Commanders: Original Vision, Crippled Default, Future ?

[aubergine]

I know pathfinding is a major bugbear but I feel there are many challenges involved in "fixing" it (in reality, making it work better than it does now) and as such should be considered a long term task. The issues apply to all units in the game, not just commander led ones, so I think we need to resolve some of the underlying issues first before we can succeed in making commander controlled units more intelligent in their movements.

[aubergine]

Another thing bolos (which is what I will call command droids from now on) need to be able to do: Ask the commander (be it human or AI) to do certain things. For example, if a bolo sees it has damaged assigned droids, but repair facilities are far away, it should be able to ask the commander to improve ability to get repairs in the field either by building local repair facility or getting some repair droids / mechanics built.

This would be done via a sort of internal chat (effectively player sending messages to themselves). With tech upgrades, HQ / CRC could possibly automate some of the responses - eg. a point may come where CRC is advanced enough to queue orders at a factory to build a repair droid (possibly even creating it's template if there isn't one already) and when built assign that droid to the commander that needs field repairs for it's team, or the HQ could set a goal of building fortified repair facility somewhere safe but in the field closer to where droids are regularly getting damaged.

[Emdek]

Maybe some of my another idea could fit here too, as some folks ranted that CRC makes not sense at all I've suggested in dedicated topic that it could be turned into module of HQ (which is also not useful for some).
viewtopic.php?f=30&t=8784

[Shadow Wolf TJC]

@ Emdek: I was thinking that maybe no building or module was needed at all in order to be able to build commanders.

[Emdek]

Personally I would let to research them and build without building, but might require one (or as module for HQ) to make them "think", as automatically pick targets etc.
Random thoughts.

[Rman Virgil]

.

I think there is an abundance of good proposals at this juncture.

Going forward I will focus on the analysis & organizing within the practical implementation framework suggested on the previous page here:

viewtopic.php?f=30&t=9119&start=30

I think that is an important practical step to eventually doing the work of making changes to prototype & actually play with in-game and not just continue with open-ended brainstorming - for as much fun as that is.

Of course those 2 practical implementation framework lists can, & will, be subjected to another round of stress-test discussion and viability analysis.

The thrust of all this is to get to a sensible doing stage and distinguishing that from what may yet remain pie-in-the-sky, however appealing, for the foreseeable future.

Feel free to continue brain-storming or contribute, if you prefer, to the near-term, long-term, viabilty analysis side of what's going on here. Both are appreciated though I will be putting my dominant focus and effort on the later for the time being and will beg-off responding to every single post and trust no one will take that as a slight on my part on the inherent merits of what they are suggesting.

- Regards, Rman.

NOTE: I have identified the 2 remaining reserved slots on the first page -

viewtopic.php?f=30&t=9119&p=97655#p97655

.

[aubergine]

I'm wondering if there are game engine performance benefits to using bolos as well.

For example, circling VTOLs are quite processor intensive because they're constantly changing the area their radar is monitoring -- if there was some sort of "air fortress" bolo that was required in order for VTOLs to use circle command, it could hover in middle of circle with big sensor that covers the effective area that the vtols want to monitor, meaning that there's only one static-ish sensor for a whole bunch of VTOLs (they effectively turn their own sensors off and rely on the bolo sensor).

[Rman Virgil]

I'm wondering if there are game engine performance benefits to using bolos as well.

For example, circling VTOLs are quite processor intensive because they're constantly changing the area their radar is monitoring -- if there was some sort of "air fortress" bolo that was required in order for VTOLs to use circle command, it could hover in middle of circle with big sensor that covers the effective area that the vtols want to monitor, meaning that there's only one static-ish sensor for a whole bunch of VTOLs (they effectively turn their own sensors off and rely on the bolo sensor).

I have always assumed there was a perfomance boost due to command group slaved units just because the units are dependent or taking their lead from the Bolo or Commander Turret A.I. It just made sense the one would be processing for the many... but I never tested or formally looked into it. It would be good to know for sure because if it's not it should be, IMHO.

- RV

.

[Rman Virgil]

Ideal life of a commander According to Iluvalar :

Level 1 (4/6 units) : The commander is subobtimal. Close to useless. It have a very expensive price for the small amount it gives to you. At least, that very expensive price make it worthy as a ~wall~ . Since it is not efficient, nobody spend ammo on that.

Level 2 (5/8 units) : It is still inefficient.

Level 3 (6/10 units) : Now you got a fair deal for your money. The bonus it gives to your units worth the price you paid. At least you are economically ok... The enemy generally prefer the "reduce" tactic. IE. Shooting the small units first and then deal with the commander because of mass effect. thx to the high price of the turret.

Level 4 (7/12 units) Now you have something ! You finally recover the time wasted at researching and planification. The commander is useful for macromanagement. However, the enemy knows it and he now try to kill the commander first. You will need to make extra efforts if you want him to survive any longer...

Level 5 (14/14 units) It became unbearable on the front line... Not only the enemy target your commander in micromanagement, but now he also do it on macro scale. Hunting him around the map. You have no choice but to retreat your commander in some defensive position to keep him alive. He now play the same role that a defensive fort. He keep a frontier close to a repair factory. The fact that he now regrouped with his units make him strong enough for the new role.

