I still haven't been able to play games but I want to work on retorts tomorrow - mainly Famous and Inquisitor, but maybe others too.
I've already tested military retorts and they were fine - not even combining all military retorts, Life magic and the best early races seemed to threaten game balance too much.
Which leaves magical retorts, and their combinations, let's see...
Alchemy - Honestly this is more of a military retort than magical - the early game benefit is free magic weapons. While trading gold for mana early is good, it merely changes one resource into another instead of creating more of anything - yes, you can have extra mana crystals but then you don't have money to buy settlers, fighter's guilds, units, etc. I consider it a somewhat above average pick but not a direct threat to game balance. Starting sawmill reduced the risk of this greatly, as it leaves way less gold that can be transformed to early mana thanks to its massive upkeep.
Chaneller - is obviously a late game retort with close to no early benefits at the cost of 2 picks. No problem here.
Archmage - A strong retort but I don't think it's overpowered. 50% more SP equals 28% more actual casting skill which is as good as a 22% cost reduction on skill alone - but mana still has to be provided as normal. As it only provides one part of the two benefits of a real cost reduction, I believe it's worth about as much as a 11% cost reduction instead - good but comparable with Specialist and not all that much better than a 9th or 10th book (they only offer 5% but also to research and spells). I believe this part of the effect is fine. The +10 casting skill is a bit more questionable - honestly, it's a lot. At least far more than an extra 28%, so the starting benefit is greater than the long term one. And it has abuse potential in the early game if you stack it with other effects that either help fuel this increased skill, or in worst case further accelerate it. This retort might be more balanced if the starting skill bonus was lower. The ability to bypass the need of 5 whole books for the same staring casting skill is probably the greatest enabler of low book high retort strategies - if they actually had to start with the skill they deserve from books, they'd fail.
Ideally, I would want this skill bonus to be equal to (or a set fraction of) the number of books the wizard owns, but I'm quite sure I have no room to implement that, so we should most likely do a flat +5 skill, or none at all. (This would also fix the Archmage AI's problem of having more skill than starting power base, ending up constantly running on 0 mana until building power buildings.
Artificier - Early trade potential has already been balanced by giving only one free spell instead of two. No threat to early game balance whatsoever, heroes are unavailable and good artifacts are as well, not to mention their cost. At best this allows early access to the Myrror plane, for hero units only - Nothing a Life wizard cannot already do anyway by researching Planar Travel early.
Conjurer - Not overpowered on its own - yes, it's effectively a 33% bonus to skill and mana but limited to one spell type only - even if that's the most frequently used type it's still only a portion of overall spellcasting. Stacking with other cost reductions and skill bonuses is more worrysome. I'll continue this analysis in the "combo" section. The research bonus is nice for flavor and most likely not a threat to balance. I don't like the maintenance reduction. Due to early game rounding, this is a MASSIVE advantage - maintaining 9 sprites for 18 mana out of 18 income is way different than doing the same for 9. It also cuts into Chaneller territory - one retort to reduce maintenance is plenty. Not sure if any other effect needs to be added or not if the maintenance bonus is cut. An alternative solution would be to "fix" the rounding - change it to get applied to the sum maintenance of the total army instead of to every unit, so 4 sprites cost 6 correctly instead of the current 4. The problem with this approach is inconsistent display - individual unit maintenance shown will ignore Conjurer which can be misleading.
Sage Master - Questionable - The +9 RP per turn is the exact same kind of early advantage we want to get rid of. Fortunately, research is hard to abuse in the early game, as it requires both skill and mana to cast stronger spells. Might be fine as is if mana and research skill problems are fixed.
Runemaster - Grants nothing for the early game and produces no resources either directly or indirectly. Safe!
Specialist - I believe I should dig out the discussion about nerfing this or not a few months ago, where we compared the numbers and concluded it's slightly on the strong side but not really a threat. Reducing the cost bonus to 12% might be a good idea but I would like to re-read those discussions - but I have no idea where to find them. Stacking this with books for cost reduction is a lower threat as cost reduction from books is less abundant - at least 9 books required. Stacking with Conjurer might be a risk still.
Cult Leader - Expensive buildings needed for any effect - no threat to the early game.
Astrologer- Effect grows by time elapsed - no threat to the early game.
Spellweaver - Fairly balanced I believe. Even if I accept Overland skill is more valuable than combat skill, Archmage at +28% to both is more (56) and comes with the +10 early bonus, so the skill part is unlikely to be worth more than 1 pick. The 33% power part is superior to the 33% of Mana Focusing, but that effect was aleady judged unworthy of costing an entire pick on the own (which is why the +8 mana bonus was added to it which broke the retort completely). I have a hard time seeing problems with this aside from the multiplication effect from spending the bonus power on skill. However if the "+10 skill" part of Archmage is changed, a reduction on the produced power might be fair to keep the two at the same power level. Now, this is for the generic effect, but idk about the early game. +50% skill and 33% mana at the same time is...fairy powerful for early game casting power to say the least, comparable to Archmage's +10 skill bonus or a ~30% cost reduction.
