Friday Facts #427 - Combat Balancing & Space Age LAN

[Alice3173]

Also a -66% reduction in damage is not that much when you think of legendary quality being 2.5 better than the original. You still get more than 80% of the original damage. That is not even counting the fact that you can have quality armor, which to me WILL benefit the PLD compared to handheld weapons, ( same for spidertron of legendary quality) if you can have 20 -25% more of them, it already makes the PLD "stronger" than in non-space age game. D

It's a lot when you don't assume the use of a mechanic that has been repeatedly stated to be entirely optional and not what the game is supposed to be balanced around.

Also a simple 2/3rds reduction in damage is the single laziest way to address the issue of over-reliance on PLDs. Many many people have suggested many different types of nerfs that would be less lazy and also address the core problem better, though it seems that the devs don't care in the slightest judging by the fact that only a handful of devs have posted in the thread and only to answer questions that weren't directly related.

I really think they just need to up the power draw so that they have similar issues as laser turrets (aka: badly stressing your power grid until you get enough accumulators or nuclear power. The equivalent for power armor being the fusion generator and the second tier of batteries), cut their range in half, and make them not target nests at all. Increased power draw would make stacking them less viable. Decreased range would make them more defensive-focused. Making them not target nests (which it doesn't make sense for them to even do in the first place) means that even if you're trying to rely on them to clear out nests, you still have to add additional methods on top of them no matter what.

Cutting the range and increasing power draw should be no more difficult than cutting the damage by 2/3rds. I'm not sure how difficult stopping them from targeting nests would be but even if it's not simple in a mod, it shouldn't be too difficult for the actual devs of the game.

[shopt]

In theory I'm liking the idea of cutting PLD range instead of DPS. But I'm also conscious that theorising from people who haven't playtested is probably less reliable than the thoughts of people who have playtested the changes as a whole. Realistically we can each mod in whatever we want, but defaults still matter when it comes to multiplayer, speedrunners, and people talking about combat on the internet.

I would address this by giving turrets (and all combat buildings) a 5 second "deploy time" where they can't operate after being built and take additional damage when attacked.

Turret creep is a "problem" that people have tried to suggest solutions to for ages. You aren't first to suggest a vulnerability window after construction, the problem with that solution is that it just means you have to start your turret creep from further out (so that you have some "warmed up" turrets ready to cover the first line that's actually in any danger), and wait longer for each line to warm up before creeping forward. Or you can incorporate quick-walling into your creep. I think we need to accept that turret creep is hard to make non-viable without nerfing turrets so hard they are useless. It can only be made slower and more expensive. The alternatives need to be faster/cheaper than turret creep. But not to the extent of un-nerfed PLDs, which basically turns you into laser turrets with shields on legs.

Turret creep is powerful as it allows you to quickly scale up your DPS in an area, and it tops out very high too. A post-constrution warm up on it's own doesn't lower the top out, or the rate of increase, it just adds a lag. The real problem is that the engineer's construction rate is insanely high. Personally I think if we want to stop offensive turret creep there needs to be a global cooldown between each turret becoming active. If only one turret could become active every 5 seconds that would stop even the most cautious of turret creeps. However that would make the game annoying when using turrets defensively as intended (ie. the cure would be worse than the disease IMO). And it feels very gamey and unrealistic. Overall I'm less annoyed about the viability of turret creep, at least it requires more skill and production than PLD.

when you think of legendary quality being 2.5 better than the original. You still get more than 80% of the original damage

I'm not sure comparing pre-nerf PLD to top-quality post-nerf PLD is useful. Top quality gear is a large production investment, and also supposed to be optional (not to mention impossible for people who don't buy the DLC). We also start going down the rabbit hole of the new buffed portable fusion reactor, and higher quality armor allowing you to equip more PLDs than you could in 1.1.

I think we can just leave it at the fact that PLD was too powerful for when it became available, especially given it needed no ongoing production or skill to use, and invalidated pretty much all other mid to late game combat tools until spidertron or artillery. Yes combat bots did deliver more damage, but they require so much more production, and they are a lot more fragile than a shielded engineer who can eat fish. They also give you far less control over where the acid puddles end up. It may be my technique, but overall I find combat bots hard to use effectively in 1.1. Either they are overkill and you could have killed the nest cheaper, or they die too fast before they have killed much, or you carefully circle strafe the nest they expire before you have killed much.

I'm hoping that post rebalance, standard quality PLD in standard quality armor doesn't carry the combat all on its own like it currently does, but hasn't been nerfed so hard that it only tickles the biters. Which is why a range decrease rather than a DPS decrease is starting to sound better to me, on paper at least.

