Smarter Target Acquisition

[XTF]

How doable would it be to improve the target acquisition code? I'd like my tanks to fire at other tanks before firing at dogs.

[DCoder]

http://forums.renegadeprojects.com/showt...hp?tid=938 , among other facts, "Threat Rating" might be helpful. If that is not sufficient, well, I know where the main targeting code is, it would just be a matter of adding logic to it for specific needs (but given that this is a really complex piece of logic, and everyone has differing ideas on how it should improve, it's not as simple as it sounds).

[XTF]

I don't care about AI players, so is threat rating still relevant?

How can the targeting code be found?

[¥R M0dd€r]

Air units have highest priority. If, say flak track is close(firing range) to an enemy tank and air unit, it will always auto fire the air unit(then infantry I think), then the vehicle.
A debatable solution is so the unit auto fire(those in range) the unit with highest cost.
Imo this is a minor issue and I had only this problem a few times.

[Mevitar]

Btw, there's also VHPScan=

[Orac]

What about units which are potentially a large threat but which a given unit might be very ill equipped to deal with?
I still target engineer rushes with every available unit, whether or not they stand much chance of killing the engineer.

[XTF]

Imo this is a minor issue and I had only this problem a few times.

Is it? It happens with tanks vs infantry + tanks all the time.

[DCoder]

On the other hand, if you do change the logic so that tanks target enemy tanks first, won't your "professional" player base raise hell about their fodder tactics no longer working?

[XTF]

Maybe, maybe not. They'd have to adapt.

[Orac]

But making such a fundamental change to the engine without allowing mods to opt in or out would be an odd junction from the usual implementation of fixes in Ares, which iirc are mostly things which can be individually customised (except the kennel hack being removed, which was a side effect of something else iirc).

...I have a pet peeve about people deciding that widely arching changes are a good idea and then that anyone who disagrees should simply accept the new way things are done, so please let me go on at a little length on that:
"They'd have to adapt" is the way the "lets annoy the fanbase with arbitrary changes to X which indirectly limits the effectiveness of Y when used as/with/during/before/after Z" path starts. Westwood went for this sort of route when doing so much of TS/RA2, and look how many headaches its given to the community, invoking strange limitations and even stranger workarounds (its got to be one of the big reasons Ares exists). The arbitrariness and unnecessarily hardcoded nature of the 'fix' you're suggesting just perpetuates a bad design cycle. You want tanks to target other tanks? What about infantry, should they all target only other infantry first? What about if it's an antitank infantry? Or an antiinfantry tank? This seems like a fiddly fix to a perceived problem, which only breaks some tactics and skews the gameplay further without giving control to the players/modders.
Forcing changes down unsuspecting people's throats is just annoying. Then justifying it with "Oh well, they'll adapt" is just downright incompetent, if not actively malicious.

It basically says "I am right, and you are wrong. Whether reality agrees with this or not, you have to see things my way".


In the end, if you want your tanks to attack the tanks instead of the infantry hold the left mouse button and drag select your tanks then move your cursor over to the enemy tanks and double click on them. This will cause the tanks to target the other tanks, and is generally referred to as 'playing the game'.

[Renegade]

While I agree with the sentiment, I do think you have to separate rhetoric from intent, here.
What you are doing is, essentially, building a straw man; I admit that, due to Olaf's original phrasing, said straw man is, to a certain extent, justified, but I am also pretty sure Olaf doesn't actually want to re-order targeting based on opponent classes, but was actually using "I'd like my tanks to fire at other tanks before firing at dogs" as a simple/humorous way to express "I would like my units to pick the optimal target first".

And that is quite a different thing.

So yes, if you take Olaf's statement literally, then his suggestion would be of questionable quality. If you think about the more specific suggestion, however, things look quite differently:
I believe all players would happily get behind a general change of "units now attack the units they are most effective against first", especially because that wouldn't look at something as simple as unit classes only, but take everything from verses, immunity, range and applying bonuses into account.
You can't tell me you honestly believe players will collectively be unhappy about all of their units suddenly being less stupid.

