No more: 100 engineers around a factory

While everything erastos said is correct, I think I should make a few comments regarding this "flexibility" argument that people are making about engineers.

When you say engineers are more flexible than factories, you are implicitly acknowledging that the same build power in engineers and factories is not worth the same amount. The same BP in engineers is more useful because of engineers' additional flexibility. This is the best argument for why engineers should be less efficient sources of build power for cost- because you are getting more value with each increment of BP from an engineer.

Also, I think it is important to point out that we want players to be able to choose between super-lean, efficient production, which gives them more actual hardware, and having greater flexibility. That should be a tradeoff. Adopting a more reactive playstyle involving scouting what your opponent is doing, and gearing your production in reaction is a powerful style, but you should be paying a premium for the ability to do this quickly.

It is important that the game not allow players to be too flexible with their production without incurring prohibitive indirect costs. Absolute flexibility, the ability to react immediately to any possible movement or action by your opponent, should carry such high incidental costs that most of the time it is unwise to adopt such a stance. You would need hugely excessive amounts of buildpower and stockpiled resources in order to insta-build whatever you find appropriate. It is more efficient to have few idle resources, and less total BP, and spend them continuously.

And lastly, the final component of the industrial equation is time. I am of the opinion that it time should be the most expensive kind of resource for which you can make an exchange, and the most significantly compromising resource to waste. What I mean by this is that it should be very prohibitively expensive to adopt as strategic doctrine the idea that strategic moves, or countermoves, can be reacted to with such speed that you don't need to have already produced the relevant assets. The time required to produce the units or other assets needed to respond should be significant enough that you need to be preemptively ready to respond, or else you are out of luck.

Put another way, if you are building anti-air when there are already planes overhead, then you have missed the time to correctly counter that action by your opponent. You need to have an army already in place, or defenses, or other countermeasures already on the board before they will be useful to you. When an enemy makes a move, your response should be to maneuver or utilize assets you already have. Waiting until the move has already been made to respond to it with production should be severely punished, as you were totally unprepared when your opponent made that move, and you should pay the price. This isn't to say mass assist isn't useful, especially for rapidly ramping up a particular project or type of asset, but it should not be a crutch to allow you to respond to threats for which you are utterly unprepared.

This is what I've been trying to say all along, but it got kind of lost in the argument on whether 'instant building' was really instant.
Late game FA gameplay does NOT punish you for being unprepared, and you can hedge your bets big time by simply building one unit over and over (engineer). I believe this contributes heavily to a gameplay style that favors economic grinding, since it reduces the penalty of poor strategy down to a minor inconvenience.

Why does everyone who is against assistance ignore continually straw man me with this 'react instantly' stuff? I have said repeatedly that there should be a window of opportunity to respond if you scout the enemy. This means getting a visual of the nuke launcher early in the build. Or spotting that they are pumping lots of resources into their air construction. It does not mean instabuilding anti-nukes while the missile is in the air. That is obviously stupid, and impossible anyway.

That's just not true. Late game threats require vast amounts of resources to produce, that's why they're late game threats. They also require proportionally large investments to investments to counter. If you have insufficient AA and try to instabuild defences as the strategic bombers enter your base, you will die. If you haven't bothered building an anti-nuke when you hear 'strategic launch detected' you'd better hope your ACU is underwater but either way that base is gone. If you have no land units when the army of Bricks start walking out of the ocean next to your base they will tear it apart long before you can produce anything to counter them.

Read my previous posts. I am in favor of unlimited assist (i.e. on your side). All the problems of previous assist systems stem from slight statistics issues, that's all. The only thing I think needs to be done is make engineers slightly less efficient sources of BP/cost than factories. You are thus paying a little bit more for assist power than for factory power.

Read mine (the one you said 'everything erastos said is correct' about). I specifically said that an argument can be made to adjust buildpower/cost ratio to favour factories.

However buildpower/cost can not be analysed in isolation. You point out that engineers flexibility is an advantage. Definitely true. Their relative fragility is a weakness which also needs to be considered. As does the cost of teching up. There are knock on consequences to messing with any of these ratios.

The rest of the arguments you make are based around a premise which simply isn't true: in TA or supcom no amount of engineers will allow you to react instantly to any viable threat unless you are so far ahead economically that you are capable of crushing them at any time. That's a pathological case and if you're that far ahead you should be unstoppable. If you are woefully unprepared when a competitive opponent attacks, you will die as you should. If you scout well and adapt your strategy in time you might not. And on the flip side, if you scout and spot a weakness having some flexibility gives you the chance to try and take advantage of their mistake. You exaggerate the utility of flexible buildpower far beyond its actual value in the existing games in your argument to nerf it.

