Ritual/site searching questions

[Baalz]

I take it there is no "auto forge" option in the same vein? It'd be nice to set up little owl quill factories and such without having to micromanage each turn.

[Nerfix]

Baalz said:
I take it there is no "auto forge" option in the same vein? It'd be nice to set up little owl quill factories and such without having to micromanage each turn.

That would be good, yes.

[JaydedOne]

You know, that shouldn't be hard to code in... Given that you can shift-m for rituals... But then, Johan O. might have my head on a spit for incorrectly stating that. ;-)

[Neophyte]

Reading the pre-release threads, I found myself understanding that the likelihood of finding that type of site in that type of terrain would be taken into consideration as well. For example, with certain sites settings, it might be more likely to find an Astral site in plains than in Mountains, and thus the auto-search would suggest a plains for Arcane Probing. Is this not the case?

[thejeff]

Doesn't seem to take province type into consideration.

What it seems to do:
When the order is given, choose the lowest numbered unsearched province.
When the spell is cast, again set the caster to cast the spell on the lowest numbered unsearched province, if the gems are available.
Two annoyances with this are, since spells are cast before battles, it will not target provinces taken that turn. Also, new gems haven't been acquired yet, so you have to have enough gems left at the end of the turn for next turn's searches.
I'm not sure what happens if a targetted province is taken. Probably the order is cancelled.

[Wick]

This is the site searcing algorithim I'd use. I doesn't check strategic considerations because that's hard. It also doesn't factor that existing searches in other paths improve the odds because I've forgotten the proof that it does.

I've used variables instead of numbers so numberOfPaths = 9 (including Holy), maxSlots = 4 etc. and if the rest isn't obvious then I haven't done it right:

for i = 1 to provinces
if provinceControl(i) = player then
for j = 1 to numberOfPaths
if searchingLevel(j) > 0 then
searchedLevel = withDeclaredSearchers(i, player)
if searchingLevel(j) > searchedLevel then
slotvalue = 0
for k = 1 to (maxSlots - knownSites)
slotValue = slotValue + (baseChance + terrainChance)^j * (searchingLevel - searchedLevel)
next k
endif
endif
searchValue(i) = searchValue(i) + slotValue
next j
endif
next i

Then search the most valuable province.

[Agrajag]

Wick said:
It also doesn't factor that existing searches in other paths improve the odds because I've forgotten the proof that it does.

If I understand what you mean, then the proof is as follows:
A province has a certain number of magic sites, for example lets say a certain province has 1 magic site.
If you searched the province for every magic other than death, and didn't find anything, then that means that the site must be a death site, so the province having been searched for other magics without finding any sites means it is more likely that what sites are in the province will be of the path you are now searching for.

[Corwin]

Neophyte said:
Reading the pre-release threads, I found myself understanding that the likelihood of finding that type of site in that type of terrain would be taken into consideration as well. For example, with certain sites settings, it might be more likely to find an Astral site in plains than in Mountains

Really? What settings and why?

Could you point me to some of the threads that you've mentioned by any chance?

[Taqwus]

Some are terrain-dependent; you shouldn't see "Inkpot End" outside of swamps, if memory serves. Another ex: most can only appear either on land, or underwater, but not both.