Proposal: New Commander Morph System

The other day CatLady was chatting with GoogleFrog about Commander balance, and GF had this to say:

https://discordapp.com/channels/278805140708786177/278908415756206080/591106219800920064

quote: GoogleFrog: the big danger of no design is that the game will just randomly walk around design space GoogleFrog: the design does not need to be complicated GoogleFrog: but the failure to make a simple coherent design for what commanders should do is a warning sign that making changes will lead to something incoherent

I had posted some thoughts earlier about Commander design in the recent thread here. I've fleshed them out somewhat into an actual proposal and I'm seeking discussion and feedback. It's just a first draft, and there's plenty still to do. But perhaps it's a starting point towards something better than we have now.

Here is my proposal: https://zero-k.info/mediawiki/index.php?title=User:CrazyEddie/Draft_Commander_Framework_Proposal

TL;DR:

For every 100 metal in cost spent on morphing a Commander, the comm will gain 75 metal worth of unit value (UV) and 50 metal worth of economic value (EV). The unit value is an approximation of the metal cost of a comparable unit, particularly in combat. The economic value is based on the metal and energy produced by the comm. This rule ensures that morphing a Commander is cost-effective, but only under the right circumstances.

The morph levels are more rigidly structured. The number of levels is limited, capped at Level 6. The levels grow progressively stronger, and each level in the new system is significantly stronger than the corresponding level in the current system. Starting Commanders (Level 1) are roughly equivalent to an Ogre, Felon, or Jack in combat strength. The highest-level Commanders (Level 6) are substantially stronger than a Grizzly, Likho, or Cyclops, but still substantially less powerful than striders such as a Scorpion or Dante.

Modules are organized into tracks, with the modules in each track growing stronger as the upgrades progress along the track. Each track has three levels of module: Minor, Major, and Ultra. The lower module levels are prerequisites for the higher module levels.

Like modules, weapons also have Minor, Major, and Ultra upgrades. This means that every primary weapon has four different levels (base, I, II, and III). Every secondary weapon has three different levels (I, II, and III).

We can use the existing weapons and modules as a starting point, but they will need a lot of work to rebalance. The goal of having this design is to ensure that, as GoogleFrog cautions, the weapons and modules can be balanced without leading to something incoherent.

The upgrade sequence does not explicitly say what kinds of choices players are making at each step. It turns out that each module upgrade has a choice, but there are no choices in the weapon upgrades.

Level 1:

• Pick a chassis.

Level 2:

• Pick primary weapon.
• Pick a track to start.
• Pick a track to start.

Level 3:

• Upgrade a choice of two tracks to rank 2.
• Pick a track to start.
• Pick a track to start.

Level 4:

• Pick a secondary weapon.
• Upgrade a choice of three tracks to rank 2.

Level 5:

• Mandatory upgrade of both weapons to rank 2.
• Upgrade a choice of two tracks to rank 3.
• Upgrade a choice of two tracks to rank 2.
• Pick a track to start.

Level 6:

• Mandatory upgrade of both weapons to rank 3.
• Upgrade a choice of two tracks to rank 3.
• Upgrade a choice of two tracks to rank 2.

I think you have bitten off more than you can chew with the concepts of stand-alone modules and heavy weapons that occupy both slots. Where do the stand-alone modules fit?

Weapon upgrades should not simply be automatic because players won't even realise that an upgrade has occurred. Is the presence of a heavy weapon supposed to change the structure of the upgrade slots? Heavy weapons just seem bad given that they remove the choice of primary weapon made back at level 2.

To solve the lack of choice in weapon upgrades and the stand-alone modules you may want to consider choices within tracks. Perhaps a player could take Rez as their rank 2 Build choice. Perhaps rank 3 even has a choice between plain build or less build with a little extra build range. Weapon upgrades could be a choice between some raw power or some utility, as with the current slow and fire weapon mods. It doesn't have to be that complicated. To start with, the special weapons could give a choice between more power or shorter reload time.

quote: The upgrade sequence does not explicitly say what kinds of choices players are making at each step.

