Login

Seravy

Caster of Magic 0.8 is now available at the usual location!
This is the last "big" update.
I've finally fixed all the bugs I was able to, and did all the changes to the game I felt necessary.
The work however does not stop here.
Next up on the list is playing the game, making sure everything works (there has been tons of major changes in these 2 updates), fixing whatever problems might be remaining, addressing possible remaining balance issues, and after all that, release 1.0 and then it'll be "completed".

A warning, as I was working nonstop for the past week and had no time to play, nothing was tested beyond the bare minimum of loading the game and forcing the new feature to show up somehow. This usually is sufficient, but as rather complex systems were changed in both of 0.7 and 0.8, some changes might have unwanted side effects that aren't immediately obvious.

Keep an eye on the AI and please report it if you notice anything seriously wrong with how they move their units overland, and/or regarding diplomacy, aggression, and personality/objective having the expected effect on these.

And now, for the list of changes and highlights.

-AI will now concentrate more forces in their cities to make the game harder. As production is more accessible in this mod, Insecticide "rush AI" philosophy didn't work very well with it. Adding a few more halberdiers only takes a few turns for the AI, not a significant delay, especially when they can also summon multiple cheap units per turn. Contrary to that, units leaving the capital to form stronger armies outside for earlier expansion wasn't a very good tactic. I guess it was good in the unmodded game, where units are slow to produce and valuable, but not in Caster where having enough soldiers to both defend and attack is normal.
-AI is now finally, finally able to split up stacks defending their nodes and send out the stronger units into a stack used for offense. This was, along with the previous, my top two goals for improving AI, and it sure took a while to do. Those disassembler files Serena shared containing all of Kryub's info sure helped a lot! No more top ranked heroes and storm giants will be wasted on guarding the sorcery node next to the wizard's fortress, at least not for long periods of time.
-Although untested, the AI should also be able to recognize poorly defended player towns and redirect patrolling units to them from a larger distance, which should help powerful AI units return to the game from node guard duty if needed. I really hope this one works as it would make the AI a lot better at stopping player invasions.
-Exp tables and item rewards have been slightly adjusted. It's still easier to level and items are still better than originally, but the changes are more balanced now. Also fixed hero level up dialog to contain only valid information (ability details simply do not appear at all)

That's it, not all that much but it should significantly improve the game, and more importantly it removes AI behavior that is not only easily abuse-able but blatantly obvious to players, too, which might hurt the enjoyment even more than actual severe bugs.

Have fun, and I'm expecting some feedback on these two updates. AI is a rather subjective matter as well as being the core of the game, so it would be nice to hear if it feels better, worse, same, stronger, weaker, more interesting, or anything, really.

Seravy · October 28th, 2015, 19:24

0.81
-Grand Vizier is unable to build Housing if populaton>=5
-Banishing a wizard now uses diplomatic event code 6 (thanks for helping), 0 (no event) or 7 (overextension warning) based on the relationship of the banished wizard with the others. Defeating a wizard still uses 7 (overextension) for all cases. Let's face it, if you take all their terriroty to trigger this, you ARE overextending, even if they are happy about the other guy being gone forever.
-Wizard's Guild, Miner's Guild, Mechanician's Guild maintenance is lowered a little.
-Slightly adjusted AI maintenance difficulty modifiers.
-Experimental : Check for more defenders needed is now run every turn instead of every 3rd (reversal of insecticide change)
-Adjusted defenders needed formula for towns : (pop/2)+4 units.
-AI to AI diplomacy : If already having an Alliance, don't roll for forming a new Wizard's Pact.
-AI to AI diplomacy : randomness reduced, difficulty level modifier increased.
-AI to AI diplomacy : “willingness to consider treaties” modifier has halved effect, it is a -200 to 200 value that changes very quickly during the game, adding it raw totally unbalances the check.
-AI to AI diplomacy : Alliance has to pass a check for the sum of modifiers >=200. Original was same amount for both Alliance and Wizard's pact (150), ironically this means the AI always formed an Alliance first, and next time demoted it to a Pact due to the other bug, then back to Alliance, oscillating every time the roll was successful.
-AI to AI diplomacy : AI is not able to propose an Alliance or Wizard's Pact if they are already at War. Peace has to be offered first.
-AI to AI diplomacy : AI is not able to trade spells with people they are at War with.
-AI to AI peace treaty : now uses the default relation based formula instead of..some unknown and seemingly unrelated variable
-Chaos Channels : Fire Breath is now strength 3. Fixed help text on Demon Skin.
-AI combat unit movement targeting : Units that have a Breath, Thrown or Gaze attack receive an extra +12 priority boost on targets reachable this turn.
-Fixed incorrectly working new feature : When AI has a new unit in a full map cell, always the weakest gets pushed away based on unit COST and level (instead of maintenance). AI relies on this procedure to make sure best units stay for protecting towns, so fixing this was quite important.

