Autsj posted:So I made the time to play 25ish turns (and then fast forwarded another 25) and I was kinda bummed by what I saw from the ai. Sadly that sounds similar to what I saw in my 3v4 dragon war game which I started early while playing the open beta. The main cause of the AI not founding cities was losing armies (including their leader) while clearing neutrals or defending vs hostile neutrals.
|
|
# ? Aug 16, 2023 23:29 |
|
|
# ? Jun 10, 2024 18:03 |
|
Lowen posted:Sadly that sounds similar to what I saw in my 3v4 dragon war game which I started early while playing the open beta. Nah that wasn't it, they all had troops and their leader running about and like I wrote they were continuously sitting on at least 1 outpost that could be upgraded to a city; they just never did until something (way later than it ought to) triggered/caused them to do so--and then they immediately added another outpost to sit on. It looks to me like the ai follows a pattern of always putting down an outpost (or 2, in VH case) a little away from their borders (by rote, regardless of expansion plans or cap), then when the desire/trigger for expansion comes up and they have room in the cap, upgrades one and builds a new outpost a little further out. Basic rote behavior that makes sense for an ai within the game's system. But for some reason, the Hard ai never triggered the expansion part of the outpost system until way late, in this case turn 40 (which is basically not playing at all). While the Very Hard ai did better, it was still doing so significantly slower than any semi-experienced player would (turn 30ish for 3 cities is a good 10-15 turns too slow). Nothing that I could tell was limiting them resource wise, only Imperium could be the problem here and they didn't seem to have any real sinks there, so it has to be a trigger/behavior problem. Fitzy Fitz posted:I haven't done any tests, but I've noticed much more aggressive behavior than that by the brutal AIs I've played against. I definitely noticed the ai venturing much further from their cities/borders than before (which was mentioned in the patch notes). I think that should help with their exploration/making contact and aggression. Though as a note: Brutal is a map setting, not an ai setting; Brutal includes lower starting resources and higher world threat, which usually handicaps the ai somewhat due to attrition. (though I think they now get some bonuses based on world threat to compensate for that) Autsj fucked around with this message at 00:31 on Aug 17, 2023 |
# ? Aug 17, 2023 00:03 |
|
I haven't done any tests, but I've noticed much more aggressive behavior than that by the brutal AIs I've played against.
|
# ? Aug 17, 2023 00:17 |
|
The AI bugs in this game are gonna drive me into an early grave, not gonna lie Thanks for the info guys, we'll look into it! General tip if you want the AI to put up more of a fight is to play on a smaller map, the AI has a limited engagement range, so being closer will cause them to engage with you much more frequently. We supplied the large map sizes because people tend to like them, but in general the game plays better on smaller maps anyways - enemy empires are smaller and closer, so less of a slog to conquer, travel distances are shorter, etc
|
# ? Aug 17, 2023 08:32 |
|
Yeah TBH most games wrap up around turn 70 or so when your builds fully come online and a larger map only prolongs the inevitable and exacerbates the whole racial transformation/enchant spam issue.
|
# ? Aug 17, 2023 08:51 |
|
I've only played one large map, and it was the snow queen assassination scenario so that I only had to conquer one corner of the map. It was still a slog toward the end. Smaller maps end up feeling much larger than they initially seem, and the encounters with the AI feel much better. I wish there was a size between small and medium though.
|
# ? Aug 17, 2023 11:59 |
|
Gerblyn posted:The AI bugs in this game are gonna drive me into an early grave, not gonna lie Thanks for the info guys, we'll look into it! Don't rock blast your ai guys too quick! I did a test game watching the Hard ai win a score victory and honestly most of what I saw looked really good. You've got an ai here that seems to be playing pretty legitimately and engages with most of the game's mechanics very earnestly. That's really impressive, most 4x games just cheat their eyeballs out. The biggest issue I saw was just that the Hard ai began expanding too late (except for 1, oddly enough, which was also the one that ran away with the game and won it). I've seen one or 2 bits that might be causing the delays but I want to run another test game or 2 with different personalities and map settings for a clearer picture. I'll post any findings in the thread here. Note: the AI did actually attempt a proper expansion victory, it just got stopped (and again it was a little late on getting started with it). Unrelated edit: Feudal Champions can actually start with a Horse and Lance now, thanks so much for that. Autsj fucked around with this message at 12:37 on Aug 17, 2023 |
# ? Aug 17, 2023 12:24 |
|
Autsj posted:Don't rock blast your ai guys too quick! That's good to hear! After all that I'm glad that at least it's working to a degree We have ID'd an issue with the AI founding cities as described though, so that's now being looked at.