Your analysis listing the choices made at each step matches my intention; sorry it wasn't clear at first.

I made the weapons upgrades automatic because it seemed like the right way to build up to a monster comm at the highest level. Much like the current morph system builds up HP whether you explicitly buy additional HP boosts or not. It seemed like aesthetically, if your comm is getting bigger and badder then the weapons should be growing along with it.

It would be easy to switch that, though, so that - for example - at L3, instead of choosing a Major and two Minors and getting the Minor weapon upgrade automatically, you could instead choose a Major and three Minors, and have the possibility of leaving the weapon unupgraded and instead start three new tracks instead of only two. Or even let the player go shopping: you've got 300 points to spend, and the things you can spend it on come in bundles of 50, 100, and 150 that have to be purchased in sequence, but otherwise, take your pick.

I do think that weapons upgrades should be distinct from generic boosts, so that each upgrade of a specific weapon makes that weapon "more like what it is" and not just generically more DPS or more range.

quote: Where do the stand-alone modules fit?

They break the "upgrade along a track" pattern. If, say, Lazarus Device was an Ultra module, then you could add it at Level 5 instead of upgrading one of your Major modules. It would mean one of your tracks doesn't progress as far as it otherwise would, but in exchange you get the one-off module of (hypothetically) equivalent value.

quote: Is the presence of a heavy weapon supposed to change the structure of the upgrade slots? Heavy weapons just seem bad given that they remove the choice of primary weapon made back at level 2.

I included the heavy weapons because the current system includes them (heavy particle beam and shock rifle). I envisioned that they would work essentially the way they do now - if you choose a heavy weapon as your secondary at Level 4, then your primary disappears. Upgrading to Level 5 and Level 6 would upgrade the heavy weapon to II and III, and the stats at those levels would reflect the combined UV of both weapons slots.

... but this is very handwavy, and very subject to adjusting the numbers as we try to make this work.

The wiki page seems to be where the design goals and constraints are kept.

I find some of the goals to be somewhat in conflict. Not in a hardcore unresolvable conflict, but definitely in tension that can and should be addressed:

Namely (numbered and cherry-picked for my own convenience)

1) The Commander should remain an important unit throughout the entire game.
2) The Commander's role should be active, rather than passive.
3) Not using morphed comms should also be a viable strategy.
4) The outcome of the game should not be decided solely by the death of the Commander.

First: if the commander dies, the game either goes on without the commander, or it doesn't.
First case breaks rule 1 because dead commander is not an important unit in this game anymore. Second case breaks rule 4.
The current state does not have this problem by not requiring commanders to stay relevant.
Could this be addressed by commander rez becoming a worthwhile or accessible thing?

Second: Rules 1+2+3 give us a situation in which an unmorphed commander is expected to stay relevant as an actively used - that is, frontline - unit.

I guess this could be addressed by giving even baseline commanders some scalable abilities.

However, there's a current design rule that commanders should not have abilities not possessed by at least one other unit; this means that whatever scalable ability can be granted to level zero commanders to make them relevant even while unmorphed as the game progresses, it can be obtained for cheaper through building the unit with that ability.

quote: No idea, but I know Anarchid has some thoughts around this.

Replacing stacked modules with tiered modules is easier to visually represent. Especially if the number of tiers is limited to 3, and each module is designated a specific body part that represents it.

Anarchid - Good point about some of the design goals being in tension.

Rules 1 and 2 should probably be phrased as "It should be possible for the Commander to remain an important unit throughout the entire game" and "It should be possible for the Commander's role to be active, rather than passive".