Erastos, you and I obviously have a vast difference in understanding of what 'late game' means. I think the poster who mentioned his record of 3 seconds to put up a fusion has it on the mark.

MY late game scenario involves a map like Seton's Clutch, where by the 40 minute mark my base would require 5-10 nukes to disable it, and I have spread out across my entire peninsula.

An army of Bricks coming out of the ocean 'next to my base' will encounter a line of defences.. and then another one.. and then another one.. and I will continue building defences up until your attack is blunted, long long before you reach anything critical in my base. Even if you manage to damage some infrastructure, the wreckage your experimental or force of heavy units leaves behind is plenty of mass (energy isn't even close to being an issue anymore), and it will only take me a minute or two to replace what was lost, and then my base will be stronger yet because I will have a new layer of defences to counter whatever you just hit me with. Heck, I've weathered situations where multiple Scathises were firing on my base, simply by building more and more shield generators ( another reason bubble shields are bad).

The ONLY strategy you as my opponent would have remaining is to build up such a massively huge force that you completely overwhelm my ability to build more defences. It doesn't matter where you attack with those forces, and all 'strategic' choices are out the window, you just have to brute force your way through to the end. He who continues expanding his economy the most, wins.

All this is behind my statement 'Engineers allow you to build an instant response to any attack'.


Your scenario seems to be one where A) Your base is small enough to be critically damaged by a single nuke B) Your economic infrastructure is very close to the front line.

Neither of which apply in a game where you're talking about fighting across planets!

replay or gtfo.

Also if its a no air no nuke no rush setons i'm gonna tell you to stfu.

So you're saying that you would already be prepared? You wouldn't need to instabuild defences because you already have them? So the thing you've been complaining about (the ability to instantly counter anything you didn't see coming because of massive amounts of flexible buildpower) is utterly irrelevant because you've built the counter in advance.

Except these points have nothing to do with each other. You are not 'instantly responding' to threats you weren't prepared for, you are building massive defences in advance. These are different.

Actually my scenario is one where an unexpected threat appears which I am not prepared for. You know, sorta the situation we've been discussing? Rather than an ultra-turtle-fest on the most easily defended map in the entire game where you have already built every conceivable defence so there can be no such thing as an 'unexpected threat'.

Yes, in pathological cases supcom does devolve into an economic grind to produce an utterly overwhelming force. However that has nothing to do with engineer assistance. If you couldn't assist you'd just set one engineer to build each of your hundreds of shields and turrets. That scenario is primarily due to layered bubble shields and maps with very limited choke points. And guess what? You can avoid this scenario pretty much entirely by playing almost any map that isn't Seton's Clutch.

Erastos, the argument I am making isn't that a player will ever actually have the situation where they can react "instantly," a close reading of that section will show that I say towards that end of the spectrum costs and inefficiency become extremely prohibitive. While theoretically possible, as in the case of the 3 second fusion, it is so wildly inefficient that it doesn't happen in real games between good players because that 3 second fusion represents a vast waste of resources unspent before then, and excess build power already produced.

It is possible in TA and SupCom to spend resources to increase your ability to build quickly. This is a good thing. At the extreme end of the spectrum, a player has spent so much, and has so much of this capability, that they can 3 second fusion or other ridiculousness. That would be possible even if engineers were slightly nerfed in BP/cost, so factories are preferable.

However, no amount of factory production will in any way contribute towards your 3 second fusion.

The only effect of this "nerf" to engineers' build power per cost is that in serious games between good players, there is an interesting tradeoff between power and flexibility. All factories all the time yields a LOT of build power, but it is not very flexible build power, and can lock you into a specific strategy. All engineers all the time lets you build whatever you want, fast, but is considerably less efficient in terms of BP yield for resource investment, meaning you will have less stuff, but will be able to build particular assets more quickly. Obviously in a real game, both will be used, probably in reasonably balanced numbers.

Contrast this with a case where engineers are equal or better than factories in build power per cost, where serious games between good players will obviously involve as few factories as possible (when unique location requires) with as many engineers as needed to spend your economy, simply because they are at least as good at building units for cost as factories.

Erastos, once again, you've missed the point completely.