0.81 is up
Most important change is I looked into the AI to AI diplomacy procedure in detail and found a huge amount of bugs which I fixed.
The AI will now
-Not trade or make pacts with each other during war
-Not always make alliance before wizard's pact due to equal needed relation but more priority
-Not make a wizard's pact if already allied
-Not base peace decisions on unknown purpose unrelated variables instead of relations
-Not trade more often the worse the relation is instead of the other way around

Seravy · October 29th, 2015, 03:14

0.82, a small but important fix :

Quote:-Removed a redundant and misplaced roll for AI peace offer check that could redirect it into trying to form a treaty instead which then obviously fails due to war
-Removed condition “nonzero event required” for peace checks. War prevents negative events and positive events are rare during it.

MrBiscuits · October 29th, 2015, 16:53

I've not had much time to play this week, but I had a quick go last night and noted a few things from early impressions.

Being able to choose life and death is great, although it seems to make all wizards neutral towards you at the start.

The map features like towers and nodes seem to clump together on tiny map size which they don't in vanilla. Have you reduced the total land area on tiny maps? It looks like there is significantly less land, I'd say around a quarter compared with vanilla. I like small maps but these are a bit too small. I think they need to be at least doubled.

Early building seem really overpowered eg. Sawmill gives +8!!! Makes the first building choice not really a choice as you always pick sawmill. It also means you can have most of your workers as farmers early on and still have a high production so the balance between production or having a large army is lost somewhat. Also it negates the advantages for high production races like Klackans. Another consequence is that unrest is less important as you get most money and production from fixed rate buildings rather than multiplying worker yields, so who cares if 1 or 2 aren't working.

I'd tone it down the sawmill to +2/3 or make it a percentage rate again or maybe add one production to workers so the increase is gradual as you grow your cities. Also it costs 1 turn for a spearman early on, so you can build an army when you see raiders coming rather than have to plan ahead. Later on you can get a swordsman in one turn too, making spearmen completely obselete.

Same deal with marketplace +8 seems far too much.

Rewards from map features seem better so far, I got a nice orb from killing hell hounds in a node. Have you changed this or was it coincidence? It's good anyway.

Ghouls can use ranged attacks even when an enemy is adjacent. I didn't think this was possible before and casters had to melee instead.

Astrologer seems to actually reflect the game more than it used to.

Reduced food maintenance makes armies even larger alongside the production boosts. Loads of troops to move around each turn which I'm not a massive fan of.

8 magic attacks for heroes is far too many.

I got Grayfairer and he had scouting 4, meaning he can see half the map immediately. Is it 16x16 squares?
Also noble gives 20 gold per turn which seems too much to me. I'm rolling in gold immediately before even building any buildings.

Mana focusing gives you a nice early mana boost which is very useful now.

Got offered a wizards pact early on which was great. I'm wondering if all these pacts will create a much more peaceful game.

Trireme has scouting 3 so again can see a lot of the map without exploring much. Sort of spoils the exploration part of the game a little bit.

Increasing settler costs is a good idea as expansion was too fast previously. Didn't really see how that affects the AI at higher levels though with their production bonuses which probably means they take over the map even easier than before.

Towers, lairs and keeps seem a lot weaker, mostly ghouls and guardian spirits.

Hoping to get to play a lot more over the weekend and try out some of the excellent new features.

