Should Bots, Tanks and hovercraft have their own factorys?

Re: Should Bots, Tanks and hovercraft have their own factory

I don't think that factories should be massively expensive, but it should still be an important consideration whether to build an extra factory or a unit that costs as much as a factory. Also I expect that it will typically make more sense to build a T2 factory than a different T1 factory type.

I'd guess that players wouldn't be able to build more than a 1-2 factory types per planet without giving away too much resource advantage in actual unit production to the opponent.

Yes. Would rather avoid tier obsolescence. Teching up to T2 should offer a moderate advantage in unit value and will help to consolidate factory type lock in. I don't think that T3 will look much like SupCom based on what we've heard from the developers so far.

Both players would be equally blind to the other player's first factory choice so it's not really an issue. That will be the case whether there are 2 factories or 20. As long as all factories have a way of dealing with all other factories it's fine. I feel this balance has been achieved in ZK, more or less.

Having multiple possible starts gives rise to the complex interactions that keep the game fresh over time. With just 2 viable starts there are only 4 possible combinations of your own and your opponent's start, which means that once you've mastered 4 strategies, you've effectively mastered the game. When there are 6 possible starts you'd need to be aware of 36 combinations. 12 starts = 144 combinations. And so on.

Again, I'm a fan of complexity and having to think on your feet at times rather than just rote learning the optimal strategies for a fixed, small number of situations possible in the game.

Re: Should Bots, Tanks and hovercraft have their own factory

Only having 1-2 factory types (and 2 tech levels) implies that you think players will only be able to support two to four factories on (half an) an entire planet's eco, which seems pretty dubious (and implies that even one factory can crash one's economy before you have significant map control).

Exactly. Because both players have no information, what to build is a metagame choice, but in order for there to be a metagame, one has to be able to reasonably predict what an opponent might do before you've even seen them do anything. An example would be building a T1 mobile AA first from your first land factory under the assumption that your opponent will probably build a bomber from his first air factory and send it straight for your base. If there's a small number of opening moves, you can actually make the choice between an offensive start or a counter to the most common offensive start (if Bomber first happens 50% of the time, then T1 AA first has a good chance of costing your opponent a bomber for little loss on your part), but if there are a dozen opening starts, then you have virtually no chance of guessing the correct counter (8%-ish) so you may as well forget trying to counter and just go straight offence with your first move.

You're confusing "starting strategy" and "strategy". Just because you've mastered the four possible starts doesn't mean you've mastered the game. It just means you've mastered the part of the game that happens before you've started interacting with your opponent, which is the first (and least interesting) minute of, say, a thirty minute game.

Re: Should Bots, Tanks and hovercraft have their own factory

I think we need to consider the context in which we are playing the game and also the type of opponent faced. Also, this discussion is moving more into the realms of unit types and what abilities they should or shouldn't have, rather than should they be in the same factory. Where as I know that it's unavoidable that this be discussed on occasion, I think the focus here is on how the factories will be implemented in the game. Either that or the topic title needs adjusting.

[1]Starting strategy has been mentioned. There are many things other than what the enemy might produce that will determine the type of initial choice the player will make. Namely the planet type and environment. Is it mostly land or all water? Is it lava or is it open fields or rocky uneven terrain? All of these things can and should change the players initial factory choice and will also effect the choice of commanders wanting to attack this planet in the future. See point 2.

[2]By this I mean building loads of land units for an attack on a water planet is pointless. An extreme example, yes - but what about the more subtle differences between kbots and vehicles? A planet of large open fields would be compatible with both unit types. however it might suit vehicles more than kbots, allowing them to use their higher top speed more effectively and give the defending player the advantage over a player attacking with kbots.

[3]For a later game situation it becomes more about the combinations of units used as to how effective they are as well as the environment. It also depends on what the enemy has specialised in and also where we are engaging him.