I don't NEED to build any more than just enough defenses to slow you down, because if I have a swarm of engineers and stored up mass it's more efficient for me to lay down defenses right in your path, rather than pre-build a massive set of defenses in areas I'm not sure of. In the time it takes you to overwhelm my outer layer, I can have a stronger line behind it, and then another line behind that, etc etc. The harder your push is, the more of my economy I can divert to my cluster of engineers and the faster they can build defenses. And then reclaim your army when it eventually succumbs.

Why SHOULD I build a huge line of defenses when I can do this?

I mean, c'mon, this is basic maths people. Unit costs are fixed. Resource income is not. The more resource income you have the faster you can create more income. This is called an exponential economy. If you have a fusion costing 1500 mass, and you're making +500 per tick, it's not wasting anything at all to be making fusions every 3 seconds. If we're talking about 100 engineers around your factory, we're talking about this level of economy and the issues it can bring.

Stop making arguments based around 10 engineers.

Ledarsi, I have to argue that in one respect you are incorrect with your statement. A large set of factories allows you to rapidly increase your mobile build power, and also makes your base that much more resilient because you can replace any destroyed engineers quickly.

Oh, I'm sorry, I honestly didn't think anyone could possibly think that could work. As I said initially, if you try and build a wall of defences as an army of bricks emerge from the ocean your engineers will all die long before you can build enough defences to matter. In order to even come close to pulling this off you'd have to have literally hundreds of engineers - which would break the pathfinding so badly you're back to not being able to do it.

Let's do some basic maths. The ravager does 328 dps. A brick has 9000 health. If you could somehow drop a ravager every 3 seconds (I'm actually being slightly generous to you here, that'd take 600 mass/second) and would require literally 100 engineers, with all the pathfinding impossibility that would create.

Starting from nothing, it would take you 12 seconds to kill your first Brick. In that time each brick has done 4500 damage. I initially said 'an army' which isn't very precise but implies many. Let's be generous and call it only 30, so that is 135,000 damage. Each of the 4 ravagers you've built so far has 7500 health, so we're down to 105,000 left to wipe them all out. An engineer has 150 health so that takes us all the way down to 90,000 excess damage. Admittedly, there will be a bit of overkill (not much, bricks aren't very front loaded) and some bricks might not be able to fire fight from the start but we're talking more 3 times the damage required to kill everything. That's also ignoring the way your engineer swarm would be shrinking under fire and the minor detail that they'd never actually complete anything because it would explode while being constructed.

Ok, so what if you built a shield first? Let's say you (somehow) get a seraphim T3 shield up, you know how long that buys you against 30 bricks? 3 seconds. Or to look at it another way, it takes up another 30k leaving us with 60,000 excess. It also takes you 7.5 seconds to build with your 600 mass/second economy so I think it's pretty clear you won't be getting enough of them up to make any difference. And even if you could, the Bricks can weather your fire long enough to just walk under it.

Oh, and by the way, those 30 Bricks took your opponent's 600 mass/second economy 64 seconds to build. Not exactly a strategic investment.

But let's ignore all that. Let's imagine that you somehow could focus 1000 engineers in one spot and drive them with a 6000 mass/second economy and you didn't have to worry about moving the engineers around as they build and you actually could build a defence that'd stop 30 Bricks after they crawl out of the water. You know what happens then? Instead of 30 Bricks the attack is 24 Monkeylords because your opponent gets the same scale of economy you do. Threats scale up. Buildpower that could instantly turn aside mid game threats is all but useless against late game threats unless you use it in advance. So we're back to scouting and reacting within a window of opportunity.

You're right, it is basic maths. It's just that your maths is wrong. As I keep repeating the scale of the threat goes up as time goes on, which means the scale of the defence has to go up too. You can stop a bomber at the very start with two mobile AA. At the start that's a significant chunk of your available resources and buildpower, not something you do lightly. In an ultra-turtle setons match with utterly ludicrous economies you can stop a wave of Bricks with a massive wall of shields and PD, which is once again a significant chunk of your available resources and buildpower.

There absolutely is a cost to building a fusion every 3 seconds when you're at +500 mass. It's called opportunity cost - while you're laying out your fusions you're not building anything else, and while excess energy isn't disastrous what you're actually doing is wasting the initial 1500 mass cost plus the 5400 buildpower for each fusion if it's not being used.

Oh, and that 100 engineers thing in the thread title? That's what's called hyperbole. The OP doesn't like assisting factories and so gave a number that's unrealistically high. In the real world 20 engineers around a factory is high due to space, range, and pathfinding issues.

