Quote:I agree I have felt this way as well. Summoning a storm giant takes me a million turns but at the same time I can cast dozens of battlefield enchantments in the same turn and bankrupt my mana supply.
I think this is fun though.
I agree. It's fun so I'm not worried about it.
Combat spellcasting has two extra constraints overland does not :
1. the end of combat. Few, quick units? combat ends in 3 turns so 3 is all the spells you get.
2. the distance multiplier. At a 3x cost, using all your skill is possible a horrible idea unless it's the endgame.
Combat spellcasting has a huge disadvantage to overland spellcasting, too. The effects are - usually - not permanent. It might or might not help you win that battle and that's it. An overland spell - like the mentioned storm giant - gives you something that stays and can be used for the rest of the game.
We could use a multiplier to reduce casting skill in combat, but that would make early game a lot worse. Unless it works like "halve all casting skill over 100 in combat".
By the way, while 200-300 skill in combat feels excessive if casting common spells, it's not that much when spent on very rares. You can only summon like 4 earth elementals from that, or cast high prayer, supreme light and one more enchantmant, like mass invisibility and that's all. Or you can Animate Dead 4-6 times, etc..
I think it's fine as is, and makes common/uncommon combat spells age better. While there effect is not as good as the better spells, you can cast 4-5 times as many of them and that might make up for it - if the combat is long enough to do so.
(June 26th, 2016, 04:19)Seravy Wrote: We could use a multiplier to reduce casting skill in combat, but that would make early game a lot worse. Unless it works like "halve all casting skill over 100 in combat".
Could just make range penalties affect casting skill as well as mana cost. Maybe change the range bands, instead of +0.5 every 5 squares distance make it every 10.
I still insist having 'amplifying tower' provide only overland skill, but about 14 (instead of 7) to compensate half a benefit. Without this building I don't remember combat skill being extreme at all, though overland skill was generally too low late in the game, hence the introduction of amplifying tower.
But the issue late game is how much you are limited to spells overland. You may have dozens of cities recruiting many top-tier units per turn, but overland you are limited to a [better] summoning unit every few turns and only if you don't decide to cast other kinds of overland spells.
Original Late Game: 150 casting skill combat and overland
Caster of Magic Late Game (assume 15 amplifying towers) - about 250 casting skill combat and overland (300-350 if sorcery with blessing of uranus)
Proposal of 14 overland-only skill per tower (assume 15 amplifying towers) - 150 combat skill and 350-375 overland skill.
Maybe it's my preference, but wizards having immense overland skill late in the game would make games brutally interesting with fast global/city enchantments vs disjunction/disenchant, fast recruitment of fantastic units and unit enchantments, mass-casting blizzards/fire storms/black wind, etc
Meanwhile, combat will not greatly deviate at the end, having wizards run out of skill eventually during each battle, requiring decent armies to finish things up. Given that battles are far more frequent late in the game, it's not like we're starved out of combat mana anyway.
(June 27th, 2016, 16:11)zitro1987 Wrote: I still insist having 'amplifying tower' provide only overland skill, but about 14 (instead of 7) to compensate half a benefit. Without this building I don't remember combat skill being extreme at all, though overland skill was generally too low late in the game, hence the introduction of amplifying tower.
But the issue late game is how much you are limited to spells overland. You may have dozens of cities recruiting many top-tier units per turn, but overland you are limited to a [better] summoning unit every few turns and only if you don't decide to cast other kinds of overland spells.
Original Late Game: 150 casting skill combat and overland
Caster of Magic Late Game (assume 15 amplifying towers) - about 250 casting skill combat and overland (300-350 if sorcery with blessing of uranus)
Proposal of 14 overland-only skill per tower (assume 15 amplifying towers) - 150 combat skill and 350-375 overland skill.
Maybe it's my preference, but wizards having immense overland skill late in the game would make games brutally interesting with fast global/city enchantments vs disjunction/disenchant, fast recruitment of fantastic units and unit enchantments, mass-casting blizzards/fire storms/black wind, etc
Meanwhile, combat will not greatly deviate at the end, having wizards run out of skill eventually during each battle, requiring decent armies to finish things up. Given that battles are far more frequent late in the game, it's not like we're starved out of combat mana anyway.
I've had plenty of battles going where I needed all 250+ of my casting skill to stand a chance so not sure if this change would be good.
In overland casting, the player is forced to decide what to cast because their amount of skill is limited, and spells you can cast on overland are generally always beneficial to you (few exceptions exist). So the decision is more "which is more important" instead of "is this worth it".
In combat, the opposite is true. You want to end the combat and win as soon as possible, casting as few spells as possible. Using all your skill every battle is going to make you bankrupt...but you do have the option for critically important battles. Here the decision is, which battle is worth the mana crystals and how many, since aside from winning or losing that battle, your invested crystals give you nothing. (Death magic breaks this rule, which is one of their main strength actually)
Also, think I already mentioned this but considering the cost of high rarity combat spells, 250 skill is fair. It's reasonable to expect to be able to cast spells at least for the first 5 or so combat turns, and most spells of this rarity cost an average of 50.