[Alice3173]

In theory I'm liking the idea of cutting PLD range instead of DPS. But I'm also conscious that theorising from people who haven't playtested is probably less reliable than the thoughts of people who have playtested the changes as a whole. Realistically we can each mod in whatever we want, but defaults still matter when it comes to multiplayer, speedrunners, and people talking about combat on the internet.

Normally I would agree with you but if there's one place in Factorio that Wube doesn't have the best track record, it'd be combat. Combat has always been kinda awful compared to the rest of the game. (And removing the whole alien artifact deal was probably one of their best decisions due to this fact.) They seem to be more keen on easy solutions rather than addressing core issues with the combat, I would assume probably due to it being less fun to deal with than the main focus of the game. They do seem to be trying to make combat more enjoyable with the new planets but those are new things rather than the things already in the game so they're probably more fun for the devs to deal with.

[mmmPI]

It's a lot when you don't assume the use of a mechanic that has been repeatedly stated to be entirely optional and not what the game is supposed to be balanced around.

I am comparing it to the late game solutions that player have at their disposals. I am assuming players can still use better solutions like artillery for hand free long range combat or follower robots for better fighting capabilities. Tesla guns i expect too.

Maybe this was missed from the FFF :

Artillery - damage has been doubled.
Rockets - Accelerate 2x faster

Even tanks that i hate to use have received some delicate attention :

Cannon shells - Increased damage and range.

It's never been said that quality would be unbalanced on purpose either. It would make little sense to release a feature even optionnal that is also known to be broken or easy to exploit, and if that can help address other things at the same time i find that -66% damage reduction in early game when you don't have PLD is not as much as in late game when you access to them, and most likely also quality stuff, things you can't do early game.

In many games when there is a nerf you don't get the option to cancel it, here in Factorio, quality gives you that opportunity, it is indeed optionnal, it allows you if you want to , to invest into a little extra puzzle and have your un-nerfed PLD's to play lasers everywhere.

Also a simple 2/3rds reduction in damage is the single laziest way to address the issue of over-reliance on PLDs.
I really think they just need to up the power draw so that they have similar issues as laser turrets (aka: badly stressing your power grid until you get enough accumulators or nuclear power. The equivalent for power armor being the fusion generator and the second tier of batteries), cut their range in half, and make them not target nests at all. Increased power draw would make stacking them less viable. Decreased range would make them more defensive-focused. Making them not target nests (which it doesn't make sense for them to even do in the first place) means that even if you're trying to rely on them to clear out nests, you still have to add additional methods on top of them no matter what.

I don't think there is much differences in the outcome if the damage is reduced or the power draw increased, to me that's the same thing when trying to math the DPS, it creates the same constraint on the power grid.

Decreased range would overlap with the discharge defense that is the thing for close-combat defense supposedly imo

Making them not target nests would make those player using it with spidertrons to kill bases instead of say artillery cry even more.

I know many solutions have been proposed, but from all i read i prefer the ones from the devs when i try to think of the consequences or read the reasons that lead to their proposal, and i'm sure there will be mods to experiments as there is already with various combat changes.

Overall I'm less annoyed about the viability of turret creep, at least it requires more skill and production than PLD.

I think turret creep is the more methodical approach for players who are more interested in the factory building aspect than in the combat where combat is a matter of logistic. I think it makes sense that it would be the bulkiest / slowest to deploy, but also the most powerfull of all. And those do benefit from target priority, and there is this rocket turret https://factorio.com/blog/post/fff-410 thing which

Shoots any type of rocket (including atomic bomb).

. That probably would require quite some skill to use safely.

I'm not sure comparing pre-nerf PLD to top-quality post-nerf PLD is useful. Top quality gear is a large production investment, and also supposed to be optional (not to mention impossible for people who don't buy the DLC). We also start going down the rabbit hole of the new buffed portable fusion reactor, and higher quality armor allowing you to equip more PLDs than you could in 1.1.

This is about space age though, i have no idea if that will apply to the 1.1, that is not mentionned. I only react to what was announced from the FFFs, i think it is useful to consider the quality thing, and yes higher quality armor allowing you to equip more PLDs or buffed portable fusion reactor allowing you to sustain them are things that comes to mind when considering the reasons for the change in PLDs.

just trying to make it more fun: Personal lasers - We found them to be too strong (especially as you can have multiple), so we nerfed their damage by 66%.