On a side note, it becomes more complexed as Uber have talked in their recent terrain video about terrain transitions - suggesting that planets can and will have more than one terrain type. This makes scouting even more important to determine the types of unit that will best suit the general conditions of the area of operations. All this being pretty much off topic, of course.

EDITED x3: Need to start checking my posts for typos before I submit them.

Re: Should Bots, Tanks and hovercraft have their own factory

Nothing to do with economy to support factory buildpower. Let's say a factory costs 500 metal, while a basic light raider tank costs 125 metal. If you build a factory and your opponent builds tanks, that 4 tank advantage might be enough to swing the battle in their favour.

Tech levels compound this problem. If T2 units do offer better value, however slight, it will almost always be more sensible to go to T2 rather than building a second T1 factory.

By the time T2 units are in play, it makes more sense to build T2 units than a second T1 factory, as the T1 units from the second factory wouldn't be able to compete efficiently with the T2 units already in play.

The investment would therefore need to be made to go rapidly to T2 in the second factory type, so the cost of 2 factories would need to be considered. It will only be in the late game where it would be more efficient to build 2 new factories rather than units to the same cost, at least in an even match.

One way to make this different to previous TA-family RTS that has been discussed is to make factories a more cost efficient source of buildpower than engineer assist. That would change the whole dynamics and lead to many more factories being built.

Now we're talking about unit selection, not factory selection, which leads us back to scouting. First unit choice should be a scout. That's true for all RTS I've ever played. Scout the bomber, build the AA tank, or scout land and build the assault tank / anti-riot tank / whatever. Pretty basic.

This differs in ZK, where significant interaction with the opponent (between good players) normally begins around the 40-60 second mark. Games can be decided before 90 seconds with sufficiently effective raiding of early econ. This is one reason why I like ZK. We'll have to see how PA plays out in this regard.

Re: Should Bots, Tanks and hovercraft have their own factory

Or you could use the SupCom method and upgrade factories instead. The bonus of this is that you have to think about more than just the resources required to build a second factory in this case. It adds the extra dimension of time to the equation. A player who goes for an early T2 jump risks not having enough T1 units to defend himself against an enemy who might have gone for a T1 rush. He also needs the supporting economy for this upgrade and then afterwards to build units from it. The fact also that the factory can build both T1 and T2 units means it has more value the player and is less likely to become obsolete.

It also adds the later game mechanic that all factories must start at T1 and be upgraded. It adds more decisions when colonising a new world. Mostly along the lines of cost (in resources and time) of development and what the player wants to achieve there.

Re: Should Bots, Tanks and hovercraft have their own factory

I think that PA could work with just 2 factories for land. I think it could work without factories aswell. Just place the nanoframes on the ground and build all units from scratch there. Although I do think that there can be some interesting strategic decisions caused by the restraints of having multiple factories.

I find the strategic decision between bots, vehicles or other types of terrain dependant units interesting. If there is a noticible transition cost to get the different unit types it forces players to choose if they want to grab the hills and forests with kbots or the lowground with vehicles for example. Although I am more interested in unit interaction so having alot, if not all unit types, buildable by 1 constructor I'd welcome that aswell if there is complex mechanics as different units clash.

This is a typical example of how factory diversity can be interesting. Hovercraft might be able to hit a player from all sides but they might lack dedicated artillery and just have specialized artillery.

In ZK the redundancy in units is small even though there is more redundancy in unit roles as every unit is tied to the unit lineup in the factory they are placed in. Terrain is the restraint that makes different units more or less favoured on different parts on the map.

Why have tanks, kbots and ships when we can just use hovers instead?
We obviously agree that the terrain should be a restraint that makes some units less useful than others under certain circumstances.
Forcing the player to evaluate the terrain and choose factories based on that evaluation is an interesting strategic decision in my opinion.

Re: Should Bots, Tanks and hovercraft have their own factory

That's not how economies work. A factory costs 500 metal, but you're getting 8 metal a second, a factory drains 4 metal a second, so if you don't build another factory, you won't get 4 tanks with that 500 metal, you'll get nothing because you're already using all your build power but not all of your mass.

