Ah, I couldn't remember if they still had that. Thanks Seravy.
|
New experimental version
|
|
Really? I had a distinct impression to start losing more buildings from neutral cities too with the change, but it might be observer bias.
(September 13th, 2017, 10:50)Nelphine Wrote: I dunno, I've chewed through 13000 mana in 3 months before, and that was only around 1409. I've probably done higher. Mana (and therefore gold) is extremely important late game. That's what makes it unimportant. You get 6 cities in a turn? You've won, some time ago already. You catch an early city at the beginning despite: - AI cheating - no possibility to garrison -> your cities are a stack magnet - having to move around all the AI stacks that are crowding your continent - scarce initial resources ...Well you deserve the city if you don't raze it then :D Realism: you go through a guild's treasury, pillaging it, you might well find some valuables. Your calculations seem to be quite off, 1k is if your troops pillage a top end game city perhaps, again: I'm proposing something for the early game. And how comes you say "for razing"? I've only spoken about pillaging - razing already has more gold I believe? Maybe a way to accommodate both PoVs - I guess that too easy an end game might also be an issue - would be to have a diminishing return mechanic for conquering. Some ideas: - more chance to raze the more buildings in the city - cheaper buildings have a lower chance to be razed than the expensive ones - getting as low as 0 for 100 cost stuff with no troops moving in the city during combat and no spells maybe (I've got this feeling that more destructive spells increase the raze chance, is that the case?) - higher chance to raze the more advanced the game is - my coin for pillaging idea being reduced logarithmically - a combination of the above This could allow you to remove the difference between neutral and AI cities too, as the AI cities are usually more developed.
Checked, the 50% rate for neutrals is still in the code.
...more importantly, this version is still new. After we each played a dozen games, we can decide if the system is good or bad and think about a solution if it's bad. I can't imagine you already played enough completed games in EXP13 to judge it - and I definitely did not.
September 13th, 2017, 13:45
(This post was last modified: September 13th, 2017, 13:49 by Nelphine.)
I don't have any super objection to having higher destruction rates for cities with more buildings (also allows us to remove the difference between difficulties, as higher difficulties will have more buildings).
However, I'd still like to understand - why do you need to be (close to) guaranteed to get the important early buildings, above and beyond getting the city itself? Just getting the city means: no settler produced, no time to get the settler into position, no time waiting to turn from outpost to hamlet. Just this advantage is huge. What justification do you have (from a balance point of view) that you need more than that? Remembering that all of those things also benefit from AI cheating bonuses, except settler speed; so you can already conquer more cities than on lower difficulties if your strategy is based around conquering cities. (This all assumes a good ratio of economic gain between razing and conquering. I talked about razing because if you raze the city, you lose all of the benefits I just mentioned, plus any buildings you do get plus any taxes from population. That's a HUGE loss, so razing the city needs to give corresponding economic rewards, or you won't ever raze a city, and we want razing cities to be an important option, but a less favorable one. Thus, I assume if you are going to give gold for destroyed buildings when conquering a city, you need yo give something to razing cities, or it will never be chosen.) If the answer isn't based on balance, its based on enjoyability of gameplay, then I have no direct response, except to say I don't think its balanced. Quote:I don't have any super objection I do. The same destruction procedure is used by Earthquake, Call the Void, Meteor Storm, Raise Volcano and god knows what else. Any change to it would apply to every single source of building destruction in the game. Meanwhile, done the second random game. Rolled 4 Chaos, 2 Nature, 1 Death, Spellweaver, Guardian, Famous. Not the best but...it's literally the same as before with 4 books swapped for 4 more picks in retorts. Spellweaver is powerful enough on its own to make a difference, and Guardian is...something I consider so good I don't even pick on lower difficulties. However both these retorts go very well with Chaos magic although they don't have any synergy with each other. Map rolled was Tiny - Guardian is many times more relevant there as losing cities is not an option when the entire map has like 12 cities/plane. On top of that, Poor so not much minerals to mess with