Currently it's not possible because morphing the commander is not cost-effective. (Okay, strictly speaking it's possible but it's a bad idea.) The point of Rules 1 and 2 is that these playstyles should be feasible, even encouraged. The point of Rule 3 is that these playstyles should not be mandatory. The Framework Balance Rule fulfills these goals by making comm morphing a strategic decision that needs to consider the risks and benefits.

quote: First: if the commander dies, the game either goes on without the commander, or it doesn't. First case breaks rule 1 because dead commander is not an important unit in this game anymore. Second case breaks rule 4. The current state does not have this problem by not requiring commanders to stay relevant.

There's levels of importance between "The comm is an expensive nanotower" and "The game ends when you lose your comm, like in TA or SupCom". Right now ZK has the first; I'm aiming for something more than that, but not literal game-ending.

But besides literal comm-ends, there's also the concern that losing the Commander will be a decisive loss. Losing any very valuable unit is a setback but is usually not decisive; Rule 4 is there to ensure the same is true of the Commander. GoogleFrog also raised this concern in Discord. My hope is that with proper design of modules and tracks it will be rare to have a Commander build whose loss cannot be recovered from.

quote: Could this be addressed by commander rez becoming a worthwhile or accessible thing?

GoogleFrog also mentioned this in Discord, suggesting that there be some way to build a new comm after losing the first, either through rez or by morphing storage (!) into a new comm, picking back up where you left off.

quote: Second: Rules 1+2+3 give us a situation in which an unmorphed commander is expected to stay relevant as an actively used - that is, frontline - unit. I guess this could be addressed by giving even baseline commanders some scalable abilities. However, there's a current design rule that commanders should not have abilities not possessed by at least one other unit; this means that whatever scalable ability can be granted to level zero commanders to make them relevant even while unmorphed as the game progresses, it can be obtained for cheaper through building the unit with that ability.

Softening Rules 1 and 2 as I did above resolves this.

The heart of my proposal is that morphing commanders is a good deal if they are used and kept alive, and a bad deal if they either stay safe in the base or are thrown away on the front lines like cannon fodder. I would not expect unmorphed comms to stay relevant as an actively used unit on the front lines. But the decision to not morph the comm and keep it safe in the base is a decision by the player, and is a valid one.

The Framework Balancing Rule ensures that Commanders have two roles, either of which could be fulfilled by building other things instead (units and eco). Morphing the Commander is not cost-effective if the Commander subsequently fulfills only one of those two roles, but is very cost-effective if the Commander fulfills both of them.

Thanks for starting on the fundamental design Anarchid.

The design goals and constraints are far too broad to point the design in a particular direction or to constrain anything. The goals are reasonable but basically amount to saying that everyone will be happy regardless of how they want to play. I expect map and game size variance to be too high to even support the goal of letting people do whatever they like with commanders.

Adding economic value to the commander is very worrying.

• It trivialises much of the energy game, with knock-on effects for constructor logistics and raiding. If you barely have to make energy generators then all bases and expansions will look similar and there will be fewer targets to raid.
• Losing your commander means losing your economy. You may as well have lost the game.

I doubt that early overdrive is enough to promote energy structure construction.

Keep in mind that value created through combat morph and leveling is much more powerful than its unit equivalent.

• Combat morph value appears on the frontline immediately. Units take dozens of seconds to arrive.
• Unit value is attained at intermediate morphs as your commander walks around and does things.
• Combat morph does not require any infrastructure. Remember the cost of the assisting constructors.

The ability to arrive immediately upon completion is part of what makes turrets so powerful.

With both these considerations I expect commander morph to be mandatory at 50 EV and 75 UV (for 100 metal) provided that your commander can walk to the enemy base in, say, three minutes. With 27 energy income for 1200 metal you should be able to mostly ignore energy production. The savings on energy and base constructors should easily get you a level 4 commander within a few minutes. For 1050 UV you can approximately buy a Minotaur which is as good against raiders as a Mace, with a Welder sidearm. This unit also ignores Lancaster's Square Law.

I think a possibly workable overarching role for the commander could a combination of all or some of these:

1) Early defense riot-assault.

