How many test attempts were there? 1000? 10,000?

As I recall, the number of test attempts is related to the required accuracy of the result (the spread of the sample). The minimum effect of an artifact or skill is 5%. Let's take that as a conditional unit of accuracy, instead of trying to find 1% differences. The larger the sample, the more accurate the result, but 10,000 or even 1,000 attempts is the developers' concern, not mine (my hand would get tired). My concern is to determine the order of effect (with a 5% margin of error), i.e., whether or not a particular factor affects the final result. For 5%, you need to run the sample for a sufficient amount of time, but you can determine the difference between 60% and 90% much faster. The larger the sample, the more evenly the attempts are distributed across the series (it won't happen that 60 zeros are followed by 40 ones). Neither you nor I know what calculation algorithm is used in the game – we can only guess. At first, I thought your conclusion was correct, but when I tested it, I came to a different conclusion. If you have time, re-check it; maybe I made a mistake somewhere. But my approach is also more practical – it's easier to add several numbers than to multiply them.:smile14: