Re: Luck scale strategy
 

thejeff:
I believe lch confirmed the province check order by code diving, but I'd have to ask him to be sure.

My best guess is that for each possible event, it runs through your provinces in numerical order and checks to see if that event occurs there. So each event is independently checked against all your provinces until it happens.

Given most turns don't see 4 events, the effect is going to be somewhat weak.

Re: Luck scale strategy
 

Hmmmm.

Does it seem as if events occur by type?

Ie., do barbarian events occur in clumps

Barb
barb
Magic item
Magic item

Or do they occur

Barb
Magic item
Barb
Magic item

There was also discussion about times when more than 4 events would occur


I know that I have tended to see clumps of events around a province #, often seeing a "Bad event" followed by a good event in a province. But I would agree that they do not seem to clump at the low province #'s.

How about this as a testable hypothesis.

A: illwinter creates a list of provinces in a players empire.

B. Illwinter randomly picks a province from the list of provinces (that has not been previously checked). So if a empire has 10 province a rnd(10) is generated to determine the province.

C. Each province has a site mask; which can create a list of eligible events (both good and bad) that can occur in that territory.

For that province, While events < 4, each event is checked to see if EACH event occurs.

E. Check and see if an event is cancelled, due to fortunetellers (Still counts against the limit of 4)

F. If events < 4, return to b.




The only way I see to really test this hypothesis: Knowing the site masks, you would know how many events a province was eligible for. Over a large sample size, the predominance of events would correspond to each provinces # of eligible events.

Ie., suppose an empire of 4 provinces
Province has an order generator, and is only eligible for 10 random events
Province b: 5
Province c: 5
Province d: 20.

If the frequency mapped the predominance of the masks - we would consider that a pretty good indication.

Now, if on the other hand, the mechanims was roll 4 provinces, and see if eligible events occurs in those 4 provinces - then the total number of events would be lower in an empire with few number of eligible events.

Additionally, this would explain those very rare occurrences when more than 4 events occur.

One final bit of testing would be interesting. Does an empire with 2 provinces (and identical dominions and site masks get half as many luck events as an empire with 4 provinces.

Re: Luck scale strategy
 

So far in my games I haven't actually recorded hard data, I probably should.

My gut feeling is that certain provinces get events a lot more frequently, I can't discern any pattern in good/bad events like you mentioned.

I'll dig through my events and see what I come up with.

btw is there a way to rewind through each turn's message summary using my current save?

Re: Luck scale strategy
 

no...