Direct answer

Server capacity is shaped by player distribution, entity AI, farms, chunk generation, mods and plugins. A single "X players per GB" claim cannot predict your build.

Recommended procedure

  1. Choose a reproducible busy period and log player count, world count, CPU, memory, GC and TPS/MSPT.
  2. Use /spark tps for tick health and /spark profiler start --timeout 120 during a representative event.
  3. Compare players concentrated in one base against players spread into different terrain or dimensions; chunk and region load changes.
  4. Note how often joins, teleports and exploration create spikes even when average TPS remains 20.

What to do next

Build a capacity baseline from actual workload and include headroom for world saves, backup jobs and events. Repeat after each large plugin or modpack upgrade.

How to verify the result

Confirm stability across several normal peak sessions before raising public player limits.

Version and compatibility notes

Folia region parallelism is not a simple "one player per thread" model. Tick time, network RTT and memory usage measure different resource constraints.

Official references

Was this article helpful?
Help us improve this article

Tell us what is missing, unclear, outdated, or incorrect. You can also suggest something we should add.

What could we improve?
Do not include passwords, server IPs, or other private information.0 / 1000