Wizards already have insane casting skill. In the save file I got from Hadriex, one wizard has a skill of 707 and even the second best was 500+. That is without the casting advantage AI players get. Yes it was very late game, 1430, but high difficulty games can easily last that long. Even Hadriex himself had over 250 skill and he was losing.
The ai is defending his capital with 3 skeletons. The ai must be over rating skeletons. I could remove him from the game with my swordsman if I had been heading that direction. He should be more careful, I saw an army of two nagas earlier moving way way out from his territory, they should probably be at home.
Also I found a massive problem with myrran ai's which explains why I found a myrran ai with 3 cities on turn 120-150ish.
They spam settlers. They don't have the population growth to do it though. Somehow the ai code needs to adapt to having extremely low growth otherwise AI dark elves will lose all their population and thus production, settle some outposts that die due to mountains and such. Not sure what other turns from this game to send you to help show this.
EDIT: updated save file. we're at war. He has big armies randomly standing on the coast that cannot fly or swim. things like cavalry. yet his cities are mostly undefended.
note save2.gam is much earlier in game than 1043.gam
edit2: barebones solution, tell the capital to not produce settlers ever unless it's population is at least 4? or some other number? 3? 5?
edit3: if this is impossible you might have to raise the minimum racial population growth rate to something like 20or30, to prevent cheating ai production bonuses from vastly outstripping population growth? I'm not sure if any race besides dark elf can run into this problem or not. some others might be able to though.
edit4: good god are life barbarians an insanely strong OP ai start :-p totally irrelevant to my point but I can't sleep! I think almost every AI who ever gives me trouble is barbarian.
(June 30th, 2016, 19:14)namad Wrote: The ai is defending his capital with 3 skeletons. The ai must be over rating skeletons. I could remove him from the game with my swordsman if I had been heading that direction. He should be more careful, I saw an army of two nagas earlier moving way way out from his territory, they should probably be at home.
Also I found a massive problem with myrran ai's which explains why I found a myrran ai with 3 cities on turn 120-150ish.
They spam settlers. They don't have the population growth to do it though. Somehow the ai code needs to adapt to having extremely low growth otherwise AI dark elves will lose all their population and thus production, settle some outposts that die due to mountains and such. Not sure what other turns from this game to send you to help show this.
EDIT: updated save file. we're at war. He has big armies randomly standing on the coast that cannot fly or swim. things like cavalry. yet his cities are mostly undefended.
note save2.gam is much earlier in game than 1043.gam
edit2: barebones solution, tell the capital to not produce settlers ever unless it's population is at least 4? or some other number? 3? 5?
edit3: if this is impossible you might have to raise the minimum racial population growth rate to something like 20or30, to prevent cheating ai production bonuses from vastly outstripping population growth? I'm not sure if any race besides dark elf can run into this problem or not. some others might be able to though.
edit4: good god are life barbarians an insanely strong OP ai start :-p totally irrelevant to my point but I can't sleep! I think almost every AI who ever gives me trouble is barbarian.
Prior to turn 50 the AI prioritizes expansion over defending the capital, since attacks are unlikely, especially ones having enough force to take out a capital despite the lightning (as the AI will get the update to not declare war on other AI until turn 40 this will be even more true). In later turns they keep their best units there to protect themselves.
Armies standing on shore are waiting for ships to pick them up. Unfortunately they haven't produced any yet. Or they might be waiting for more units so the stack gets big enough to attack something.
AI is already limited to not make settlers under a certain population amount, I think currently set to 2. Not having more cities can also be the result of failing to build a ship to leave the continent (if it's small). If you send me like, turn 120 I'll try to figure out what happened.
I sieged his city after the second save. He never moved out. Then died. It took me dozens of turns because I tried to expand a bit at the same time. I also intentionally let his settlers slip out and settle outposts in the hopes I'd get them as free cities, and I did, but the red player stole them from me. On turn 110 (I took a break from the game for real life and I am not yet on turn 120) he's been dead for a while and I'm fighting to keep his cities from an attack by the red wizard.
I think you should change population 2 to population 3. The problem is that if you only have 3 population you build your settler too slowly, then you go down to 2 population and build everything too slowly. If that minimum population number to build settlers got +1 or if every race in the game got +10 growth I think the problem would go away.
Even though in the most recent save the wizard doesn't exist you can surely see the file. I'm doing quite poorly overall.
edit: This might be the only time I've seen this happen to an ai. However I have seen ai's die before turn 50, and I have seen ai's with almost nothing on turn 120 so things similar to this do occasionally happen. I cannot begin to guess though as to all the many causes. Is there a patch that makes awareness on at all times for the player? (I know it would be OP to have awareness on turn 1 but maybe it would help spot/test these occurrences just hitting next turn rapidly with awareness on?)