Level 6 (16/16 units) You had many of such commanders that survived... you covered more ground than the opponent with such defensive beast. If he still alive, the commanders can make the final assault all together.

final tough
I believe that mostly, the reason why the commanders are unused is because they are balance for level 6, from a single player experience. The ai dont target the commander, it doesn't try to hunt it... so of course, it become very useful near the end...

But for MP games, we need to balance it for level 3-4. Not over...

This strikes me as one of those candidates for implementation in the shorter-term for MP viabilty testing.

- RV

.

[dak180]

I know pathfinding is a major bugbear but I feel there are many challenges involved in "fixing" it (in reality, making it work better than it does now) and as such should be considered a long term task.

Pathfinding in warzone is actually not bad; the real issue is pathwalking.
It would still require a lot of work (most likely a full rewrite of that part of the code) which is why the list of volunteers to do it is currently mostly nonexistent.
That said, if anyone does want to work on this issue, checking out OpenSteer would likely be a good place to start.

[aubergine]

OpenSteer looks amazing!

[effigy]

Hmm... I could swear I posted in this thread recently. I came back to read the current thoughts and clarify my own, but I can't find a post I was sure I had made I'll do my best to repeat my phantom post.

Lots of reading to do in the links posted above I've skimmed through it a bit, and saw an interesting idea I hadn't thought of already: making the Commander beam an EMP weapon. I think it's a good fit, but sounds like better solution for late game. Though, perhaps EMP tech could be moved to earlier in the tech tree in parallel, since it's also an underused (and in ways I won't get into here, ineffective).

Anyhow, I've recently been trying to work Commanders into as many MP games as I think I can get away with. So far my experience with that has been a mixed bag.

The following thoughts I think are short term solutions, that could be implemented sooner, rather than later.

The unit cap should be removed. In the past I've suggested increasing the starting value to 10+ with a bigger increase bonus per rank. After revisiting commanders in-game, I'm now thinking that's not enough.

REMOVING UNIT CAP:

• benefits early-game on low oil maps
  • These are the first few minutes of the game when it's not a benefit to make commanders, in favor of micro'ing small groups to claim/steal oil, & before we have a decent body researched to build the command turret on. These [soon to be] obsolete units can be assigned to the Commander after one is made with out sacrificing units just so you can make better units
    • Recycling units 1 by 1 is excessive micro, and the alternative of recycling all units at once makes you vulnerable
• benefits maps with a lot of oil
  • this one is simple: more oil = more units sooner.
    • this almost always means your factories will stop producing units for a commander long before he gains rank -> unit cap goes up
• generally beneficial
  • this provides a solution for the frustration of loosing a commander so easily in MP.
    • Rank can still provide accuraccy and damage bonus, but you won't need to worry about what to do with 12/18 units when your commander is ganked.

My other "big idea" is one I've seen in the links above: the commander beam should operate as indirect fire.

INDIRECT FIRE COMMANDER BEAM

• Keeps the commander safe[r]
  • commanders could hide behind hardpoints and still tell indirect fire units to shoot, and send direct fire units to attack
    • this is a new issue, created in 3.1 targetting
  • commanders could hide behind terrain and tell units to fire/attack

[aubergine]

With regards to indirect fire on commanders, does it not make sense to have commander be an upgrade to sensor turrets?

[Mirefrost00]

Rman, your sincere attempt to apply real velocity to this congregation is commendable. I shall assist where possible by not derailing, but rather supplying thrust with directing ideas.

Regarding the CRC, one of the big concepts with Bolos is that they were always, *always, without fail*, attached to a human commander, or operator, if you prefer. Sometimes, the commander would prefer to ride into battle inside the unit, but could also stay back at Brigade HQ or a similar facility and direct the Unit just as well from there. The consoles at the CRC could be exact replicas of the command deck of a Bolo, therefore we could assume that all human commanders commune with their Bolos from the CRC.

Here's another thing, something that I should have mentioned earlier. What really makes Bolos seem like the ideal Commanders in WZ is this: in the story "Desert Fox", (which can be read here for free; I love Baen books!), Unit RML-1138 displayed a prowess for directly controlling light mechs which his operator had personally rewired to listen on one of RML's command frequencies. In other stories, Bolos have, in the event that they cannot raise HQ on comms, assumed that HQ had been compromised, and began coordinating with living, friendly troops in the field to throw the enemy's assault into disarray and try to turn the tides. Which they typically succeed at, being of comparable mass and ordnance to a battleship.

I'm not saying WZ needs true rolling fortresses, but we could handwave miniaturizing the psychotronic brains (or whatever you wish to call the tech that makes this all possible) and still get the job done. However, I would think that they really should be one of the most powerful, most heavily armored units on the field, given that we will have a limited number of them in theater.

[effigy]

With regards to indirect fire on commanders, does it not make sense to have commander be an upgrade to sensor turrets?

hmm... well, in a sense, I suppose. Indirect fire/no unit limit removes all but 1 advantage to using sensor units vs. commanders: cheap, disposable leader.

Am I contradicting myself? I'm all of a sudden remembering a posting somewhere around here about making sensors behave like commanders.