You're trying to simultaneously argue that T1 units don't become obsolete (because ZK is perfectly balanced according to you and every unit is useful all the time) and that T1 units become obsolete here. You can't have both.

Forged Alliance 3598:

1 T1 Land Factory = 240 Mass, Build Power = 20
8 T1 Engineers = 408 Mass, Build Power = 5 * 8 * 50% Factory Assist Efficiency = 20

You were saying?

One does not simply "Scout the bomber". The first thing you know about a bomber rush is it blowing up one of your engineers. Maps are big and T1 scouts are fairly crappy at intel. If you start building counters to openings after you've seen what's going on, you've already lost out because that counter won't be finished by the time it's needed. That is why this is a factory selection issue and not a unit selection issue, because you effectively have to make your decisions before you've even built the factory. It's all metagame.

60 seconds are a minute, you know, so this doesn't actually contradict what I said.

Re: Should Bots, Tanks and hovercraft have their own factory

We do not suggest that a factory necessarily should counter another factory. So if the player starts hovers, tanks or bots it should be evaluated by the terrain and there you have a metagame.

This sounds like it's specific to SupCom. We don't know if bomber rushes will be viable in PA.
It is also pretty irrelevant as both hover, kbots and vehicles are not as fast as planes and an enemy rush could propably be scouted before it reach your base and then you have time to prepare a counter.

Re: Should Bots, Tanks and hovercraft have their own factory

So you're suggesting players zoom in on the procedurally generated maps and evaluate the relatively hilliness of the map in order to determine whether to go with bots or tanks?

If bomber rushes aren't viable, then air first isn't viable, since there's no way of engaging enemy ground units (you know, most of their stuff) without bombers.

Re: Should Bots, Tanks and hovercraft have their own factory

SupCom is not the only game which has bomber rushes

And thats the point about rushing with bombers, your enemy has NOT enough time to build counter measures, especially not if it would require him to set up a second factory first.

He may have time to produce 1-2 mobile AA after the bombers have been scouted, before the second strike, but only real chance is to build stationary AA defenses in advance.


Btw.: I still don't see why there should be multiple factories at all? It's not about balancing as everyone will build bot factory first, simply because light assault bots are cheaper and faster than heavy tanks. It's also not about preventing synergistic effects between the different techs, you have such effects between tier 1 and tier 2, but not between different movement types.
It's just an artificial complication made under the false assumption that we would have so many units, that we couldn't stick them into only 1-2 factories. We are probably going to have a total of "only" 50-100 different units and buildings after all, that makes no more than 10-30 different ground units, already including special purpose units. This hardly justifies for having more than 2 or 3 factories peak.

Zero-K has far over 200(!) factional units, plus regular buildings and regular unit types. Thats not even close to the number of units a stock copy of PA will have.

Re: Should Bots, Tanks and hovercraft have their own factory

If this game uses similar AA vs ground weapons like TA (Tracking missiles and lasers being the only viable AA, but units can too) you could probably get 5 units per factory.

Engineer
Raider
Tank
Artillery
AA/Sniper

While this is an assumption, it is the essential set of unit types for a normal land factory

Hover/bot/spider configurations would be modifications of these unit types in ways that balance the mods.

Hovercraft losing HP for amphibious movement, Bots losing firepower for speed, and spiders losing speed for all terrain movement up and down cliffs.

As would be the differences between each unit set, not great but overall diverse for different environments and strategy's.

Re: Should Bots, Tanks and hovercraft have their own factory

I don't know where the information has come from on the numbers of units for PA. I'm assuming it's just speculation. There will be a whole group of units that fall under the shields category that will be omitted as the team have stated they have no intention of putting shields into the game. However:

Why would you assume only 5 units per factory? That means 10 factories to get the alleged 50 units (leaving roughly half that for structures). Sup Com had over 80 units per faction after a quick count on the database -I couldn't find the information readily available anywhere at a glance- so I would say that 100 units isn't far off the mark if we consider this information.