This is fairly obvious: a commander in place should make that place rather expensive to attack with raiders due to its innate raider-defeating abilities. This can give us these constraints:
- Commanders should not be easily able to escape the riot-assault role early on.
- Commanders should be fairly rioty by default even without user intervention to make them more rioty.

2) Secondary tech pick.

A faction of one

This is somewhat less obvious. The commander morph system as-is already implements more of a proper tech tree, complete with tech tiers, than the rest of ZK combined. The commander is also a pregame tech/race pick much as the factory: in current ZK, at the start of the game, the player picks a factory (with its tech) and the commander chassis (with their module loadout restrictions).

I feel that these two things could work together to a further extent. I really like how raaar uses commander as part of an army composition, performing an unique role that synergises with the rest of his units. I feel like this should be the default way for commanders to function: providing tech that is not available within factory in a way that synergises with the units of their factory. Here's some examples:

- Godde using a HMG commander back when Redback did not exist to provide riot-dps to his spiders loadout.
- Raaar using a long-range missile commander to snipe enemy skirmishers and long-range-fire-supports to allow his Recluses to dominate everything else
- A quick dgun build providing antiheavy utility to factories that lack it
- An area cloaker with inbuilt repair facilities

This can be formulated into these goals and constraints:

- The commander should not have entirely unique abilities, but
- The commander should be able to equip a lot more utility tech than is possible now: djinn, capture, gravity, lobster throw, spider-leg, anti-air, black hole, float, anti-sub, whatever is needed to plug that annoying hole in the factory loadout.
- The commander should not be a comparably cost-effective source of those abilities compared with regular units, nor should commander builds be much more powerful than the units the abilities were taken from.
- There should be less quantitative modifier modules and more qualitative ability modules.

As with losing a factory, losing this tech unit loses you the tech that it was providing to your army composition. But as with losing a factory, it should be able to recover the key functionality by rebuilding or resurrecting the commander.

Things that are "too universally good to be locked to a factory" and are implemented as morphs from statics could instead be commander abilities. Sparrow could be a commander recon drone ability of some sort, for example.

Under this lens, i don't think any commander income - including what we have currently - is a warranted or desirable feature; it should be either ejectable into a vanguard socket building, or otherwise detached from the commander.

3) Alternate counter provider.

We have adapted

This is pretty much 2) but remembering that this is ZK and we have flat tech.

If the new design even has chassis, you're still stuck with the chassis you have, but you could be able to morph your commander to counter some bits of the enemy composition, provided adequate intel, lead time, and resources to afford it.

This could somewhat help with the "stay relevant" goal: but instead of staying relevant by growing in power and importance, in this mode the commander stays relevant by staying flexible.

This morphs the system with the following extra rules:
- Modules should not mostly be level-locked or have prerequisites.
- Modules could be possible to reclaim and respec.
- To maintain sanity and ease of use, modules could be assigned slots, represented visually as body parts, both in UI and in-world.

Instead of an incremental RPG skill tree, we could have a somewhat swappable mecha loadout. Less Diablo; more LANCER.

(Or, less Diablo Skill Tree, and more Diablo Inventory Doll)

"Currently i have a Shield Generator Crown on my head, but an ally made a bunch of snipers, and an Intel Visor would allow me to spot for them - so i'll use the Head slot for that instead, by selling the Shield at a discount, and installing the Visor"

Under this lens, the levels are meaningless and provide few if any innate bonuses. Modules almost entirely determine what a commander is (and what it looks like, hopefully). There's less incentive to keep getting levels because eventually your kitchen sink is complete and you can't add more forks.


TL:DR

Totally Loco: Do Radical

- Remove economy except for buidpower.
- Lock early states into a default riot-assault loadout.
- Be relevant by providing tech not available to your factory.
- Stay relevant by staying flexible (by being able to swap loadouts).
- Be possible to rebuild or respawn without permanent malus (like current loss of income).
- Don't overshadow units for cost: getting a unit ability via factory/unit should be cheaper than getting it via respawn/respec.