Seravy · October 29th, 2015, 17:26

(October 29th, 2015, 16:53)MrBiscuits Wrote: The map features like towers and nodes seem to clump together on tiny map size which they don't in vanilla. Have you reduced the total land area on tiny maps? It looks like there is significantly less land, I'd say around a quarter compared with vanilla. I like small maps but these are a bit too small. I think they need to be at least doubled.

Vanilla does not have Tiny land size at all. Tiny is about 40% smaller than "Small" I think it's 150/250/400/600/800 for each option. Stuff gets clumped up due to lack of space. Vanilla only has Small, Medium and Large, Insecticide adds Huge on top of those. Tiny is new (although some other mods have it)

Quote:Early building seem really overpowered eg. Sawmill gives +8!!! Makes the first building choice not really a choice as you always pick sawmill. It also means you can have most of your workers as farmers early on and still have a high production so the balance between production or having a large army is lost somewhat.

That was the intention actually. Game was far too slow paced at the start and you could build close to nothing in insane amount of turns due to lack of population to have any workers at all. Not a fan of "Hit End Turn 50 times then start playing". Don't worry the AI also knows this and will always build the Sawmill first.

I received a suggestion to just have a flat +X production without needing a building instead (just have it appear for free) but I think having a building making it is better. 8 is a considerable amount, but it's still only one swordsman unit every 3 turns and that's...nothing much, really.

Quote:Rewards from map features seem better so far, I got a nice orb from killing hell hounds in a node. Have you changed this or was it coincidence? It's good anyway.

Yes I changed it, items are set to be 60% better than originally, and monsters (at least the harder ones) worth more treasure points too.

Quote:Ghouls can use ranged attacks even when an enemy is adjacent. I didn't think this was possible before and casters had to melee instead.

Any unit does that, unless their melee is much higher than ranged (I think double), or the target is immune to missiles.

Quote:I got Grayfairer and he had scouting 4, meaning he can see half the map immediately. Is it 16x16 squares?

Scouting 4 extends radius by 4 so adding the default 1, he has a range of 5. That's roughly the size of the game window, I think 11x11 if we look at it that way. It's far from half the map, but half of a smaller continent, yes. The map is 60x40 large I think.

Quote:8 magic attacks for heroes is far too many.

It actually depends on the hero, 8 is most common but some have 6, 10 or 12. Archers also have this many, so I don't see much of a problem. Same archers even have Caster so it's not like that makes them better. I guess no range penalty helps but hit is easy to get for heroes. To be honest I have more problem with melee heroes that can attack an infinite number of times, and with enough defense stats or a first strike/breath/thrown ability the can do it without taking damage at all. Casters are at least vulnerable.

Quote:Towers, lairs and keeps seem a lot weaker, mostly ghouls and guardian spirits.

Towers are random (there are only 6 in a game so the chance of them all rolling weak is not that low)
The others, roughly half are set as "weak" the rest as "strong". It's disappointing when there is nothing that can be defeated for the first 100 turns of the game.

MrBiscuits

(October 29th, 2015, 17:26)Seravy Wrote: Vanilla does not have Tiny land size at all. Tiny is about 40% smaller than "Small" I think it's 150/250/400/600/800 for each option. Stuff gets clumped up due to lack of space. Vanilla only has Small, Medium and Large, Insecticide adds Huge on top of those. Tiny is new (although some other mods have it)

Ah I see. I was playing the Zitro mod that has it, but had significantly larger land than the tiny on yours (which I'd say is too small to be playable). If you could up it by 80% of so it would probably be perfect.

Quote:That was the intention actually. Game was far too slow paced at the start and you could build close to nothing in insane amount of turns due to lack of population to have any workers at all. Not a fan of "Hit End Turn 50 times then start playing". Don't worry the AI also knows this and will always build the Sawmill first.

I understand the reasoning, but it seems to put the whole early city choices out of whack, it means assigning workers to production is pretty much pointless as their contribution is much smaller than the sawmill. Same with managing unrest. It also means you can build every building very quickly reducing choices and making the "early" races like barbarians much less useful. I'd favour population growing faster and perhaps give your capital a few production buildings at the start to get things moving faster instead.