(July 1st, 2016, 18:10)namad Wrote: I sieged his city after the second save. He never moved out. Then died. It took me dozens of turns because I tried to expand a bit at the same time. I also intentionally let his settlers slip out and settle outposts in the hopes I'd get them as free cities, and I did, but the red player stole them from me. On turn 110 (I took a break from the game for real life and I am not yet on turn 120) he's been dead for a while and I'm fighting to keep his cities from an attack by the red wizard.
I think you should change population 2 to population 3. The problem is that if you only have 3 population you build your settler too slowly, then you go down to 2 population and build everything too slowly. If that minimum population number to build settlers got +1 or if every race in the game got +10 growth I think the problem would go away.
Even though in the most recent save the wizard doesn't exist you can surely see the file. I'm doing quite poorly overall.
edit: This might be the only time I've seen this happen to an ai. However I have seen ai's die before turn 50, and I have seen ai's with almost nothing on turn 120 so things similar to this do occasionally happen. I cannot begin to guess though as to all the many causes. Is there a patch that makes awareness on at all times for the player? (I know it would be OP to have awareness on turn 1 but maybe it would help spot/test these occurrences just hitting next turn rapidly with awareness on?)
To enable awareness for debug purposes, go to the MAGIC screen, hold alt, and type in RVL.
Sawmill generates 8 production, assuming a 50% AI bonus and no terrain, it's 12. Not high but not that bad and completes a settler in about 12 turns. On higher difficulties it's even faster, and if the population growth is bad, there is a good chance it's caused by mountains which give a lot of production. In general, terrain with 0 production bonus is rare if even possible, so it's safe to assume they take 8-10 turns for a settler (at +50% production for difficulty which is the value planned for use on hard in 2.5). Other stuff generally costs less so is faster.
If you took the new cities, don't be surprised they don't have any o_O
Dark elves are slow, if not producing settlers they would be even slower. The AI doesn't rely on population all that much, they have production bonuses and enough gold to even buy buildings every once in a while.
Mountain terrain is ideal for the AI to build the settlers fast, and population growth probably won't keep up with it.
Under normal conditions the other cities are supposed to grow and eventually produce armies, buildings or take over building settlers, or the continent should fill up and the AI then stops producing settlers since they still have some they cannot use up - unless they have ships or towers in which case they'll get transported elsewhere for use.
During a war if the settlers are killed, they'll replace them and the population loss will not stop, which is unfortunate but I have no way to prevent that. Reducing the priority of settlers would make the AI a lot weaker in the early game, and making many gives them a chance to build a city somewhere the enemy doesn't find until it's too late to destroy.
In general the AI works with the assumption that they haven't yet lost the game, if they're reduced to 1-2 cities with any units leaving immediately killed, they have no way out anyway so it's pointless to make it smarter to handle that situation. If another player is capable of preventing expansion it makes no difference if the lost unit is a swordmen or a settler or even a warlock on the long term.
edit :
Quote:Also I found a massive problem with myrran ai's which explains why I found a myrran ai with 3 cities on turn 120-150ish.
Quote: On turn 110 (I took a break from the game for real life and I am not yet on turn 120) he's been dead for a while and I'm fighting to keep his cities from an attack by the red wizard.
Now which one is it...I'm confused but think there is no problem here, you just destroyed their cities and found them before they had a chance to stop you.
looked into the other save again. You found them on turn 40 and they already had two cities and a third settlers sent out (I assume Leer was theirs and you took it).
so here is what happened :
first they started on mountains, so low population growth kept the capital at 3 pop (they only made 3 settlers total at this point yet were at 3 pop)
They built a sawmill (few turns) then a settler (another few turns, expecting like 10 total).
The settler got sent to build Leer, 14 tiles away from the capital. Normally they would go to a closer spot but that was far superior : ores, river, grasslands, wild game while all nearby area had mountains and desert and forests only.
So at about turn 25 they make a city. Elves grow slowly, it turns into a city at, idk, turn 30? spends another 5 turns building a sawmill then another few for spearmen or whatever...and we are at turn 43 when you arrived so they lose this city due to warlord+buffed ultra elite flyin swordsmen.
There is no way the AI can beat that on hard.
Actually several of their settlers died to mountains. As in they founded an outpost but the food bonus and max pop was too low due to mountains and lost the settler. A mountain start that lets you build settlers faster than your population grows is likely to have settlers settling in mountain locations. I'm not sure I actually killed any settlers at all. Their outposts that lived took dozens of turns to turn into hamlets by which point not only were they banished but another player came in to steal the hamlets I'd intentionally ignored for dozens of turns (which triggered the final death of the banished blue mage). Mountain starts I guess are just very bad for dark elf ai?
I don't see a solution to this though other than to disable outposts from killing themselves entirely for all players.
While I'm complaining would it be possible to nerf AI combat against neutrals? A great drake and 8 fire giants were taken out by a stack of berzerkers with almost no loses to the berzerkers in this game. (although since I am not all seeing I could be wrong about this statement).