Me, too.



As for the topic, there should be a more efficient way to upgrade factories (but not for free, SupCom2's factory veterancy was very slippery-slopey). You only need those huge bunches of engineers because engineers in SupCom and even TA are, mass for build power, the most efficient way to build things. Factories sucked at the only job they had (build armies all game long!!!) and their upgrades were costly and the build power increase was insignificant / should have been detached from the tech level from the start.

A factory build power upgrade that increases build power equal to assisting with an engineer, and that you could queue a large number of times (at least 20 or so, better 100 or unlimited) and that's cheaper than said engineer (because it's less flexible!) would easily remedy the engy spam issues.

Something like that could work, though I do think engineer vulnerability needs to be considered as well as the raw mass/buildpower ratio. My gut feel is that making them both similar in mass/buildpower could end up working well - more factories makes your build capacity far more resilient at the cost of flexibility. Although that's true even with TA and supcom's bias towards engineers - with a single T3 factory if it blows up you're kinda screwed until you can get it replaced.

I'm not so sure a fast-building, static tower a la nanos in BA/ZK is really a good design. Because having a specific unit for each specific use is a bit boring - it's better that units can do more than one thing well and can be re-tasked if needed.

If most efficient buildpower is factories - then you'll get a lot of factories which easily opens the door to multiple unit types. Potentially interesting.

If mobile constructors are the best, you'll build them and assist. Most of the time the ones on assist duty will also stay there, but you've got the option to send them out to build too (instead of waiting to build new cons), but at the cost of greatly reduced unit buildpower.

If a static base builder tower is most efficient, they can do nothing else but build units, economy and defenses near the spot they're on. This is a very straightforward unit, you have to pay extra for any versatility, even when versatile units arguably make for more interesting gameplay.


I think the best way is to have just cons and factories, where factories are only slightly less efficient in cost:bp ratio. Though a builder tower has merits in ease of use, maybe give the cons an option to temporarily root themselves in place to gain more building range.

Construction towers can play a good role in your base, that's true.

But think about making them more efficient at unit construction than construction plants. If something's more efficient, then that's the way to go. It would be foolish to spend more resources with less return value.

So instead of a base with 100 engi's around the factory, you'll end up with 100 contowers around your factory because it's just better than a corresponding amount of factories. Better than engi's, but still...

Why do we need all those clunky intermediate solutions to improve a factory's efficiency if we can just improve the factory's efficiency?

On the subject of flexibility which still seems unclear:

If your factory is more efficient than engi's, then the right way to change from mainly air to mainly sea units is to build (no surprises here) additional harbours. If a harbour can construct ships at the rate of X engineers and the cost of X/2 engineers, then not building additional harbours is foolish. Since you normally have a lot of engi's late game, or you can construct engi's very fast in your efficient factories, it doesn't take that long to build new factories.

It's the most intuitive without intermediary solutions: need more ships => build more harbours.

If you've got a swarm of engineers, why not have them assist your factories?

It's the most intuitive solution.

Also when you say improve a factories efficiency, I think you mean improve a factories buildpower. Improving its efficiency is another thing entirely.

So... You've completely missed the point about how flexible buildpower is a good thing because it lets you vary your focus based on changing circumstances within the game, rather than locking you in to a particular strategy as soon as you decide which factories you're gonna build. Nice work on ignoring the last 18 pages of discussion! I'm seriously impressed that you can miss the point for that long!

Erastos, lose the attitude, please. How old are you?

Every comment you make is just "Blublublub you always miss the point you're whining you can't make an argument you're unable to understand".

Go to the zoo if you want to feel superior, though I guess the monkeys will still be laughing at you.

Is it really that hard to discuss without having to resort to insults? Grow up, dude. I'm not replying to that filth anymore.

@Zordon
End-game you will have a lot of engis in any case, because they're good for building/repairing and that needs to be done in end-game as well. If you have a lot of idle engineers, then you've built too many. If you're assisting a factory with X engineers while they're less efficient than X/2 factories, then you're wasting resources.

And if you improve factory buildpower, then obviously you would improve efficiency, because your output in relation to factory cost has improved. But you might as well lower costs.

If this is not clear enough then I don't know what will be:

As others have said, this is ok. The extra cost of the engies is accounted for in their flexibility. You're not wasting resources in the extent that it's BP you can use on any project across your base. Of course, building 100 engies to only build land units would be a waste, because you're wasting that flexibility.