Bayes' theorem asks: given a positive test, how likely is it you really have the condition? The answer is not the test's accuracy — it also depends on how common the condition is (the base rate).
Told as natural frequencies the puzzle dissolves. Out of 1,000 people, only a handful truly have a rare condition. A 90%-accurate test still flags 10% of the vast healthy crowd — so the false alarms can swamp the true cases. That is base-rate neglect: reading the 90% straight off and forgetting the crowd.
The posterior is just the area of the gold tiles divided by the area of all the stained tiles: TP / (TP + FP). Countable, not scary.
One test rarely settles it. Fold a positive result into the next test's prior and repeat — agreeing tests drive belief toward certainty. That is sequential updating, the engine of learning from evidence.
Something in the simulation stopped unexpectedly — the lesson continues without it. You can move on; nothing you did was wrong.