real_tabasco_sauce

Well the way I did it, the closest unit is selected as the target and it will be targeted until destruction. Whatever is closest is attacked until death, so If the target is a building, it has to be destroyed. Obviously, this solution is really just a proof of concept so I could make the video above. Ideally, the script could check for any soldiers within range before resorting to attacking a building. Also ideally, one attack's worth of arrows (default+garrisoned arrows per 2 seconds) would be sent to the closest target in targetUnits. After all the arrows are sent, the closest unit would be searched for again, rather than shooting the now more distant unit to death. The second task might not be too hard, but for the first, I have no idea how to check the class of an entity in targetUnits. And then there are the freezing issues. I can open a patch later today if people think we should go ahead with this idea.

My little hack seems to work for buildings and siege towers until the crash, I haven't tried it with ships. yeah it works with ships too. I had to remove the weighted list for targets for it to work, so currently buildings are equally considered alongside units (since it's really just by proximity). Would it be fine for towers to shoot buildings if they are the closest unit? Or would it be better to first find a soldier within range and only shoot buildings if there are no soldiers? I am very inexperienced with javascript (this was my first time writing js), so should I open a WIP patch or would someone with more experience rather take up the idea?

I remember someone told me it was related in some way to the custom rating mod, but I can't confirm.

Here is a video of what I made. It works as intended until the spearman crosses the minimum range, then 0ad stops responding.

My thoughts were more to the effect of: structures with more default/garrisoned arrows (forts, ccs) would have more spread, while towers would have less spread. This way towers are more precise at longer ranges, with less risk for over kill. Forts and CCs on the other hand would get a slight spread increase compared to towers, with forts probably more than CCs. Then going further, I think it would be better to allow player control of the building arrows than provide stances to adjust building behavior, but that's just my opinion.

I think with the right spread values, this could be optimized to be less of a problem, but I see what u mean.

what do you mean by snapping around? I'd say ideally, ships could end up with something more like unitAI for fighting (I think we talked about this in the naval overhaul discussion).

I've long had my qualms with the random arrow damage provided by buildings, ships, and siege towers. While I guess it might appear more realistic, random arrows make defensive buildings unfavorable: defensive buildings do little to deter small groups of units because damage is spread across all units. These units may still deal the damage their full numbers allow. In addition, they can just go home and heal. On the same thought, this makes them less interesting for offensive purposes. defensive buildings are very effective when enough towers and forts overlap on an enemy army. Since all arrows find damage (no overkill), the towers accomplish area denial more than simply aiding the defender to beat an army. I think building defenses would be more interesting and relevant to gameplay if all building arrows of a given attack targeted 1 unit, which by default would be the closest unit to the building. In theory, this would also make player controlled building arrows a little more logical. Out of curiosity, I made a little hack on BuildingAI to try this out and it "worked", with arrows targeting 1 unit at a time. However, it was not consistently the closest unit, and it caused freezes. Is anyone else interested in this kind of change to building AI? That is, non-random targets? Has it been attempted before?

Hmm thats a good idea. you could calculate the minimum distance from the destination a unit should achieve based on the number of units in the selection.

I'd support a small-ish increase to the pushing values.

These look great. I am fine with either nerf option discussed for the super cc, as this is one of many han strengths. Perhaps we could also consider https://code.wildfiregames.com/D4952 here? It's a similar balance/minor content addition patch for maury.

If it's the Han super CC, this building has a bonus for batch train time. Maybe there were ministers garrisoned as well?

Well, to show the upgrades, I did "gift from the gods", built a forge, then clicked all the upgrades.

dont mind the gather rates in the image btw, I just used a cheat to get them ranked up quickly, and this also increases their gather rates a lot.

Right clicking on the image shows the units base stats. if you want to know the full details of the unit after any upgrades and hero auras, hover your mouse over the sword and shield icon after selecting a unit. This is found at the bottom of your screen. Also, if you want details on a unit upon train time, be sure to select "detailed tooltips" from the options under "session"

Quoting from the other discussion to include the above ideas here. In general, I don't think a large-scale revision is needed here. I think just some changes to unit pushing and adding/adjusting existing formations can be perfectly effective.

Well thats pretty interesting, but I was really just suggesting some refinements/additions to the existing formations. Perhaps we have diverged from the thread too much. Would it be a complete departure from the formation code to allow something like this? I know the box formation seems to sort melee units to the outside with healers and ranged units inside. What do you think about this?

I think the intent was that giving the cavalry a buff just because they are in a triangle shape seems artificial/forced. Instead, one could increase unit pushing values so they cannot stack super well as a blob. Then, you could increase the compactness of some formations so that an easy way to get a nice compact force is to organize. For example, box and wedge could be made a little tighter, and the phalanx even more so. (perhaps disable unit pushing for formations? not sure how that would turn out) Formations are used in competitive play sometimes, but its just box. There are cases where I would like to use a formation, but it's not quite the shape I am after. There are also some kind of redundant formations. In addition to the previous idea, you could diversify some of the more generic formations (one super wide single file line; maybe a tightly packed outward facing circle) Here is another idea: formation for splitting melee from ranged: enemy -Melee units- some midsized gap -Ranged units-

To be fair, you started it with ping pong, how is that relevant?? my point is that we will never be able to (nor should we) simulate real strategies used by generals. Maybe this can be featured in campaigns. My recollection is that a23 battles were fairly similar to a26, except there was pretty much no need for melee units at all. In a24, battles were longer, units slower, and buildings stronger. That being said, I don't disagree with reducing ranged units' damage, I said previously that it might be necessary after melee buffs.

I respectfully disagree with both of these. I didn't like the 45 minute average games, and there is absolutely logistics and reinforcements in a26. Good analogy. On the other side of that coin, how about we make the game take actual days to play, make the romans better than every other civ, make women actually give birth to get soldiers, individually put soldiers on horses, make the maps earth sized? Sound fun?

yes, this will be the primary issue to fix. I'm pretty much ready to submit a few merge requests when the community mod is set up for a27, one of which would address this. It should help with the "meat shield" meta people talk about.

@wowgetoffyourcellphone I think i've mentioned it before, but I plan on changing the ranged vs melee balance in the community mod. I think this way the fear of impacting the balance can be alleviated. Probably first will be the melee damage/armor change, then ranged damage reduction afterwards if needed.

I'd say the real reason is because those "tactics" don't help in 0ad because 0ad is a videogame, not a simulation of history. In game currently, you actually have many chances to control your units, as your mouse can make hundreds of clicks within the minute you say the battles are lasting. I recommend using the "option" key to order one unit at a time for sniping. If you want more tactics you can actually click and drag a group of units to adopt a shape: for example, order your cavalry archers to adopt a semicircle in front of the enemy. There are a lot of "tactics", of course they are videogame tactics, not real historical ones, but they work well and there is time to do them. I think you just are not aware of them.

you have discovered "overkill". like @alre says, there is nothing bad about it. If you do increase fire rate and decrease damage, then you have many more range queries in one second, and there will still be overkill when a larger group of units shoot one unit. I consider it a property of RTS games in general. The default behavior of your units may not be the most effective approach, depending on the situation. Thus, it is up to the player's skill to control units so that they fight more effectively. @JC (naval supremacist) was probably taking about sniping, which is the technique used to 1) minimize overkill and 2) kill ranged units past some melee "meat shield" I really like the fast-paced gameplay 0ad offers and I think slowing down fights would push us toward a24 gameplay, which longtime players will tell you was terrible for competitive play.

Good luck with that I've made much more reasonable propositions only for them to be outvoted.