real_tabasco_sauce

new poll

You aren't butting in, thanks for your input! However, I am really confused @ChronA. I can't tell what you mean.

@chrstgtr You are slamming me and my change to the mod without simply offering a clear way forward.

Ok then you want to go back to a26, surely. Frankly, I don't understand why you say this. I see perfectly content players in the community mod. You are massively overstating the effects on rushes, people are still doing rushes very effectively, its just different. One player yesterday told me that rushes are slightly more difficult. thats about 1% of the complaints. And your own complaints above are purely based on theory as you insist. Right now I am hearing that archers are too weak. I heard this from about 15 people when asking the lobby. I heard elephants are not quite effective enough from 5 people. And I heard 1 player say that spearmen were much more impactful when defeating a rush than the building arrows! Frankly I think you are very confident about your opinion solely based on a26 vanilla, and you fail to accept what is actually happening in the mod. It is not a disaster as you make it out to be, and that is why you called for "more experimentation" even though you said above "it appears to have been rejected." If it was really that bad you would actually advocate for a revert.

That is massively mischaracterizing the intent of changing buildingai. I did think the 1:30 chicken rushes were imbalanced but above all else, the purpose of the buildingAI change was just progress. To further the enjoyability of the game. I don't know what to do with this information. I don't think it is helpful at this point to use the past as a reference: Its better to observe current games and adjust from where we are at right now. So it doesn't help to tell me I am wrong and my development is wrong simply because things were not problematic in the past. Nevertheless, I think I identified the two main issues you raised with the current version and I agree with both. But ok, I give up on it for now. The next version will just address the other variables and then we can take another stab at rebalancing arrows sometime later.

anyone know how to refresh a poll? I would like to start a new one with more detailed questions.

These are indeed very different....I am repeating what others have said. I pointing to votes of other people. You not only are saying what other people MIGHT think--you are saying that people will change their mind. I honestly don't see how you can continue to insist on knowing the opinion of others. Ah, then you must be misunderstanding me. I was explaining that 0ad NAIVE players would expect and prefer a system similar to the one I developed over random arrows. You are the one saying "People don't like X, Nobody complained about Y" Ok this is closer to what I was asking for. Before you were just deflecting to how things were previously to describe a problem in a26.6. So I am gathering this from your paragraph: "Non-random building AI is too effective at stopping rushes" Is that what you mean? If so, I agree. Changes must be made to CC arrows and sentry towers so that rushers can still find kills. And lastly, let me address your last sentence here. Maybe people didn't complain about it, but my observations told me there was a problem. Players were able to dive under the CC for extended periods of time with no consequences. The CC should (according to the success of other games) be able to be a somewhat safe space if there is a ton of pressure. I would just argue that it is now too good at countering rushes when it is garrisoned now that it actually gets kills. It seems pretty clear then that the solution is to change the CC arrows and not revert to a26 arrows. With the right balance, rushing should be more dynamic since it brings in arrow micro, dodging, weak unit sniping, and quick dives under the CC. You could compare it to a26 where early rushes basically only involved buildings for uprising rushing cavalry with soldiers, other than that, they basically were just a timer saying "ok in XX seconds you need to leave the cc range." About your second paragraph, I would summarize it as: late game effects are complicated, but in general there is more turtling. Is that accurate? If that's the case, I would also agree and argue that arrow balance could fix it. However, given that there is so much uncertainty, how about I put together an update that addresses all the non-buildingAI stuff which is easier to address, and then we can reassess?

Come on...Also, see below. You talk on behalf of others, and I say what would be likely if the decision would be made with no status quo. These are very different. The concept doesn't work if you think rush balance was fine before the mod. Full stop. It's inspiration is a false premise. Thats not an issue though. Give me an issue with the way it is currently.

new, more detailed poll: also, anyone feel free to chime in on this:

I think you are making this decision for a lot of people. You are taking me out of context. I said these things before the mod was out. Now that it is out, players understand it. This one? In which only 10 people voted? It was 6-4 non-random. The only written complaints I have seen that didn't come from haters were yours. I would hardly call that a fact. If it was fine before, why don't you advocate for a full revert? After all, you suggest that If that is the case, why do you suggest to further change the system ... Let me make the rest of this reply about why going halfway as you suggest won't work as well as one might expect: Not an elegant solution: As you suggest, there would have to be two distinct buildingAIs, one with random arrows and targeting, the other with non-random arrows and targeting. Either this, or a new class for non-random buildings that gets read by buildingAI and then activates a non-random version of firearrows(). How to balance random and targeted arrows: Allowing a building to be random when it becomes non random upon targeting is bound to fail: If non-random arrows are stronger as you say, players would benefit from constantly targeting manually. You would have to ensure both the targeted arrows and the random arrows are balanced, which would (obviously) need to be handled differently. Inconsistencies: Asymmetry between units is fine and we call it differentiation. When mechanics are so massively different when players expect consistency, we have a problem. You call for differences between targeted and un-targeted arrows, as well as further differences between CCs/Forts and towers. I think that would be a confusing mess. ->Please tell me some issues with 26.6 building arrows that are not due to arrow count balance.<- What makes you say this? I saw complaints during week 1, but now I see none in the lobby. Surely this means there is no problem, right? Jokes aside, I don't declare no complaints to mean there is no issue, that is why I plan to rebalance the arrows.