balance - I got Orihalcon on my starting city despite it. Even my race was the same, so this game satisfies the conditions you set and I think this should have been winnable on Master, but I was playing Expert anyway. (yes, Famous isn't very useful but it does help the economy in the early game so it's not worse than an extra book until much later...) Now, the game... First, found an enemy wizard fortress about 8 tiles from mine on an adjacent island. First target! My sprites didn't find any useful nodes or lairs for a while but the last patch of land I actually scouted had a Sorcery node - and it was an easy one! With only 2 sprites I was able to clear it, while preparing my Ghoul army to take over the enemy wizard's island. Spellweaver or not, the AI summoned things much faster than I could, but ghouls in large enough groups beat bears and sprites so...good. I was able to take over the two cities that wizard had built on their island, only the capital left. Meanwhile the 2 other Arcanus wizards showed up and started random attacks. I didn't lose anything of high value, but I did lose my smaller settlement (I had room for two on my island, the second was a 9 pop tile, not very good) and I did also lose the small one conquered out of the two. There were far too many units walking around, and they were faster than my ghouls, and I couldn't spare enough on that island to protect both and still have something going around killing things. Took some time but I completed summoning 9 Ghouls. Flandre (the first enemy wizard) still only had 9 Sprites in their capital - easy win for ghouls. Or so I thought. Unfortunately the AI is no longer dumb and used Mana Leak, plus had city walls, so my Ghouls lost and 5 sprites survived. This was clearly my mistake (not used to seeing mana leak from a wizard using sprites but this one was nature/death). Keeping my positions wasn't easy during this, as anything that wasn't a large stack got picked off by water walking or wraith formed bears (3 move) or spiders (5 move), or heroes (3 move). I also lost my node twice, leaving me without any additional source of magic power for prolonged periods. As I saw things are starting to turn ugly unless I do something, I pushed research - Chaos is stronger if you get higher tier spells. Was hoping for Fire Storm and got it. Built Sage's Guilds everywhere so I researched in in about 20 turns. My casting skill was a horrible 20 (The single node I sometimes had didn't produce enough to raise skill - I could only afford producing mana to summon ghouls and use combat spells.) but luck was on my side : A Sorcery conjunction hit. Halfway during it I was even able to reclaim my node because the hero...was called away for doomstack-building, leaving it empty. That raised my skill to a massive 76, so I was able to finish 9 ghouls and fire storm without having to wait another 5 years. Burned away all the sprites, only one hero left. Not even high level or anything, easy, right? Well, no, not really. It had an Elemental Armor weapon and 11 resistance so ghouls did zero damage in ranged...and in melee as well unfortunately, even though the hero only had 5 shields - city walls raised it to 8 and that's very hard to damage with 4 sword ghouls. Fire Bolts could have won the battle anyway but...I couldn't afford them. I did have 70+ skill thanks to the conjunction but no mana, and I assumed it's an easy win so didn't wait for more. But hey, I had 3 undead bears close by. In just another turn, I could attack using them - 7 swords from bears with a hit bonus has to kill that lame hero...guess what? In that one turn the AI summoned a Nigh Stalker and a unit of Sprites in the city. The Sprites + Lighting+spells was enough to kill my bears and with the Night Stalker inside, I don't feel confident in trying further attacks - I did have 2 Gryphons ready to strike as plan B another turn later... Either way, this is how the graphs look like : Despite the Tiny land and Poor resources, the AI is way ahead, so much that I conclude it's my loss. This time I tried offense and conquest and it failed - last time I tried peaceful expansion which also failed (graphs looked similarly). And while I did take some risks by those attacks I lost, the AI is developing so much faster I feel pressured to take those risks - if I don't attack, I fall behind even more. Also, luck was on my side, the wizard I rolled was above average (heck, I would have played this on Extreme last year, minus the Famous - Spellweaver Guardian has amazing synergy with Chaos.) and the map denied resources from the AI (Tiny/Poor) not to mention I had a free node from Sprites. And it wasn't enough to beat the level 2 below the highest, probably wasn't even close. Second red stamp to the "Expert" column on the testing results. Not a good sign. ...or should I continue? My capital can produce Gryphons so maybe combining those+fire storm could still win.
