alre
-
Posts
1.281 -
Joined
-
Last visited
-
Days Won
15
Posts posted by alre
-
-
can you point us the review process you are talking about, the work done, or at least the language?
-
@hyperion, @LetswaveaBook, wow, your opinons look pretty extreme to me. I'd say it's cool that not all games must finish in age 3, but it's also cool that one can go to phase 3 just by booming, before the deciding part of the game even starts. that has to be allowed, and many players would go for it anyway.
-
31 minutes ago, real_tabasco_sauce said:
alre maybe I should elaborate: with all players using auto start, automated eco, and auto sniping like you suggested elsewhere, you begin to homogenize gameplay. Its inevitable.
not quite. you can take radical economic choices with ProGUI activated, and feel more confident about them because of the mod; you can and you would very often turn off the auto-production, to then turn it on again. not every game is the same in any ways more than with vanilla.
it's a bit like when autociv's autoqueue was out (and now it's in the game), except ProGUI is more helpful.
-
2
-
-
2 hours ago, real_tabasco_sauce said:
I can tell you for sure that with automated eco, and even more so for automated sniping, the game would simply be a lot less fun.
for you.
-
1
-
-
1 hour ago, Yekaterina said:
This idea will likely work on AI games only.
not. AI does not work on saved games unfortunately.
it would be very cool if it was possible to play saved games multiplayer.
-
2 hours ago, seeh said:
i talking about the default value. you have default values in all fields. i suggest use also here a default. you now understand?
Still, that would be a change inconsequential in everyone's practice.
-
Just now, wowgetoffyourcellphone said:
What's the use of having interesting names if the maps don't look like what their name suggests?
if the map depicts a H-shaped terrain with square seas, I'd argue that it shouldn't be called Corinthian Isthmus whatever the biome. Isthmus is fine, it's not boring (then why would you think a better name should make a map look better, if it's only loosely related to it?).
-
39 minutes ago, real_tabasco_sauce said:
Releases should still be well balanced. We don't want to divide the community as it is with the current community mod.
balancing stuff can be done in the community mod, why not, but it is more helpful and impactful to try more experimental things in the mod. Basically we can try stuff that will really shake up the meta (and possibly balance) without risking a release being poorly balanced.
you can use different ad-hoc mods for those tests, like valirhant did with territory mod.
and then merge the feature into community mod.
-
39 minutes ago, real_tabasco_sauce said:
I support this decision.
As for balancing, there is not a whole lot changed, so it shouldn't be too hard. Some things to consider:
- Check how the ele nerf turned out.
- consider removing elephant friendly fire.
- catapults overnerfed? if so, consider adding a small splash radius of 1 or 2 meters.
depending on timing, maybe it would be worthwhile to consider @borg-'s updated sparta patch for a27?
shouldn't balancing rather go on with the community mod? A lot of development can be made trough that project.
-
1 hour ago, hyperion said:
This active pause sounds a lot like simulation speed 0, which you can set via console or add an option for via mod.
you can also set it in the gui.
-
yes but you need a mod (could be autociv, I don't remember)
-
1
-
-
Personally I would make capturing unaffected by garrison. Garrisoning is currently op and quite micro-heavy.
Then again, A27 will be yet another thing: capturing will be buffed and less encouraged at the same time, but what can you say, devs will be devs.
-
1 hour ago, real_tabasco_sauce said:
I just think that level of control of your units should not be automated. I agree that whoever has the faster clicks should not always win a fight and that sniping is not a desirable meta, as @BreakfastBurrito_007 pointed out, but can we let the-fine tuned control of units remain a learnable skill?
there is fun in trying to find the lowest HP cav that rushed your early eco and being able to kill it as they retreat.
Also there is fun in sniping if you know it will be effective because you have learned when it is effective.
Adding a stance to do automatic sniping will add a level of automation that doesn't fit with the behavior of other stances, or any other configurable behavior in the game.
Also, when the sniping meta is fixed, automatic sniping will probably just mislead new players into thinking ranged targets should be prioritized, when in reality it depends on many more factors.
perhaps you would take interest in attack ground? I would find this to be a much more generally applicable attack mode. Its got a few nuances in terms of implementation, like how to smoothly choose the radius while selecting the area to attack, and whether or not the attack should be one volley or endless attacks.
there still will be fun in finding new tactics and micro techniques, just not this one.
-
1
-
1
-
-
11. in maps with long and complex unpassable obstacles, the AI starts acting weirdly, only considering euclidean distance for espansions and attacks.
45 minutes ago, hyperion said:10. Gathering resources under enemy tower.
I think this is the kind of things you may want the easy AI to do anyway. if it acts dumb it makes you feel smarter.
2 hours ago, NitroVicky said:The AI, PetraBot, certainly can be improved. It often uses bad strategies and is weak; I consistently beat 7 AIs all teamed up against me. So I am trying to improve the AI behaviour by tweaking its code and I would appreciate any inputs / suggestions.
Firstly, I will identify a list of imperfections with the current AI:
- Suicidal early attacks - sends some units to attack you but then suicide all of them at your civic centre. It doesn't know how to retreat. The long march also delays their development by a huge amount.
- Wasting resources on non-essential buildings and techs: Greek theatres, towers, wonders (even though it's far from reaching the limit population and needs to spend resources elsewhere)
- Wasting time on capturing / damaging non-essential buildings, especially suicidal attempts at fully garrisoned fortresses.
- Poorly optimised army: too many siege units, too many mercenary units but not enough real fighting force.
- Does not interact with the player - they don't seem to care about what you are up to; they just attack walls and towers as they wish.
- Unwise expansions - too many civic centres placed at poorly chosen areas when they really need to defend their main base.