Additionally, we can consider that (hopefully) the guys at Uber will spend some time making more units with a greater variety between unit types, making each unit type have strengths/weaknesses based on the environment and maybe even abilities that are not consistent between unit types. For example, heavy arty would only be available to vehicle types. Hover, kbots and vehicles all get light arty. This, although still slightly off topic, makes sure that there is less obsolescence between unit types and promotes the idea that players would have more to choose from between unit types than just what they like to build. I would also like to see less obsolescence between T1 and T2, making it viable to use more of the units for more of the game.

I think it's reasonable to expect that if we used the sup com factory upgrade system that with a proposed six unit types (kbot, hover, vehicle, air, naval and orbital) that we could simply get away with six factories, all able to build scouts, a base of approximately 5 or 6 standard units as T1 and a further set of 4 or 5 specialised units as T2. I don't think that would be too much to ask. It's still about a third of the work required to put together the units for a game like Sup Com.

EDITED: To improve clarity.

Re: Should Bots, Tanks and hovercraft have their own factory

I'm assuming in my quoted example that both players are good enough to use all available resources and limited by metal, not buildpower. If buildpower is a limitation then yes, building another factory is one way to overcome that, although building more engineers is slightly more efficient even if buildpower/cost is equal as you mention below. An engineer takes less time to build than a factory and so can begin to contribute buildpower more quickly.

I can have both. There is no T2 in ZK. In ZK every unit is potentially useful all the time. In FA T1 units become useless with the advent of T2/T3. TA was somewhere in the middle, but closer to ZK than FA. I'd like PA to be more like ZK or at least TA than FA in terms of unit obsolescence.

Good, in FA the excessive efficiency of engineer assist has obviously been recognised (presumably after patches). However my point was that for a true change in dynamics factories would need be more efficient than engineers, not equally efficient.

In TA and ZK engineer assist is more efficient than building more factories. In another Spring game, NOTA, engineers can't assist factories at all. That's too extreme for my taste but it might be interesting to see factories given a slight efficiency advantage.

As Godde said, this is a problem for FA. It sounds like in FA you do have to make blind decisions on both factory and unit choice. I'd like rapid scouting / raiding to be a possibility in PA where two players start on the same planet. Factory choice will always be blind and I don't see that as a problem, so long as factories aren't balanced rock-paper-scissors, which they haven't been in any TA lineage game.

Well... ok. I wasn't aware interactions in FA would normally start as early as that but if they do, then that's a good thing, and I'd like to see early interactions be possible in PA (between players on the same planet).

Re: Should Bots, Tanks and hovercraft have their own factory

I'm not sure you would have to zoom in to determine that. Maybe players will recognize the type of terrain after some experience or the terrain type will be predetermined before the game starts.
Sure there might be important to notice if there are cliffs that hovers can't traverse or beaches where hovers can go up on land but that might also be a certain terrain type.

Air first can still be viable in alot of cases. For example if map is big enough so that the mobility gives the airplayer the ability to scout first, expand faster with air constructors and enough time enough to deploy counters to land units.
Maybe fighters will be able to attack ground units. Maybe there will be gunships or maybe you could just use bombers to counter their raids.

Re: Should Bots, Tanks and hovercraft have their own factory

I don't see how bomber rushes is relevant to this discussion about the number of factories unless you suggest that both land units and fighters should be buildable in the same factory.

Typically I see bots as slower than tanks with the ability to traverse steeper slopes. This gives tanks the ability to outmanouver bots on even ground while bots can take refuge in hills and outmanouver tanks there.
That is how it worked in TA with the addition that tanks/vehicles got more heavy weight/were beefier aswell.

