Last updated: September 21, 2026 · Category: Guide · Editorial review · About the authors
Key Takeaways
- A real-to-fake join ratio has no universal "good" number. It only means something compared to a server's own recent history.
- A sudden shift in the ratio has two common causes: a real farming attempt, or a changed anti-fake threshold reclassifying the same underlying joins.
- Checking which cause it is takes one step: look at whether the threshold setting changed before assuming the join pattern itself changed.
Once anti-fake filtering is running, a server gets a new number to look at: the ratio of joins classified real versus fake. The real-to-fake ratio is the percentage split between joins an account-age filter accepts and joins it rejects over a given window, and it depends entirely on where the threshold is set, since the same account-creation timestamps, encoded directly in each account's snowflake ID, retrieved 2026-09-21, get classified differently depending on the cutoff. The number on its own answers less than it looks like it does.
Why There Is No Universal Benchmark
A 70/30 real-to-fake split means something completely different depending on context. A small, invite-only community with no public presence and no rewards program might see almost no fake joins at all, since there is little reason for anyone to farm it. A large server running a public giveaway is a much more attractive farming target, and a 70/30 split there could represent a real, ongoing problem or simply the normal cost of running a public reward at scale.
Comparing a ratio across servers, or against some external "industry average," answers a question nobody actually needs answered. The useful comparison is a server's own ratio against its own recent history: did this number just change, and if so, why.
What a Sudden Shift Actually Means
A ratio that has been stable for weeks and then moves sharply has, in practice, one of two causes.
A real change in join behavior. Something happened that created a farming incentive: a reward was announced, a giveaway went live, a milestone role became newly achievable. Auditing the join timeline around the exact window the ratio shifted usually confirms this quickly, since farming produces a visible timing cluster.
A configuration change, not a behavior change. If the anti-fake threshold itself was tightened, say from a 3-day account-age minimum to a 14-day one, the exact same population of joins gets reclassified differently. Nothing about actual join behavior changed; the measuring stick did. This is easy to miss because the ratio number moves in exactly the way a real attack would move it.
Checking Which One Happened
The fastest way to tell the two apart is to check the threshold setting before checking anything else. If the anti-fake threshold has not changed recently, the shift is almost certainly behavioral, and an audit of the join timeline around the shift date will confirm what happened. If the threshold was changed in the same window the ratio moved, that is very likely the full explanation on its own, and no further investigation is needed.
This single check, threshold changed or not, resolves most cases of "my fake percentage suddenly went up" without requiring a deeper audit at all. It is worth doing first, since it is faster than a full timeline review and rules out the most common false alarm.
Support tickets that describe a spiking fake-join percentage as an "attack" turn out, more often than not, to trace back to a threshold change the server owner made and then forgot about, sometimes weeks earlier while testing a stricter setting. Checking the Discord verification level, retrieved 2026-09-21, and any anti-fake threshold change log together before assuming an attack saves most servers from a false alarm.
Frequently Asked Questions
What is a good real-to-fake join ratio on Discord?
There is no universal number. A ratio only means something relative to a server's own history, so the useful question is whether it just changed, not whether it matches a benchmark.
Why did my fake-join percentage suddenly increase?
The most common cause is a reward announcement that created a farming incentive overnight. Check whether the spike lines up with anything the server announced in the same window.
Does a rising fake percentage always mean an attack?
No. A stricter anti-fake threshold reclassifies more borderline joins as fake without any change in actual behavior, so check whether the threshold itself changed before assuming an attack.
How often should I check my real vs. fake ratio?
Weekly for most servers, and immediately after any reward announcement or unusual growth spike, since that is when the ratio is most likely to move.