|
# ? Aug 17, 2023 13:40 |
|
Definitely agree that the game plays best on small maps, I like the low distance option with about 4 players.
|
# ? Aug 17, 2023 14:47 |
|
Gerblyn posted:That's good to hear! After all that I'm glad that at least it's working to a degree Not just to a degree, from 2 test games now I would say the ai is mostly looking very functional and solid. Definitely in regards to issues that were raised in the past and were noted to have been looked at/addressed. A quick list of things I noted that were working well: Tested on a continents and coastal map, 6 and 7 players resp. , Hard difficulty, Normal World Threat. -The ai was properly clearing nodes, even nodes it skipped due to army power it would later return to to clear. The ai would even clear nodes during war times if convenient (something it often neglected before). -The ai was properly prioritizing resource improvements, rarely if ever building a local resource improvement where a global resource improvement could be build. (and mostly preferring research posts over conduits on magic nodes outside adjacency situations, which I would say is the generally preferred focus for development). -The ai was properly placing special resource improvements, prioritizing low value tiles and fairly consistently using adjacency bonuses. (this one really pleasantly surprised me) -The ai declared wars regularly starting turn 30ish, massed armies, sieged cities, attempted and several times succeeded in eliminating other ais. -The ai engaged in diplomacy and made declarations, alliances and the lot, things would shift sometimes even outside of the victory emergency diplomacy. -The ai would attempt both Magic and Expansion victory types, and seemed to understand defending or retaking these objectives if threatened. Though magic seems like a tall order for the Hard ai, and the Expansion type might have a bug, note below. -On slightly more packed coastal map, the ai seemed to have no difficulty engaging with any of the other ais. -AI interaction with research, echantments, transformations, and affinity perks all seemed to be working fine, though this wasn't a particular a focus of my attention for these tests. Overal that's pretty good. Really the main issue holding this ai back really is the significant delay it has in expanding, and since AoW is a pretty tight game where early pays off a lot, this is a serious handicap. On that note a few bits that weren't working so well: -The Warlord ai personality might be at a bigger handicap than most. From observations it seems like it does not want to settle any cities itself at all, choosing to build an army, declare war, raze cities and then rebuilding the ruins. This leads to some issues: first it takes too long to declare war, leaving it lagging behind; then it still needs to successfully siege, raze, and rebuild, meaning that it will start getting small and underdeveloped cities not until the halfway point of the game. Assuming it is even successful in its plan which is a tall order since it is already starting from a deficit here. I think it might need some adjustments in its willingness to fill up its starting city cap when not (yet) at war. -The ai has a few oddities in placing outposts/cities. Its early cities it will often spread a little too far out, really stretching forward in a line. While later cities it might just plop somewhere where there is almost no space, sometimes 1 tile away from another city/outpost with only 2 or so tiles to actually grow into. -On the Continents map, the ai seemed to enjoy random sailing trips a little too much; probably a pathfinding issue and not easily fixed, but of note. -Though the ai is engaging with Free Cities, some might not be doing so as well as they could. I noticed a Diplomat ai doing alright, but a Spy ai (supposedly also a diplo/vassal type) seemed to only be engaging with 1 Free City at a time, despite having more right near it, and despite starting with an extra Whispering Stone. (Preliminary observation that might need more testing, only noticed this late, might need to pay closer attention from the start of the game here.) -Bigger infestations were a real struggle for the Hard ai, even with just Normal World Threat, both test games had at least one ai severally hurt by attacks and raids from an infestation. -The ai does not seem to realize that a Demon Gate, Chrono Gate, and Teleporter all serve the same primary function, and would happily build 2 of a kind in a city. (haven't seen 3, would expect that it could) -The ai seems to have a logistical error or bug in terms of defending the Beacons of Unity: several times I noticed the Beacons beginning to be razed, the ai massing a force to defeat the razing forces and doing so successfully but then _not actually stopping the razing_. They would move their forces off while the Beacon continued to be razed, this stopped every expansion victory attempt. -The ai would also not rebuild Beacons, it would simply build a new Beacon somewhere else, leaving a trail of ruined Beacons in its territory. Finally, somewhat adjacent to ai behaviour if not exactly the same: -The new Free City tile expansion isn't properly evaluating what resource improvements to build on its tiles, causing it to spam farms, quarries, and foresters, in that order of preference. So a Free City might just end up with 2 rings of endless farms (and a pity quarry). This causes a couple of issues both in terms of how fast they grow, how valuable they are as vassals, and if an ai intergrates them leaves the ai with a lot of badly developed tiles (as the ai doesn't seem to redo improvements outside of building SPIs). This behaviour seems to change to the normal (improved) tile evaluation behaviour once the Free City becomes a vassal, leaving me to believe it is directly related to the adjusted tile expansion behaviour it has pre-vassalage. I realize that these sorts of feedback lists were the sort of thing you probably really would've preferred to see during the Open Beta, I'm sorry, I didn't really have time/energy for it at the time. I hope this is still helpful for the future though. Autsj fucked around with this message at 17:22 on Aug 17, 2023 |
# ? Aug 17, 2023 17:04 |
|
All feedback is good feedback, thanks a lot!
