Hey Fabian. I think while you have a benchmark, stories are related by complexity and not time leading to the same discrepancy