- Suicidal support units: they send women and traders into battle or under enemy towers
- Not playing: when checking the replays, I often see the AIs doing absolutely nothing for some time.
If you spot any additional flaws, you are more than welcome to append to the list.
there have been some work on better AI in the past, which was never merged into the main game, @Yekaterina knows more. Catilina was pretty good I remember.
-
1
-
22 hours ago, Vantha said:
What do you mean no?
there are AIs ment to help developers coding, but they are only far relatives to chatgpt. besides, coding is not a "translation" of human language into machine language, code has a rigour that natural languages lack, and if you can't code some feature, chances are you can't express what you want from that feature in natural language either.
the only thing LLMs can do, and IMO will always be able to do, is to give very generic indications, or creating very generic pieces of content, that you can either spam as they are or use as a stub for later rework. this applies to code as well as to other text.
models that generate images are somehow different because pictures are fuzzier and AI works better there, the hand of a painter is not as precise as the pen of a writer or the hands of a programmer.
-
5 hours ago, Stan` said:
The reason is minimum resolution, but yeah for some users it's wasted
I think he meant the bottom-leftspace, which is left for chat, that doesn't apply to single-player.
-
1
-
-
1 hour ago, Vantha said:
The whole point of LLMs is to, instead of humans having to adapt to the computer language("to code"), enable computers to adapt to human language. Right now when coding you need to "translate" human language into a computer/programming language. One of the goals of LLMs is to do that "translation step" for humans(not just in programming, but also in other areas) so everybody can directly interact with computers in human language.
And even though GPT 4 is very very advanced, it can't write large working pieces of code from just a simple prompt in "human language". Even GPT 4 is not good enough for this. Somebody with no coding expierience can't really use GPT 4 to write code right now. But in the not so distant future there will be LLMs that can do that.not really, no.
-
for any nomadic civ, territory should be either not present or somehow secondary, I'm hoping for a civ that actually plays nomadic.
10 hours ago, BreakfastBurrito_007 said:AoE4 had Mongols as a nomadic civ, but basically all it meant was that their buildings could pack up into carts, and it was quite gimmicky. I would prefer to maintain the functionality of the cc (territory root, training) while it moves, and have it move very slowly, also other buildings should be mobile as well but faster and with a pack up pack down feature. moving cc territory would be very challenging without making some features to help make it feasible, such as a dedicated territory root building, permanently mobile storehouses and farmsteads, and/or buildings not decaying to gaia once they are out of cc territory. Nomadic society doesn't simply stop when they are on the move, so the cc should remain functional as it moves. Having a slow but still functional cc also means players need some planning and strategy while moving around the map, instead of zooming around and using the cc to snipe enemy units like in AoE4.
I agree that packing/unpacking shouldn't be abused, and pure mobility could be used instead for a lot of units (that would be buildings in other civs, houses maybe).
-
give me a nomadic civ and I'm happy with that, I ask nothing more.
-
2
-
-
15 hours ago, real_tabasco_sauce said:
metadata.json 20 kB · 1 download commands.txt 152 kB · 1 download
Alright the first posted multiplayer game of a27 rc.
Not very balanced, and I had a rough eco.
The elephants I trained as @Philip the Swaggerless attacked did unreasonably well, so I think a nerf is warranted. I am thinking 20 hack 20 crush for the splash, and then add back the 10 crush to the direct attack.
maybe they wouldn't have been so effective if he had known how effective they could be and he had taken counter action? just an idea - micro play can be pretty effective against splash damage in general.
-
start by making a simple attempt. then show what you have done and collect feedback.
46 minutes ago, ShadowOfHassen said:Also, where will we put the writing in a mod first? Or is there a better way?
yeah, I don't see a better alternative. However, if you want to give visibility to your work, simply posting what you have written here seems best to me.
-
1
-
-
2 hours ago, Stan` said:
Problem isn't so much them gaining too much points, is you losing a lot to a 1100 player when you are 1800 (while in reality they are 1800+)`
Any automatic matchmaking should avoid matching players with strong differences in rating.
Also, new players should be matched against new players when possible.
By the way, there is a flaw, I believe, in the way the current rating algorithm computes the updates to the ratings of players losing or winning against new players. The fact that that player is new is not taken into consideration (compare this to how the new player rating receives stronger updates instead). The opponent to the new player should receive finer updates, because the new player true ability is not known.
-
1
-
-
1 hour ago, Stan` said:
I did not mean it as a dismissal. I meant that somehow players are playing single player and seem to be gaining something out of it.
@Langbart and @SciGuy42 did a great job porting a campaign to A27, hopefully will be available through mod.io
Yes. But from what I read you seemed to mean that the player base would increase, rather than not decrease. Sorry if it wasn't what you meant.
it's pretty much the same thing, because there's a lot of people trying out the game, but not as many that stick. I'm sure a single player campaign would also show in the players number.
-
3 hours ago, Stan` said:
How so? It will not create new players.
Do you know "player retention"?
3 hours ago, Stan` said:We know that there are at least 10 times more player that play solo matches than multiplayer ones (1500 per day). And those are the ones that enabled feedback.
So we're obviously doing something right, even though we're not sure what it is.
Considering how huge is the number of daily downloads, one could say you are doing right by keeping the game installer online and giving it visibility.
proGUI, the day after
in Game Modification
Posted
well progui is a bit like that (but not too intelligent) because it works on a logic that makes it choose which unit to train based on army composition specifications. I would actually like mentula's solution better because I would enjoy retaining control on which barracks trains which unit.
at that point a gui mod that makes you keep track on production (exactly like boongui always did) would be a nice option for those who want to improve their game without any automation.