chart shows only 20 NSFW AI characters while the catalogue behind it keeps growing. The observer choosing between two unfamiliar cards has only the thumbnail, the tag row and the counters to work with, and two of those three are effectively empty for a fresh entry.
What a new card does have is a conversion ratio that can be measured from a very small sample. A card with 400 views and 12 favourites shows a rate of 3%, far above the 0.029% recorded on the current leading card. Small numbers make small samples, so the percentage is unstable, but it is the only positive signal available in week one and it is what a browsing reader will see next to the counter.
A second pattern matters to anyone watching a launch from outside. The chart mixes rate and volume in a way no single counter shows. A card can enter on a wave of views and disappear again within two weeks if its favourite rate is ordinary, or hold a place for months on a modest but steady stream of saves. Only both counters, read across consecutive Thursdays, separate those two tracks.
What early engagement actually looks like
The practical unit of early engagement is the favourite, because it costs the reader a deliberate action. Views follow attention; favourites follow a decision. A card that collects favourites faster than its view count would predict is being read closely, and that is the pattern a chart position eventually reflects.
Timing matters more than total. Because the list is rebuilt on a Thursday, a first-week surge that lands on a Friday has a full seven days to accumulate before the next rebuild, while one that lands on a Wednesday is being compared against a week it has barely entered. Nothing about the arithmetic is complicated here; it is simply that a fixed weekly window makes the day of a launch part of its outcome.
Three levers a creator controls
The first is the tag row. Tags are the vocabulary the catalogue is indexed with, and a card whose tags describe it precisely will keep surfacing in search long after its chart appearance has faded. The second is the opening description, which has to carry the premise in a line or two because that is all a scanning reader spends on it. The third is the creator name on the card. A creator with a track record starts with a queue of returning readers; a new name has to earn that queue, which is why the first card is always the expensive one.
What lies outside creator control is just as clear. A newcomer cannot manufacture a four-year accumulation of views, and it should not try to compete on that number. The competition is on the rate of deliberate saves, measured over a window nobody can extend.
| Input | What it decides | Figure or condition |
|---|---|---|
| Favourite rate in week one | Whether the card looks closely read | 12 favourites from 400 views is 3% against the top card's 0.029% |
| Day of launch | How much of the weekly window is usable | A Friday launch keeps seven days; a Wednesday launch keeps two |
| Tag precision | Whether search can still find the card later | Tags on ranked cards: Unfiltered, Submissive, Fictional, OC, Female |
| Creator recognition | Size of the first wave of returning readers | Card signatures are published as "by …" under each entry |
| Description length | How quickly a scanning reader understands the premise | Roughly 15 seconds is what a chart scan allows per card |
Worked example: what a first month has to produce
Illustrative assumption: a new card is opened by 400 readers in its first month, and 12 of them favourite it. Its ratio is 12 ÷ 400 = 3%. Compare that with the leading card's published counters, 7.1K favourites from 24.3M views, where 7,100 ÷ 24,300,000 is about 0.029%. The new card's rate is roughly 100 times higher, on a sample about 60,000 times smaller.
Turn it around to see why the ratio alone is not enough for a chart place. At the top card's conversion, 40 favourites would imply about 137,000 views — 40 ÷ 0.00029 ≈ 137,900. Early favourite density can be spectacular while the headline counter stays tiny, and the weekly list mixes both readings. A card that reaches double digits in favourites inside its first month has done the hard part; the rest is repetition.
Watching the chart without chasing it
The published nsfw ai list is a poor scoreboard for a single card because it only records the top 20 of a much larger catalogue. Read it instead as a sample of what is currently converting, and check it on Thursdays when the order changes. Readers who want the creator view of the same data can use the weekly rankings boards, where Characters and Creators are listed separately.
For a creator, the honest summary is that early engagement buys a chance rather than a position. The mechanics reward a fast favourite rate inside a fixed window, and they reward precise tags long after that window closes, which is why a launch plan that ignores the tag row tends to look successful for a week and invisible for a year.
Frequently asked questions
How long does a new character have to make an impression?
Seven days is the unit that matters, because the list is rebuilt each Thursday. Within that window the useful measurement is the ratio of favourites to views rather than the total of either.
Can a card with few views outrank one with millions?
On rate, yes: 12 favourites from 400 views is a 3% conversion, far above the 0.029% recorded on the current leading card. On ranking position, not usually, because a 20-slot list has to weigh volume as well as rate.
Do tags matter once a card leaves the chart?
They matter more. Tags are what search matches against, so a precisely tagged card keeps being found for years while a chart appearance lasts a single week.
Build the launch around the window
Publish on a day that leaves the full week available, describe the premise inside two lines, tag it with the vocabulary the catalogue already uses, and read the ratio rather than the raw counter for the first month. Then leave the card alone and let reader behaviour accumulate, because the numbers that matter early are the ones a small sample can still measure honestly.
Data and sources: counters and signatures on the Crushon trends page[1], the search and leaderboard pages[2], and the platform's illustration of how characters recall earlier conversations[3], read 22 Sept 2026. Ratio arithmetic is computed from published counters; the 400-view launch case is an illustrative assumption. Intended for adult readers aged 18 and up.