All Activity

The Crush damage type is a mess because it incorporates three different kinds of attacks: Small blunt objects, e.g. stones Blunt objects with a lot of mass, e.g. maces Anti-Building attacks, e.g. rams This will always lead to weird gameplay implications. E.g. currently Slingers are more effective against "lightly armored" targets as they should be, but they are also somewhat of a siege unit. Macemen behave more like arsonists with high anti-building damage but low damage against human units, especially heavily armored ones. How to fix this? Idea 1: Two "new" damage types: Thump and Crush_new Slingers deal (mostly) Thump damage, and resistance of units is similar to current Crush Resistance. Low for weakly armored units, significantly higher for heavily armored units. Like 1-5-10 for infantry ranged, infantry melee / champion ranged, champion melee. This tries to simulate that the Sling loses comparably a lot of its lethality against armored targets. Crush_new is the damage type Macemen and partially Axemen deal. Crush_new resistance would be fairly low, like 1-2-3. This tries to simulate that maces and to a lesser degree axes have so much kinetic energy that armor is a lot less effective against it compared to how well it protects against e.g. swords or arrows. With two new damage types it would also be easy to incorporate them into Forge techs; Pierce Armor upgrades increase Thump Res, and Hack Armor upgrades increase Crush_new Res. Siege units get their own special damage type, or a bonus against buildings. E.g. a Ram deals Pierce and Crush_new damage and has a huge bonus against buildings. One issue with this idea is that very few units would use the Thump damage type. Mostly just Slingers, and maybe Onagers. Idea 2: Blunt and Piercing Both Slingers and Macemen deal Blunt damage. Blunt Rest is similar to current Crush Resistance. Macemen also deal a little bit of Piercing to simulate them being good against heavily armored targets. Units would have 0 Piercing Res, and buildings a very high one. This would also enable the option to make other units relatively good even against heavily armored targets, e.g. Bolt Shooters. Against, Siege units would either get their own damage type or a damage bonus against buildings. One issue this idea has is it's hard to decide with Forge Armor upgrade should affect Blunt. Idea 3: Crush_new Crush_new Res is something like 1-2-3. Only weapons particularly effective against heavy armor have the Crush_new damage type, e.g. maces. Slingers deal only Pierce damage, but have a bonus damage against ranged units to simulate the Sling being effective against unarmored targets. Siege units deal either Hack, Pierce or Crush_new damage, and have a damage bonus against buildings. This doesn't add any new damage types, but has to heavily work with bonus damage. The issue is if Forge Armor upgrades should affect Crush_new Res, and if so which one.

Also @victorcrimea is working on dedicated servers.

I've seen several posts that suggest that the current Petra is too difficult, after my changes to its behaviour. My question is, should I make new PRs to nerf it Also what about creating more granular difficulty levels? For example we can lower the very easy to even lower and add some more extreme levels. Currently the only cheat of Petra is in terms of eco. We can give the extreme levels other cheats such as bonus to atk, unit hp, building hp, unit movement speed, training time discount, projectile accuracy, armour, price discount...

This is the best scale I have seen so far and the most useful

I was wondering literally today why they overlap in the first place! this will be nice to have @Atrik I believe you used the wrong link for the pull request

this scale is okayish as well but it seems a bit too generous. in your scale you assume them a bit better than I do,which is close enough to be still almost same as mine

but if i were a host doing TG i would judge rating like as follows: <1000 as useful as maginot line in WW2 (complete new,as useful as a tiny roadbump) <1200 knows how to get full pop at some point <1300 can get 200 pop at min 17 with a few forge upgrades <1400 mature enough to play the "elite" TG if hes lucky. can get 200 pop in 15 mins and all forge upgrades in 17, but dumb at choosing what fights to take,how to fight in a battle,that (actual fight) sort of thing <1500 has enough IQ to understand what is rushing and how to play with/against it pretty well. has eco good enough to reach 200 pop before min 12 and all forge upgrades in min 15, can be officially considered as a "strong player" in most TG games these days <1600 old enough player to know what is OP,what is gimmicky,what works great in a given specific situation, what is bad asf, and etc. <1700 knows the very rarely known things (e.g what type of army to make against a certain civ, what kind of resources your particular civ takes up a lot, how to heal heroes at giga fast speeds without any temple) and knows how to play atleast one civ perfectly well, in the best way that civ can be played tactically 2000+ right now = vinme,vali,borg,wolf

Enhancement to freehand formations: units now actually order themselves by distance, and don't cross each other's paths. Nice when microing ranged cavs. Before : After : Committed for integration in R29 too, as part of https://gitea.wildfiregames.com/0ad/0ad/pulls/8525 (use common sorting functions).

as a bourgeoisie of this scale, I must say that I divide people in classes by using petra: >1100 = can't defeat very hard petra >1200 = can defeat very hard but will be tough 1300 = can defeat multiple very hard petra bots 1400= can defeat 4 very hard petra bots but barely 1600= can defeat 4 very hard petra bots with time but easily enough 1700= can defeat 5 very hard petras alone. this scale is calibrated for a27 dumb petra,not the gigachad we have right now thanks to troops scanning for units even while attacking buildings (lucky if you did it in a27, 1400+ is nigh impossible in a28)

I think the biggest problem with the rating system is its too easy to "farm" points and then refuse to play against bettter players. So in a lot of ways most of the ratings you will see don't mean anything. For 1x1 it should be more like SC and you get assigned a player based on your rating, so new players can only play new players etc (For rated games) that way you cant get to 1500-1600 by only playing 1200 players I try and find rated games of ~100 of my current rating, but its harder cause a lot of 1400/1500s don't want to play rated, mostly cause they probably aren't 1400-1500!