Similarly, you misconstrued Olaf's comment that "They'd have to adapt" as a general "screw the others, I want it my way" remark, when, in actuality, said remark was quite obviously made in the narrow context of tactically-challenged n00bs on XWIS. Again, a straw man argument.


I acknowledge that textual communication poses certain challenges, and I admit that it's everyone's own responsibility to ensure he is communicating clearly, however, one can expect a certain amount of goodwill and common sense from the reader.
Therefore, I would appreciate it if you read twice before getting your torch and pitchfork next time.


I would also like to point out that Ares has a number of universally applying changes beyond the kennel hack removal, such as Secret Lab Boon Weighting/Too Many Secret Labs, Hijackers are reimbursed when a unit is grinded, Unit Instances Not Counting Towards BuildLimit and others. Point being, while we usually do offer a way for the modder to decide which way he wants it, there are multiple precedents for universally applied changes, and you can construct debates about each of them. Ultimately, it's a question of "how likely is it there are people who'd want the old behavior?".
And it's not very likely there would be people going "Heyyy! How dare you make my units smarter without asking me?! I liked them wasting their effort and letting the enemy destroy me!!".


So yeah. Feel free to discuss the merits of a general targeting optimization and whether it should be optional or not, but please don't clutch to the clear simplification of "tanks must shoot on tanks first". It's an image. A symbol. An example. Nothing more. Nothing worth fighting over.


(Also note that this is currently nothing but a "Would it be possible..." discussion, and there is no promise any of this will ever be implemented by anyone. Just sayin'.)

[Orac]

Ah, I understand.

[XTF]

Ah, I understand.

Renegade already addressed your post, but I'm still curious, why did you make all those negative assumptions?
Did you misunderstand the context of this request / topic? I merely posted quite a general idea and you make all kinds of specific assumptions.

[Orac]

Well, it seemed to me that your idea would come down to some sort of hierarchy of which targets to attack first and which ones to ignore, which would result in some sort of blanket change which would of course not be able to cope with some outliers (and modders just love to add outliers).
This mostly came from the "I'd like my tanks to fire at other tanks before firing at dogs." comment, which set the baseline from which my assumptions were working, that what you were thinking of was a stringent divide between different units and different roles.

Then in your later post, I might have taken you slightly out of context to some extent, to which I also apologise. It was mostly used to fuel my already strong aversion to the viewpoint that "everyone should accept my point of view", which might in hindsight not have been what you were aiming for. The bottom line was that I didn't (still don't) like fixes to things which aren't technically broken. While I agree that sometimes there are flaws in it as target acquisition as it stands in RA2/YR, the benefits of a different system seem negligible to any player who is even half awake and willing to reassign units to different targets. Again, maybe my assumptions are wrong.

The other thing I was trying to get my own head around, and which I didn't actually articulate, is what would you replace the "old behaviour" with? I can sort of see the calculation of what the max/min/average % of the enemy's health which a unit's weapon can remove being a good place to start, since it would be reasonably sensible in any circumstances. But I'm still not sure where you'd go from there since there are numerous cases where the % is misleading.

[XTF]

Let's first clear up the context: Red Alert 2 Multiplayer on XWIS
Yes, RA2.

While I agree that sometimes there are flaws in it as target acquisition as it stands in RA2/YR, the benefits of a different system seem negligible to any player who is even half awake and willing to reassign units to different targets. Again, maybe my assumptions are wrong.

Have you played online competitively?

What would be the ideal tactic?
You could take into account only yourself (the unit searching a target) and enemy units. But you could also take into account potential targets/threats of enemy units (engineers, snipers, etc, an engineer is harmless when he can't capture anything, an anti-air unit is harmless when ...).

The first step would actually be documenting the current behaviour.
The next simple step might well be just adding an exception for anti-tank tanks (Grizzly, Rhino, Apocalyps) to see how that'd work.
After that you could think about using weapons effectiveness.