Combinations for early game potential (the number represents how much base cost worth of spells the wizard can cast per turn using only the starting books and retorts as resources) :
12 books no retorts :
Mana capcity : 24
Skill capacity : 24
Specialist+11 books
Mana Capacity : 28
Skill Capacity : 29
Specialist is a fairly powerful pick here, although it probably should be - it is meant for this sort of setup.
Conjurer+11 books
Mana Capacity : 32
Skill Capacity : 33
About 33% better ratio than 12 books but only on summons. Probably acceptable.
11 books, archmage
Mana capacity : 21
Skill capacity : 32
Looks acceptable - hard to fuel that much skill
Specialist+Conjurer+10 books
Mana production : 36
Skill capacity : 40
Looks overpowered - 50 to 66% higher capacity to summon than 12 books.
10 books, Spellweaver
Mana capacity : 26.666
Skill capacity (overland only) : 33.3333
Only marginally better than 12 books and worse in combat. I think safe.
9 books, Spellweaver Archmage
Mana capacity : 20
Skill capacity : 45
Problem is obvious, 45 skill is huge. This has already been disabled by "mutually exclusive" for late game problems - three multipliers on casting skill stacked is too much, so, yay, good job...I guess. Crude solution, but it works...
9 books, Conjurer, Specialist, Archmage
Mana capacity : 27
Skill capacity : 50.9
Massive skill potential - I dare say this is "broken as hell". If Archmage is limited to +5 skill, it's still 41 skill capacity, and at none it's 32.
Surprisingly, mana potential is massively worse than without Archmage.
8 books, Conjurer, Specialist, Spellweaver
Mana capacity : 26.666
Skill capacity : 40
Less overpowered than the Archmage variety, or equal to it if it is reduced to "+5 skill".
As a conclusion, I believe the two critical points that need change are
-The +10 bonus on Archmage should be +5 or none (used to be +5 before 1.31 MoM, lol)
-The way Conjurer, specialist, and books cost reductions stack is problematic, it should be changed somehow. If the effect was multiplicative instead of subtractive, it could improve balance, in particular for the strongest offender, 10+Specialist+Conjurer, it would change Mana capacity to 31 and Skill capacity to 34 - still high numbers but only 50% more than what 12 random books can do...and this setup requires both specialization to a single realm and focusing on summoning spells only to actually have the advantage.
I've already tested military retorts and they were fine - not even combining all military retorts, Life magic and the best early races seemed to threaten game balance too much.
Which leaves magical retorts, and their combinations, let's see...
Alchemy - Honestly this is more of a military retort than magical - the early game benefit is free magic weapons. While trading gold for mana early is good, it merely changes one resource into another instead of creating more of anything - yes, you can have extra mana crystals but then you don't have money to buy settlers, fighter's guilds, units, etc. I consider it a somewhat above average pick but not a direct threat to game balance. Starting sawmill reduced the risk of this greatly, as it leaves way less gold that can be transformed to early mana thanks to its massive upkeep.
Chaneller - is obviously a late game retort with close to no early benefits at the cost of 2 picks. No problem here.
Archmage - A strong retort but I don't think it's overpowered. 50% more SP equals 28% more actual casting skill which is as good as a 22% cost reduction on skill alone - but mana still has to be provided as normal. As it only provides one part of the two benefits of a real cost reduction, I believe it's worth about as much as a 11% cost reduction instead - good but comparable with Specialist and not all that much better than a 9th or 10th book (they only offer 5% but also to research and spells). I believe this part of the effect is fine. The +10 casting skill is a bit more questionable - honestly, it's a lot. At least far more than an extra 28%, so the starting benefit is greater than the long term one. And it has abuse potential in the early game if you stack it with other effects that either help fuel this increased skill, or in worst case further accelerate it. This retort might be more balanced if the starting skill bonus was lower. The ability to bypass the need of 5 whole books for the same staring casting skill is probably the greatest enabler of low book high retort strategies - if they actually had to start with the skill they deserve from books, they'd fail.
Ideally, I would want this skill bonus to be equal to (or a set fraction of) the number of books the wizard owns, but I'm quite sure I have no room to implement that, so we should most likely do a flat +5 skill, or none at all. (This would also fix the Archmage AI's problem of having more skill than starting power base, ending up constantly running on 0 mana until building power buildings.
Artificier - Early trade potential has already been balanced by giving only one free spell instead of two. No threat to early game balance whatsoever, heroes are unavailable and good artifacts are as well, not to mention their cost. At best this allows early access to the Myrror plane, for hero units only - Nothing a Life wizard cannot already do anyway by researching Planar Travel early.
