Docs unclear about minReplica and queueLength

[maxiedaniels]

I understand that the usual calculation is your job count / queueLength is how KEDA determines how many replicas it needs. But let’s say you have 50 jobs and queueLength is 10, therefore target is 5 pods.
If you then set minReplicas to 3, when does it start scaling? Would 50 jobs result in five additional pods (so a total of 8)? Or does it stop the first three triggers instead, so it would be three replicas until it hits 40 jobs, and then it would scale to 4 pods, etc?