Astrologer and Mana Focusing

[Tlaloc]

Well I'm a bit late to the discussion, but I agree with Catwalk that mutually exclusive retorts aren't a great design. If Archmage and Spellweaver need to be mutually exclusive, it's a sign that one or both need be changed so that's no longer the case.

[Nelphine]

Agreed. Although I'm apparently in the minority that spell retorts are too strong.

[Seravy]

Well I'm a bit late to the discussion, but I agree with Catwalk that mutually exclusive retorts aren't a great design. If Archmage and Spellweaver need to be mutually exclusive, it's a sign that one or both need be changed so that's no longer the case.

It's the multiplication. When the bonus is multiplicative, having both together will always be a greater benefit than simply adding the bonus amount. And two casting skill bonus retorts that apply in different phases of the casting skill calculation will always be multiplicative.
It's certainly not a good design but there is nothing better if we want two casting skill retorts (and I believe we do - overland skill, and generic skill are entirely different).
Magical resources go through a pretty long "pipeline" - inevitably any retort that applies to any part of this will multiply the effect of others.

The amount of spells you can cast per turn is ~ (1+Casting cost reduction)*(SP bonus)*(Skill Bonus)*(Power Bonus) for each point of power gained and spent on skill.
The only way to avoid this sort of unbalancing effect is - aside from mutually exclusive retorts - is to pick one of the above four and remove all retorts that do the other 3, or at least 2 of them. Two categories are slightly unbalanced, three are outright broken.

If you have a better idea, let me hear it, but I believe we can agree that we want at least one generic and one overland casting skill retort?

[RefSteel]

Assuming overland just means "outside of combat," would it be possible to have an overland casting skill retort and a (separate) combat casting skill retort, such that picking both would be the equivalent of a single "generic" casting skill retort?

[Seravy]

Assuming overland just means "outside of combat," would it be possible to have an overland casting skill retort and a (separate) combat casting skill retort, such that picking both would be the equivalent of a single "generic" casting skill retort?

That's an option but we have Chaneller for combat mana costs and I didn't want two "combat spells" retorts.
Also, Archmage is something I always liked and don't want to remove unless there is a very strong reason to do so.
(besides, having to pick 2 retorts to have a generic casting skill bonus is not the same, as it costs more picks and retort slots. Since Spellweaver costs 2, this would put the total cost of the "generic" option at 3 picks - far too expansive.)

[Domon]

what about changing archmage into giving you 33% more casting skill instead of better expenditure? then the modifiers can be added together instead of multiplied.

also split the +10 to casting skill between the 2?

[Seravy]

what about changing archmage into giving you 33% more casting skill instead of better expenditure? then the modifiers can be added together  instead of multiplied.

also split the +10 to casting skill between the 2?

Well, the current +50% spending is worth about 28% in actual casting skill so...that would be the number to use.
But this is really hard to implement - right now the code pretty much looks like this :

entering combat :
If spellweaver, Combat Skill=Skill*3/2. Add 1 if the remainder is nonzero.

Your suggestion would require something like
If spellweaver, Combat Skill=Skill*3/2. Add 1 if the remainder is nonzero.
but if spellweaver and archmage, Combat Skill=Skill*(1/1.78)*(1.28) instead. Add god knows how many and when to have correct rounding with those numbers.

I'm happy with the current solution, it's not perfect but not bad enough to make me want work another whole day on replacing it.

[Catwalk]

I agree with RefSteel's suggestion, it makes perfect sense.

[Seravy]

The amount of spells you can cast per turn is ~ (1+Casting cost reduction)*(SP bonus)*(Skill Bonus)*(Power Bonus) for each point of power gained and spent on skill.

Giving further thought to the issue, this also explains why Conjurer (especially paired with specialist) has problems. Cost reductions stack with skill bonuses the same way as two kinds of skill bonus can.

The current conjurer effectively does:
+33% summon spells cast
+25% research received on summons (marginal effect)
about 1/3 less maintenance paid due to rounding (strong in the early game, not so much later).

This is fair on its own (except the maintenance problems with rounding - half on 2 cost units is a big deal in the early game) but if you stack additional casting skill or cost reduction retorts, it becomes problematic.

[Nelphine]

Agreed.