[Saber Cherry]

SaberCherry, can you tell me how my reasoning is invalid (if it is)?

Hmmm... yeah, that's invalid. The intital event probability based on all neutral scales is assigned a unitless 100, no matter how common events actually are. This does not mean there is a 100% chance of events, it just means that by default, we're calling the frequency of events you get with neutral scales "100" for convenience, because it makes the math work out easily. You could call it "1" or "50", but that would not change the results.

So taking Order-3 lowers your relative event probability to 70, or 70% of the probability compared to neutral scales. If you normally had a 30% chance of an event per turn with neutral scales, Order-3 would reduce the chance to 21%, NOT to 0%. Turmoil-3 would increase it to 39%, not to 60%.

If neutral scales have a relative "100" event frequency, then it also has a relative "50" good event and "50" bad event frequency. Order-3 Unluck-3 has a 100*70%=70 total relative event frequency, with a relative 70*20%=14 good event and 70*80%=56 bad event frequency.

Thus, if every (for example) 40 turns you got 100 events with a neutral scale, 50 of them should be good, and 50 bad. But with an order-3 unluck-3, you should only get 70 total events in the same time, of which 56 are bad and 14 are good.

So you get a tiny (12%) increase in bad events, a huge decrease (62%) in good events (which is OK, because they aren't as potent), and a 21% (or maybe 30%) increase in income. For free. ASSUMING all the published formulas are correct. And the income boost can be even greater if you take a good castle, growth, or productivity.

-Cherry