Quote:I received a suggestion to just have a flat +X production without needing a building instead (just have it appear for free) but I think having a building making it is better. 8 is a considerable amount, but it's still only one swordsman unit every 3 turns and that's...nothing much, really.

But before you had to decide whether you wanted that extra swordsman, but now it seems that you can just build as much stuff as you want.

Quote:Yes I changed it, items are set to be 60% better than originally, and monsters (at least the harder ones) worth more treasure points too.

It's a lot better, although I did get some very nice things from very easy nodes.

Quote:Any unit does that, unless their melee is much higher than ranged (I think double), or the target is immune to missiles.

Really? I'm really thought that bowmen always go for melee only once there is an enemy next to them in vanilla.

Quote:Scouting 4 extends radius by 4 so adding the default 1, he has a range of 5. That's roughly the size of the game window, I think 11x11 if we look at it that way. It's far from half the map, but half of a smaller continent, yes. The map is 60x40 large I think.

Yeah I was exaggerating a bit, but it really seems to be far too much for one unit. It's better than a oracle which can't do any of the stuff a hero can do and which won't show up in the first few turns like Grayfairer did.

Quote:Towers are random (there are only 6 in a game so the chance of them all rolling weak is not that low)
The others, roughly half are set as "weak" the rest as "strong". It's disappointing when there is nothing that can be defeated for the first 100 turns of the game.

I agree it is annoying in vanilla when you can't tackle any nodes at all for ages, but it is nice when you can plan ahead to the midgame about which features you want to tackle. Perhaps make it 30% weak and 70% strong or something (it is a shame there is no "medium" option).

Quote:It actually depends on the hero, 8 is most common but some have 6, 10 or 12. Archers also have this many, so I don't see much of a problem. Same archers even have Caster so it's not like that makes them better. I guess no range penalty helps but hit is easy to get for heroes. To be honest I have more problem with melee heroes that can attack an infinite number of times, and with enough defense stats or a first strike/breath/thrown ability the can do it without taking damage at all. Casters are at least vulnerable.

The problem is that once casters are leveled up and kitted out 8 shots makes them able to just machine gun everything to death from the back row. I'd say 4 for magic casters and 6 for bows.

Seravy

Quote:Yeah I was exaggerating a bit, but it really seems to be far too much for one unit. It's better than a oracle which can't do any of the stuff a hero can do and which won't show up in the first few turns like Grayfairer did.

He is the "Scouting" hero, even was in the original. Other heroes only have Scouting II for a total of 3 sight range. You got lucky, in most games you need to explore for quite a long time before having more than Scouting II.

Quote:I agree it is annoying in vanilla when you can't tackle any nodes at all for ages, but it is nice when you can plan ahead to the midgame about which features you want to tackle. Perhaps make it 30% weak and 70% strong or something (it is a shame there is no "medium" option).

I tried and it was still too much "strong" ones, although strong is adjusted by difficulty too. Higher difficulty boosts the strength and value of these. I can adjust weak to be somewhat stronger but it felt about right for me so far. I still found a lot of powerful items and stuff, usually even 2-4 books per game and I only cleared one plane on average.

Quote:The problem is that once casters are leveled up and kitted out 8 shots makes them able to just machine gun everything to death from the back row. I'd say 4 for magic casters and 6 for bows.

Well, that's what ranged units are for. They are weak to fast moving units and direct damage in exchange, which melee units are not. 4-5 moving units like cavalry and most racials can reach them in 2 turns or so, although the hero can outrun them with some movement items unless outnumbered, but running and shooting while keeping a safe distance is tricky.

8 might be a little bit high but any less than 6 would feel too restrictive. Maybe 6 could work better, I'll consider it. I'm still more worried about melee heroes, with an armor and some decent melee abilities they can become unstoppable easily.

zitro1987

Seravy, the problem with your sawmill is that it doesn't give you a fixed +8 production, it gives +8 * terrain bonus.

I'm thinking of a test run with your mod and that's the first thing I noticed: my sawmill gave me +12 production! (city surrounded by mountains)

Your typical terrain bonus is around 20-35%. If you plan to give +8 production in average, please change to +5 or +6

Seravy · October 29th, 2015, 21:01

(October 29th, 2015, 19:12)zitro1987 Wrote: Seravy, the problem with your sawmill is that it doesn't give you a fixed +8 production, it gives +8 * terrain bonus.