...decided to look at the map as it seems hopeless to win. The Myrran wizard is about 25% more powerful than blue. Blue had 3 nodes, and pleny of Night Stalkers and Shadow Demons. Red has Gorgons and Lizards.
Speaking of which, I think the AI is now strong on Tiny maps instead of weak, thanks to the doomstack procedure and the overland spellcasting which casts more "waterwalk" type spells. Moving between continents or even defending anything is close to impossible when stacks of water walking spiders or night stalkers can appear any time on the sea.
Hard for me to judge that game. I consider chaos hard mode in the first place, you've got a wasted pick in famous, and a wasted pick in death (though getting ghouls was lucky, and could have made up for that), and only 2 picks that synergize in retorts, even if you consider all the retorts good. So, I'd say that should have been a very hard game to win with that build, but, that's based on my opinion of chaos which isn't entirely fair.
To be honest, ghouls in 1410 is a problem. Maybe I'll try some random games whenever I get my surrender from the AI. (Though I'll have to ixnay chaos due to personal bias.)
I think one reason why Chaos feels less powerful is the doomstacks.
Before those, the AI was good on defense but weak on offense. Chaos is strong on offense, so it can crush the AI's defenses, and that's all you need to win. Now, the AI is still fairly good on defense but even better on offense. So games don't reach the point where the "strong on offense" comes into play. Tho, the irony...if I had known Disrupt, I could have won that battle against that fortress. Not sure if winning that battle would have been enough to win the game but...enough to at least continue I guess. But I didn't have it so I lost to a city wall...(that +3 defense is brutal in the early game) But yes, good catch, I shouldn't have summoned ghouls. I should have summoned gargoyles. Unfortunately they showed up after Fire Storm. I should have picked some common to reveal all my uncommons first but I was already pressed for time. So, maybe I wasn't playing perfectly but perfect play should only be required for the last two levels of difficulty...a few smaller mistakes should still be acceptable on a "medium" difficulty level. Now that I think about it... We had hard, extreme and impossible. I wanted master to be same as extreme, lunatic same as impossible, and expert halfway between hard and extreme. So we set up numbers for the AI matching those : master = extreme, lunatic = impossible, expert ~= hard+extreme/2. ...and that sounds correct except...it misses the point completely. If I want "master = the extreme level I know" we'd need numbers much lower to account for smarter AI overland play (doomstacks, new procedures, much less wasteful overland attacks - player cities are no longer black holes that kill incoming AI stacks, as any attack is a major threat likely to kill parts of the garrison if not all) and more balanced basic game mechanics which makes it harder to gain advantage (amp tower cost, destruction rate, sorcery no longer being a pushover in early game, idk what else?). Likewise for all levels, maybe excluding "Advanced" which did go down by a little from the original "Hard". I'm going to play a lot more before any further difficulty changes, but I suspect the end result will be something like "lunatic = what master is now", "master = what expert is now", and "expert = something between expert and advanced", possibly with advanced becoming even weaker. ...and I really hope at least Normal is ok - Hadriex losing one of those was alarming and that was pre-doomstack... I strongly recommend you to try some (a dozen) Advanced/Expert difficulty games with medicore or subpar wizards, or random generated - while Lunatic might not have changed all that much, I feel these middle levels got a hell of a lot harder due to AI changes (more waterwalking in the first 50 turns is one of those). In fact, while it might be stretch, I feel the AI is now playing the game better than a medicore human player, mainly due to complexity : The AI will never miss a chance or interaction while human players usually do miss something (myself included, I forgot about the possibility of Mana Leak). (oh and before we conclude I had bad luck with everyone having waterwalking or wraithform on Tiny, there was no Sorcery wizard so no "super nagas". I guess Life and Chaos could have been better opponents with no water movement spells but that still means 60% chance per picked realm to have one)
PS : if you want to watch, I recorded the game, here : https://www.youtube.com/watch?v=_Il3KEUJ...e=youtu.be
I definitely could have played better but I don't think I was doing that poorly. At least not so much to say it's insufficient for this level of difficulty. For Master, yes, but this is Expert only. |