I meant unique champions, like the Mauryan Elephants, the Persian Inmortals, and the Mauryan Amazon women (I forget their real name). These units are a lot more unique in play style than the basic units, and I feel like because the stats are so generic across the civs, the game really just becomes who can boom faster, whereas fantasy RTSs often end up becoming more strategy focused with varying tactics against various civs.

Acknowledging that following call is not relevant to 1v1 and no-team matches: Don't forget the impact of Alliances with strong Team Bonuses.

0 A.D. does cosmetic uniqueness well, by making every building and unit of each civilization look different. Champion units are nothing special, they are only stronger versions of basic unit templates. Also, each champion template is the same across civilizations, with some small differences unique to some civs. Again, the game treats uniqueness through special civilization-specific bonuses and upgrades. Every unit template is the same and has identical base stats. If an Elite spearman of your civ looks like he's stronger than an Elite spearman of another civ, he just looks stronger. They are the same unit template, so they have the same stats.

I’ll probably test my ideas out in a small mod before merging them with something else. I feel that they should be further exaggerated, not to the point of cartoonishness, but so that the differences are noticeable to someone who has maybe not played every race extensively. For example, using the cavalry example you mentioned above, the Carthaginian cavalry is only barely faster, and so it has to be directly compared against a standard cavalry unit to notice the difference. Cosmetic uniqueness is important in many ways, it makes each civ feel like a unique race, it allows players of varying play styles to pick the race that most fits them, and it allows players of lesser skill to beat more advanced players simply through choosing the right race for the right situation (though again, over exaggeration must be avoided here). As part of this uniqueness, the ‘baseline roster’ should be changed for each race, so that the game is interesting long before champions are unlocked in city phase (champions are atm the one real unique units in the game). Again, not over exaggerated, but for example the changes I have above should be generally applied to the majority of units for each civ. You all can expect my mod sometime soon, most likely after I get the next release of Hyrule Conquest: Revival out.

You are good enough for TGs and I know how to balance you. That's all that matters.

There is no bubble. It is what it is. Get used to LLMs

Using a VPN could also solve it for you. The two possible solutions would be - dedicated servers: https://gitea.wildfiregames.com/0ad/0ad/issues/3556 - support for IPv6: https://gitea.wildfiregames.com/0ad/0ad/issues/4301 both of which, yeah, aren't trivial to implement.

after the ai bubble bursts,hopefully the prices come down enough for crowdfunded people like WFG to get smth like it

I don’t mean to suggest that your observations aren’t a valid concern, I’ve also experienced connectivity issues after switching ISPs, and I’ve felt some frustration. I’m not the right person to give a definitive answer. However, my limited understanding of the topic leads me to think that moving from a P2P system to a server-based one is not as trivial as it might seem. I hope I'm wrong about this

For ELO to work near to correctly, you need matchmaking system. Otherwise, as guerringuerrin said, this is small community - you will find out your relative skill level by playing with others. Btw, i'd not use ELO numbers, rather Tier 1, Tier 2, Tier 3, etc. Or Level 1,2,3....

P.S. i wanted to try new release and play Rated to update my MMR a bit, but this really bothered me I kinda lost will to find hack to play. Also lag/performance... Really essential things imo.

what's cost? How OpenRA has tens of servers (they don't do peer-to-peer but every game goes via server afaik)? I mean we don't have thousands of players, these days processing power shouldn't be expensive, is it? And isn't this main priority? If guys already invested thousands of hours in developing this great game, shouldn't it be finished to work properly? God knows how many players are annoyed by this... I mean, IK it's open source but... there must be a way

My 2 cents: Ratings is not that accurate as it seems until you get 2000 with 1000+ games

Some Rome vs Rome civil wars It was tough for a while, if the AI managed to get its army together before attacking, it would've won on that first push. Grouping logic needs to be re-implemented. RomeVsRome_M.zip

@BeTe The main difference is that StarCraft relies on Battle.net servers to help players connect and, if needed, even relay the connection, so you usually don’t have to configure anything yourself. In contrast, 0ad uses direct peer to peer connections between players, which means it depends much more on each person’s network setup, router, and ISP. That’s why StarCraft tends to “just work,” while 0ad can require extra configuration in some cases. Now, one might think “why not just build something similar and add a server to handle this” It sounds simple, but in practice it’s quite costly, both in terms of development time to implement it properly and the ongoing resources needed to keep it running reliably over time.