I'm thinking of a test run with your mod and that's the first thing I noticed: my sawmill gave me +12 production! (city surrounded by mountains)

Your typical terrain bonus is around 20-35%. If you plan to give +8 production in average, please change to +5 or +6

Is that few points of difference really that much of a problem? This would take some time to change, including all the displays and everything, so I want to be sure about it before I start.
I don't see how even 10 or 12 is that significant when a full town is in the 40-80 range and anything remotely good and valuable costs 100+.

If this was a player only advantage then sure, it would be something to worry about, but...AI gets the same amount (well several times more actually because of the difficulty modifier) and will also build it first like human players. In fact, Sawmill helps the AI a lot to get a good early start, which helps games become interesting. Weakening the Sawmill would directly weaken AI players.

Also consider that a Sawmill is not very cheap (costs 100 to build) at 8 production total, it would take 12 turns to earn back the investment before getting anything out of it...and that's not counting the rather big maintenance of 4 gold/turn. If I also consider that 2 gold buys 1 production, then a sawmill is actually producing 2 less advantage per turn than how it appears. Which would make it only a flat +6/turn with your suggestion, at a cost of 100 that would require 17 turns to reach the break even point before it starts to be profitable. That's a lot for a building that was designed to speed up the early game, too much, even. And I didn't even consider that 4 gold can equal 8 production if doing it the trade goods way, in which case the sawmill roughly pays for itself, getting you only a per points extra.

I haven't actually calculated but building a Builder's Hall and housing might be almost as effective by building up your workers, and the population is more versatile than the Sawmill, can get you taxes, and has no maintenance. I admit there aren't really much good options other than these two for the start, but you get plenty afterwards. Although slow growing races might even want to consider a Granary/Marketplace/Farmer's Market instead, not sure. Housing might be better than Sawmill for fast growing races now that I think about it. In just 12-16 turns you get 4 population which will get you 8 production, some taxes, and no maintenance, only needing 4 food. Building the Sawmill takes this long too for roughly the same bonus, unless you buy for gold which does let you get one earlier but then no chance for a hero offer as you lack gold.

Overall, while I admit Sawmill is quite powerful, I'm having doubts about changing it.

Seravy · October 29th, 2015, 21:20

0.83

-Fixed bug : If a treaty offer was immedaitely accepted, before the secondary offer containing gold or spells was made, the player still got the gold or spell anyway, from the offer they never saw happen.
-AI treaty offers now contain 20*random(10)+(AI total gold)/8 gold if that option was triggered, instead of (AI Army Strength/(random(3)+1)+CurrentTurn) which was a really poor amount.
Note that the AI still does not lose the offered gold, and does not require to actualy have it in the treasury for the offer, but the player will receive it properly. Due to code space and time issues, I rather not try fixing this one. The AI gets more than enough gold advantage, this small amount doesn't really matter.
-Fixed bug : Player does not get the offered spell if spell code was greater than 128.
-Fixed bug : Spell name is not displayed if one is offered in exchange for a treaty by the AI. (Caused by no text being displayed at all unless gold was offered...which is not possible if you already get a spell)
-Fixed bug : when requesting a peace treaty, chance of receiving a spell or gold offer is higher the stronger the AI is compared to the player instead of the other way around.
Formula still does not match other treaty offers but at least both reward the player for being stronger now.
One is 30*Power/PlayerPower<random(100), the other is 30*Playerpower/Power>random(100).
The former can have a 0% chance of success if AI is strong enough and 100% is unreachable, the latter is 100% chance if the player is strong enough and 0% is unreachable. Overall the new treaty offer has more favorable odds towards the player than the peace treaty offer.

Seriously, just how many more bugs will there be in diplomacy? This is getting ridiculous, did a pile of monkeys made this game or what.
It's 3am so I'm quite a big tired, and have not time for extensive testing now. Please report if you find any serious problems and I'll attempt to fix them tomorrow morning, although let's hope there aren't any. I double checked when restoring the code from the test code that forced the treaty to trigger every turn, it should be all right.

Navigation

Extra Menu

About us