Skip to content

[Warlock] Model Dominion of Argus demon summon RNG#11532

Merged
nyterage merged 1 commit into
simulationcraft:midnightfrom
millanzarreta:wl-rng-doa-demons
Jun 25, 2026
Merged

[Warlock] Model Dominion of Argus demon summon RNG#11532
nyterage merged 1 commit into
simulationcraft:midnightfrom
millanzarreta:wl-rng-doa-demons

Conversation

@millanzarreta

@millanzarreta millanzarreta commented Jun 25, 2026

Copy link
Copy Markdown
Contributor

During a Dominion of Argus window, several demons are summoned from a pool of four possible demons. Until now, this was implemented in SimC as an approximation: a 20% chance for Sacrolash, a 20% chance for Grand Warlock Alythess, a 30% chance for Inquisitor, and a 30% chance for Jailer.

After collecting more than 207 hours of sample logs, we found that the RNG system used to select which DoA demon is summoned follows these simple rules:

  • The system always performs a flat 25% base roll among the 4 demons.
  • Rolling Grand Warlock Alythess while one Grand Warlock Alythess is already active redirects to Jailer (summoning a Jailer instead).
  • Rolling Sacrolash while one Sacrolash is already active redirects to Inquisitor (summoning an Inquisitor instead).

And that is all. The analysis was somewhat tricky because this redirection effect made it look as though there was some kind of memory or BLP involved, but that does not end up being the case. The final results strongly favor a flat 25% roll with the redirection rules described above. This is also why there can never be more than one Grand Warlock Alythess active at the same time, and the same applies to Sacrolash.

A separate test was also performed without the final talent point of Dominion of Argus, which makes these demons last 10 seconds instead of 14 seconds. In that case, the observed minimum time between summons of these same demons (Grand Warlock Alythess → ... → Grand Warlock Alythess and Sacrolash → ... → Sacrolash) dropped from 14 seconds to 10 seconds, indicating that the restriction is most likely based on whether that same demon is currently active, rather than on a fixed ICD or something else.

The following ZIP contains the results of all the tests performed in-game. In total, this includes more than 207 hours of testing, 11,890 DoA windows, and 52,080 summoned DoA demons:
DOA_ARGSUM_TESTS.zip

@nyterage nyterage merged commit 2b31327 into simulationcraft:midnight Jun 25, 2026
28 checks passed
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

2 participants