|
# ? Aug 17, 2023 17:27 |
I started a new game after the release was out, set to max difficulty - the demon rat warlord guy just besieged and took my second city, so that's good. No idea what the rest of the map looks like.
|
|
# ? Aug 17, 2023 18:21 |
|
I have other things demanding my attention but how's the Watcher patch y'all?
|
# ? Aug 19, 2023 10:38 |
|
Thyrork posted:I have other things demanding my attention but how's the Watcher patch y'all? The underground is real good now, I much prefer the new tome setup, there's a rakeload of good tweaks to weak skills (Channel Power is real good now). Personal lowlights: Taking an extra hero near the start knocks your per-round imperium gain to just below my personal fun threshold.
|
# ? Aug 19, 2023 10:54 |
|
Ranged dominance getting defanged hard is good, builds are more structured now due to affinity requirements and I like that.
|
# ? Aug 19, 2023 11:47 |
|
KOGAHAZAN!! posted:I find myself being in the awkward position of liking the idea of huge sprawling multi-dozen city empires on like an aesthetic level and viscerally loathing city build queue management gameplay in practice. Like, in every 4X. I hate it more with every 4X I play. There's just something fundamentally painful about cycling around the map clicking in to all these submenus. This is pretty much what Old World does. Give it a try.
|
# ? Aug 19, 2023 22:53 |
|
Do people play on high world threat? I find it completely insane and not really fun. The infestations get defense armies and start sending out attack parties so early. And the neutral armies are very strong, especially on magic materials. Later in a game they are often harder to take than gold wonders.
|
# ? Aug 20, 2023 09:20 |
|
High world threat is kinda bad because even the AI can't beat them. Normal is probably best for balance, though I also run a mod that makes it into a good medium between vanilla normal and high so the stacks remain decent challenges as opposed to like 3 tier Vs and whatever.
|
# ? Aug 20, 2023 09:49 |
|
The ai is supposed to get bonuses vs independents now based on word threat, I haven't really tested this since I don't like high world threat but they theoretically should do a little better now there. I think high world threat primarily exists for 2 reasons (outside the typical modularity that these games kinda like to do: just let the player choose and set what they want) First is for people who want to play a more hostile enviroment, to create more of a player vs the world kind of game. And the second is for the old school pbm crowd who auto vs other players (or at least used to before the modern cloud system) and thus their tactical challenge comes/came from the player vs indies, and a big part of that format was pushing your tempo by developing strategies for high strenght/high risk nodes. Higher world threat to increase the challenge there makes a sense.
|
# ? Aug 20, 2023 14:59 |
|
Higher world threat also affects stuff like, say, the availability of experience on the map for powerleveling. If you can manage to establish some tempo in your clearing that + the rewards from cannibals corpse looting etc are much higher, and some players may prefer the dynamics that creates.
|
# ? Aug 21, 2023 10:49 |
|
Explorminate did an interview with Lennart "Sauce" and our boy Gerblyn. Goes a bunch into the development of the game and the ideas they went over and with. It was neat, I liked it. https://podcasters.spotify.com/pod/show/explorminate/episodes/Interview-With-Triumph-Studios-on-Age-of-Wonders-4-e28l72g
|
# ? Aug 29, 2023 10:01 |
|
New open beta: https://forum.paradoxplaza.com/forum/threads/watcher-update-1-2-open-beta.1598456/
|
# ? Sep 8, 2023 14:01 |
|
With the various balance changes, what's a good / overpowered build to take into Grexolis?
|
# ? Sep 17, 2023 22:04 |
|
You’re going to want to take the bonus that gives you free units for killing stuff, which means a lot of your army is going to cost mana upkeep, so plan to generate a lot of mana. Your allies are actually quite good at stalemating the enemies with giant pyrrhic battles, so you will likely only have one enemy that is actually able to send stacks at you, so focus on killing that guy and then you can just send assassination death stacks at the enemy capitals. You likely won’t be able to hold the territory easily, so just get three or so core cities and just burn everything else down, Chaos affinity is the natural play here.
