pub fn record_forge_decision(seats: usize, elapsed: Duration)Expand description
The rules work between a seat answering and the next prompt appearing, split by seat count. Measured over two days of captures, four seats run this at a p99 of about 2.9s against 0.35s for two, and put 2-5% of decisions over two seconds where two seats put none. A fleet-wide quantile averages the many clean rooms against the few bad ones and shows neither.
There is no bot count. It was recorded from the engine’s own AI seat list, which is empty for every hosted game because bots join as ordinary relay clients, so the label read zero on every series it ever produced. Only the relay knows which seats are bots, and the node is not told.