I'm not going to speculate on how many units that stock PA will contain but I do believe that they can fill up a hover factory, bot factory and vehicle factory aswell as sea, orbital and air factories if they deem it worthy.

Re: Should Bots, Tanks and hovercraft have their own factory

Your argument appears to be based on 240 being equal to 408. I'm sure even you can see the issue here.

You can't have both, because then the hypothetical scenario you're describing is internally inconsistent.

If ZK doesn't have blind choices being made, all that means is that its players aren't good enough for a viable metagame to emerge. Building counters to common strategies only after they've been scouted costs minutes.

You've never tried to see hills on a SC map, have you? There are a lot of pretty substantial elevation changes in SC, but nobody notices them because it's extremely difficult to see that sort of detail while zoomed out.

On big maps land first is actually better, because drops decide things there, and you'll want something to put in your transport when it comes out, and going land first gives you some units which can accomplish something while you're waiting for your transport, while going air first gives you a useless transport that has to wait until you've got some units to put in it. This isn't really a thing in TA because in TA artillery drops just aren't worth the effort (8 Fobos cost 432 mass and a Courier costs 120 mass, but a Thudd costs 170 metal, and a Valkyrie costs 117).

Re: Should Bots, Tanks and hovercraft have their own factory

Well the numbers are up for discussion, but I ma glad that you understood my point.

Upgrading factory's is something I am a little shaky at, coming from SC2 where I prefer my factory's to be undisturbed, and to simply build more.

Re: Should Bots, Tanks and hovercraft have their own factory

This is a little bit of a problem, mainly because there were only about 5 units to a factory in SC2 that were kept current throughout the game by performing research. Each faction only had a relatively small number of units each and that kind of format would really not be applicable to this game because it utilises a different system - namely tech levels.

EDITED: I should add here that I don't have an issue with the way SC2 works and I don't think it's a problem in that game. I enjoyed the game and it worked well with everything else that was built around it. But I don't see it working in a game that is based more on TA and SC:FA.

Re: Should Bots, Tanks and hovercraft have their own factory

Oh. I need to stop skim reading. I thought that the final values were normalised build power/cost. That's interesting. I take it that high level players do not normally assist factories in FA?

In what way is it inconsistent? They're different games. ZK does not have tiers. I don't know how I can be more clear. In ZK all units are useful throughout the game because there are no tiers in ZK and so no unit obsolescence. The designers have called this approach "flat balance".

PA has been confirmed to have T1 and T2, which brings with it the probability that once a T2 factory is built, it will be more efficient to produce T2 units than T1 units.

Good luck building a counter for every possible strategy then. I don't understand your argument here any more. If you don't know what your opponent is doing then you can't effectively counter it. Scouting is key to all RTS.

Re: Should Bots, Tanks and hovercraft have their own factory

They do, but they only assist high tech factories, which is due to a different factor. That being that low-tech units are more efficient sources of build power than high tech ones. Which is why T1 spam builds have, like, ten factories while rapid tech builds have about four. But PA isn't likely to play anything like high tech SC due to the lack of the tier obsolescence.

The situation you were describing implies that flat balance is impossible, yet you also claim that flat balance exists. You've already described a situation where T1 and T2 are equal in efficiency, and then you go on to say that T2 will probably be more efficient.

Wow, you finally arrive at the point I was making at the beginning. Took you long enough. When there are only a small number of dominant starts (there's no point in countering every possible strategy, but every probable strategy is a small subset of that), it is possible to at least mitigate all of them. When there is a large number, it is not. And you never know what your opponent is doing because these are all decisions you have to make before you've made contact with your enemies. You go on and on about scouting being important, but what are you scouting with? When you're making these decisions you don't have any scouts. You cannot reasonably scout an opponent's first move; the numbers just don't work out. A T1 land scout can see about 2.5% of an (average size) 10x10 map. Without metagame knowledge of where your opponent is likely to send his units, your scout basically has no chance of finding any useful information early on.