These are both symptoms of the fact that buildings get way too strong only when garrisoned. Take the CC for example: 33 pierce goes up to 253 at the moment. This has always been an issue IMO, its just now that it is actually killing units it feels OP. Previously, you would just retreat after losing HP across all your units. I think you need to look at the behavior and the arrows objectively and individually, throwing away all previous balance ideas, and ask yourself which is the superior system? One that acts as a damage aura, or one that actually kills individual units? Then, ask yourself which is the issue: is it because buildings become 5 to 8 times as powerful when garrisoned, or is it because units are killed 1 by 1? Without any nostalgia or fear of change, people would prefer what I have suggested to random arrows. Lastly, what do you think should be done now? Full revert? partial revert?

yeah and then someone turned around and made towers 11 pierce XD. I don't think all buildings should have to behave identically wrt arrows. is that so? I assume u mean non-random arrows here, not the above updates I suggest. I assume you mean non-random arrows, not the changes I suggested just now. Please tell me some problems with non-random building ai that are not due to the arrow count balance. My understanding was that you found particularly the CC to be too strong at preventing rushes, and that turtling was too effective. There were problems with random arrows, its just that people tolerated them: Buildings did not effectively deter rushes/raids, and healer hero auras could invalidate building arrows.

To give some examples, currently these are the changes from un-garrisoned to garrisoned: Sentry: 11->44 is 4x stronger Defense tower: 11-> 66 is 6x stronger CC: 33 -> 253 is almost 8x stronger Fort: 44 -> 264 is 6x stronger with the changes above, towers would only be about 2x as strong when garrisoned, the CC would be about 2.5x stronger, and forts would be about 4x stronger.

Ok here is my plan to address current concerns with buildings. My understanding is that most of the problems people have are when buildings are fully garrisoned. For example, a CC gets almost 7x as powerful when fully garrisoned, so it can be too powerful early on. Also, many players are leaving towers garrisoned indefinitely in order to make the most of their arrows for turtling. These differences were not really problematic when buildings shot randomly, but when all arrows are directed to 1 unit at a time, they become problematic. I think this is an example where we should look to aoe2 as an example. CCs, Towers, and castles get stronger when fully garrisoned, but not by an order of magnitude like they do in 0ad. So the plan is generally this: decrease the fully garrisoned damage of most buildings, while keeping the ungarrisoned damage the same (or maybe boosting it for forts?). The defensive buildings is that they provide firepower without the need to garrison, and that garrisoning can be used to add arrows if you are in a rough spot. (for that matter, I think the un-garrisoned arrows should be even better, but I am sure most people would want a more cautious approach like this). Towers: fire rate: 2->4 default (un-garrisoned) arrows: 1->2 Now for CCs, people say they are too strong when fully garrisoned, so I would do a little more: fire rate: 2->4 default (un-garrisoned) arrows: 3 -> 6 max arrows: 23 (basically the same as a fort) -> 16 For forts, I think they should remain pretty tough to crack, so I would do it a little differently: fire rate: 2->3 default (un-garrisoned) arrows: 4 -> 6 (or maybe 8 to boost un-garrisoned arrows) I think this is all I would do for now, but in the future we could consider stuff like letting tower upgrades apply to forts, or phase dependent building damage.

It may be possible that the community mod might contribute, but I can't think of a mechanism. I think this is more likely a case of more reports arising from this discussion and trying to get to the bottom of it. For example, I only uploaded the replay and thought to ask about memory specs because I knew this discussion existed.

any updates on the mac logs/replays/mods location for a27?

@Feldfeld do you have a github where you are working on feldmap? id be happy to try and get the other badosu maps working with your balancers for some future release.

Windows 11 host (not me) with 4x32 ddr4 ram crashed today at ~30 minutes metadata.json commands.txt Everyone was in the OOS error and it all happened at the same time. Also, the game was not playable afterwards with everyone losing connection. The host said the game didn't crash.

I would be fine with giving skirms some armor as a way of differentiating them from other ranged units. Ranged units don't have the same diversity that melee units have. However, I strongly disagree with giving units dual attacks en masse. All this will do is make the gameplay confusing. The core units should remain very simple, so I would only be ok with 1 or 2 unique units with dual attack abilities. Ie only legionaries can do the pilum throw, but not pre-reform swordsmen.

yeah, my concern was that some people would have the old version, but it looks like the last release seems to have been for a25 anyway. I think the zip file name was the same, but maybe the mod.json wasn't correct. Idk. I'd like to merge it with feldmap with @Feldfeld at some point anyway.

Ok, the updated mod is up on the mod downloader now, and I tested a few generations. The only issue is that for some reason it is called balanced-maps-2 after downloading, which was a previous name of the mod found here: https://github.com/0ad-matters/balanced_maps_2. I probably forgot to change something. Nevertheless, it does work as far as I can tell, and it includes old favorites like slopes and cross.

Do you mean other players? Are there settings which minimize memory usage? maybe turning off model appearance randomization? Maybe I should recommend windows users try those settings to help prevent oos?

@borg- @chrstgtr @wowgetoffyourcellphone and I have been doing this somewhat. https://trac.wildfiregames.com/wiki/Alpha27

@chrstgtr maybe see if one or two of the other players can post their replays here.

I must be in a huge minority thinking 1v1s should be played at 200 pop.