|
# ? Sep 17, 2023 22:26 |
|
Empires & Ashes DLC is coming with the Golem update november 7th. Looking forward to it. https://www.youtube.com/watch?v=MO-TmGySy9E
|
# ? Oct 11, 2023 17:30 |
|
Nice, I've really been looking to get back into this game once I finish Lies of P. And MK1. And Alan Wake 2. And my next playthrough of BG3.
|
# ? Oct 11, 2023 17:32 |
|
Bird people! Machines! Item forge! War coordination system(?)!
|
# ? Oct 11, 2023 17:48 |
|
Gerblyn, will it be possible to create custom items that can then show up as loot in subsequent games? As much as I enjoy making items, but I also really enjoy finding them from quests, clearing sites, etc. The current pool of items is pretty small, and while I'm sure it's being expanded, it would be great if I could also populate it with my own items.
|
# ? Oct 11, 2023 17:55 |
|
Excited for the QOL/feature updates and guns
|
# ? Oct 11, 2023 17:59 |
|
Hryme posted:Empires & Ashes DLC is coming with the Golem update november 7th. Looking forward to it. I have faith in dog dev.
|
# ? Oct 11, 2023 18:08 |
|
Guns in a fantasy setting? Come on Devs give me some Hippo racial features now.
|
# ? Oct 11, 2023 18:33 |
|
Fitzy Fitz posted:Gerblyn, will it be possible to create custom items that can then show up as loot in subsequent games? As much as I enjoy making items, but I also really enjoy finding them from quests, clearing sites, etc. The current pool of items is pretty small, and while I'm sure it's being expanded, it would be great if I could also populate it with my own items. I’m afraid not. We wanted generated items, but ran out of time before we could do it. I’m hoping it will be a possibility next update tho. Adding player made items to the pool is a good idea, I can suggest it!
|
# ? Oct 11, 2023 18:47 |
|
Gerblyn posted:I’m afraid not. We wanted generated items, but ran out of time before we could do it. I’m hoping it will be a possibility next update tho. Gerblyn posted:Adding player made items to the pool is a good idea, I can suggest it! ------------ Looks like trait points are all in one bucket now, yep? I get it but *waves at stuff I've said elsewhere*
|
# ? Oct 11, 2023 19:14 |
|
Gerblyn posted:I’m afraid not. We wanted generated items, but ran out of time before we could do it. I’m hoping it will be a possibility next update tho. Thank you for considering it!
|
# ? Oct 11, 2023 19:20 |
|
PaybackJack posted:Guns in a fantasy setting? Come on Devs give me some Hippo racial features now.
|
# ? Oct 11, 2023 19:29 |
|
I've been out of this because *gestures at 2023* but I'd like to dig in again for this next dlc Did they patch the diplomacy trade annoyances with the UI?
|
# ? Oct 11, 2023 19:59 |
Fitzy Fitz posted:Gerblyn, will it be possible to create custom items that can then show up as loot in subsequent games? As much as I enjoy making items, but I also really enjoy finding them from quests, clearing sites, etc. The current pool of items is pretty small, and while I'm sure it's being expanded, it would be great if I could also populate it with my own items. There's a pretty good player made mod that adds more items. "Wondrous Item Pack" I feel like it's decently balanced and quality and adds some interesting items.
|
|
# ? Oct 12, 2023 03:00 |
|
guns...guns...GUNS...GUNS...
|
# ? Oct 12, 2023 03:17 |
|
Lowen posted:There's a pretty good player made mod that adds more items. "Wondrous Item Pack" I feel like it's decently balanced and quality and adds some interesting items. Oh that's a good idea. I haven't really looked at the mods since pretty soon after launch.
|
# ? Oct 12, 2023 03:22 |
|
|
# ? Jun 10, 2024 18:03 |
I haven’t really ever touched mods either so I started a new realm with a bunch thrown in. The Hero Classes mod is very fun but likely hilariously busted balance wise. I think I’m going to try to limit it to maybe just my ruler because it sure seems like it’ll get out of control quickly. But that’s fun! The game is such a great receptacle for more content, I hope we get a second round of expansions (and mods keep being awesome). The unique veterancy mod is also v cool and honestly something I wish was implemented baseline. I know it’s a ton of work but it makes it so much easier to care for even my lamest dudes when they have cool veteran upgrades.
|
|
# ? Oct 12, 2023 03:34 |