Conjurer - Not overpowered on its own - yes, it's effectively a 33% bonus to skill and mana but limited to one spell type only - even if that's the most frequently used type it's still only a portion of overall spellcasting. Stacking with other cost reductions and skill bonuses is more worrysome. I'll continue this analysis in the "combo" section. The research bonus is nice for flavor and most likely not a threat to balance. I don't like the maintenance reduction. Due to early game rounding, this is a MASSIVE advantage - maintaining 9 sprites for 18 mana out of 18 income is way different than doing the same for 9. It also cuts into Chaneller territory - one retort to reduce maintenance is plenty. Not sure if any other effect needs to be added or not if the maintenance bonus is cut. An alternative solution would be to "fix" the rounding - change it to get applied to the sum maintenance of the total army instead of to every unit, so 4 sprites cost 6 correctly instead of the current 4. The problem with this approach is inconsistent display - individual unit maintenance shown will ignore Conjurer which can be misleading.
Sage Master - Questionable - The +9 RP per turn is the exact same kind of early advantage we want to get rid of. Fortunately, research is hard to abuse in the early game, as it requires both skill and mana to cast stronger spells. Might be fine as is if mana and research skill problems are fixed.
Runemaster - Grants nothing for the early game and produces no resources either directly or indirectly. Safe!
Specialist - I believe I should dig out the discussion about nerfing this or not a few months ago, where we compared the numbers and concluded it's slightly on the strong side but not really a threat. Reducing the cost bonus to 12% might be a good idea but I would like to re-read those discussions - but I have no idea where to find them. Stacking this with books for cost reduction is a lower threat as cost reduction from books is less abundant - at least 9 books required. Stacking with Conjurer might be a risk still.
Cult Leader - Expensive buildings needed for any effect - no threat to the early game.
Astrologer- Effect grows by time elapsed - no threat to the early game.
Spellweaver - Fairly balanced I believe. Even if I accept Overland skill is more valuable than combat skill, Archmage at +28% to both is more (56) and comes with the +10 early bonus, so the skill part is unlikely to be worth more than 1 pick. The 33% power part is superior to the 33% of Mana Focusing, but that effect was aleady judged unworthy of costing an entire pick on the own (which is why the +8 mana bonus was added to it which broke the retort completely). I have a hard time seeing problems with this aside from the multiplication effect from spending the bonus power on skill. However if the "+10 skill" part of Archmage is changed, a reduction on the produced power might be fair to keep the two at the same power level. Now, this is for the generic effect, but idk about the early game. +50% skill and 33% mana at the same time is...fairy powerful for early game casting power to say the least, comparable to Archmage's +10 skill bonus or a ~30% cost reduction.
Combinations for early game potential (the number represents how much base cost worth of spells the wizard can cast per turn using only the starting books and retorts as resources) :
12 books no retorts :
Mana capcity : 24
Skill capacity : 24
Specialist+11 books
Mana Capacity : 28
Skill Capacity : 29
Specialist is a fairly powerful pick here, although it probably should be - it is meant for this sort of setup.
Conjurer+11 books
Mana Capacity : 32
Skill Capacity : 33
About 33% better ratio than 12 books but only on summons. Probably acceptable.
11 books, archmage
Mana capacity : 21
Skill capacity : 32
Looks acceptable - hard to fuel that much skill
Specialist+Conjurer+10 books
Mana production : 36
Skill capacity : 40
Looks overpowered - 50 to 66% higher capacity to summon than 12 books.
10 books, Spellweaver
Mana capacity : 26.666
Skill capacity (overland only) : 33.3333
Only marginally better than 12 books and worse in combat. I think safe.
9 books, Spellweaver Archmage
Mana capacity : 20
Skill capacity : 45
Problem is obvious, 45 skill is huge. This has already been disabled by "mutually exclusive" for late game problems - three multipliers on casting skill stacked is too much, so, yay, good job...I guess. Crude solution, but it works...
9 books, Conjurer, Specialist, Archmage
Mana capacity : 27
Skill capacity : 50.9
Massive skill potential - I dare say this is "broken as hell". If Archmage is limited to +5 skill, it's still 41 skill capacity, and at none it's 32.
Surprisingly, mana potential is massively worse than without Archmage.
8 books, Conjurer, Specialist, Spellweaver
Mana capacity : 26.666
Skill capacity : 40
Less overpowered than the Archmage variety, or equal to it if it is reduced to "+5 skill".
As a conclusion, I believe the two critical points that need change are
-The +10 bonus on Archmage should be +5 or none (used to be +5 before 1.31 MoM, lol)
-The way Conjurer, specialist, and books cost reductions stack is problematic, it should be changed somehow. If the effect was multiplicative instead of subtractive, it could improve balance, in particular for the strongest offender, 10+Specialist+Conjurer, it would change Mana capacity to 31 and Skill capacity to 34 - still high numbers but only 50% more than what 12 random books can do...and this setup requires both specialization to a single realm and focusing on summoning spells only to actually have the advantage.