I think side stepping projectile and doing close combat is fun and relaxing sometimes, it also feels rewarding, it's a bit like bullet-hell game, where you can just use another method to beat the game. I think if you want to "break combat" you go turret creep, if you want to "win all fights in a quasi god-mode, you'd better use combat robots than attempts to use PLD"s".

I think it's useful to think of the top quality thing, because PLD's are strong and are late game, it's difficult to top something that is a death ray killing everything in range of your character given enough time. I think it fits as the reward for the "optionnal quality quest". You don't have to go to the rabit hole of the max level quality to have stronger PLD's than before the change to already benefit from the safety of never forgetting any ammunitions when travelling from planets to planets.

I think we can just leave it at the fact that PLD was too powerful for when it became available, especially given it needed no ongoing production or skill to use, and invalidated pretty much all other mid to late game combat tools until spidertron or artillery. Yes combat bots did deliver more damage, but they require so much more production, and they are a lot more fragile than a shielded engineer who can eat fish. They also give you far less control over where the acid puddles end up. It may be my technique, but overall I find combat bots hard to use effectively in 1.1. Either they are overkill and you could have killed the nest cheaper, or they die too fast before they have killed much, or you carefully circle strafe the nest they expire before you have killed

I don't feel this way personnaly but i can understand it was looking like too powerful/ the obvious strongest weapons to many players, even the devs said so, and all the other changes announced on the FFF are buffs to other weapons.

I think combat robots allow you to free up some space in the armor grid for more legs to get in and out of range of worms and neglect the slow from acid puddle more easily, making it easier to fire a weapon like a rocket launcher with nukes in the center of bases and letting the robots killings the approaching biters. To me the most useful combat bot is the basic one made from red ammo that cost nothing and in large numbers, because it uses the same bonus for physical damage than gun turrets and it is somewhat multiplied when you also research follower robot counts. You get it before PLDs or even oil, which is significant in slow game mode like marathon or playstyle. When you have a swarm of them , they effectively act like a PLD, killing everything in range of the character given enough time. Their downside to me is that they can't be used with remote controled spidertron, which are fun to use for combat, but i supose with the RTS tool you can just add more spidertron easily. And by the time you are making armies of spidertrons for your defenses, i think you are already in the "optionnal" territory, like deep in the sandbox/rabbit hole. At the point where thinking about increasing in quality rather than in number becomes an interesting choice.

[sarge945]

Why not, instead of tying it to quality, it's restricted to late-game through research instead.

The proposed idea is something like:

PLD: -66% damage of now, but gains up to 2.5x through quality.

What if it was instead

PLD: -66% damage of now, but gains up to 2.5x through successively more expensive researches.

This would allow the player to "specialise" into personal laser defence through investment, some of which would require late game tech (maybe the final 2 upgrades would require the science pack from the final planet, or, if playing without Space Age, would require white science), without needing to engage with the optional quality mechanic to make PLD viable.

Turret creep is powerful as it allows you to quickly scale up your DPS in an area, and it tops out very high too. A post-constrution warm up on it's own doesn't lower the top out, or the rate of increase, it just adds a lag. The real problem is that the engineer's construction rate is insanely high. Personally I think if we want to stop offensive turret creep there needs to be a global cooldown between each turret becoming active. If only one turret could become active every 5 seconds that would stop even the most cautious of turret creeps. However that would make the game annoying when using turrets defensively as intended (ie. the cure would be worse than the disease IMO). And it feels very gamey and unrealistic. Overall I'm less annoyed about the viability of turret creep, at least it requires more skill and production than PLD.

You uhhh actually make some very good points.

Although, personally, I don't really see the problem with the 5 second global deploy time. To me it seems like it would mostly be a non-issue overall, since in the typical case you would be putting up turret defenses around your base. Unless you're literally being attacked at that exact moment, there should be plenty of time for them to come online before the next attack wave. Although, if you're putting up 100 turrets, I can see that being a major concern, because the last one won't be online for around 5 minutes(!!!)

But overall, yeah, stopping creep through a deploy timer is probably a bad idea.

I agree that the real issue is that the engineer just builds too quickly overall, but there's probably not much that can be done about it.

I'm not suggesting this as a real idea, but one thing that might be interesting (probably as a mod, it definitely doesn't belong in the base game) would be a way to encourage turret survival. What I was thinking was, turrets could start out extremely weak, but as they gain more kills, they calibrate their targeting systems, allowing them to target enemy weak points more effectively, which essentially gives them a significant damage boost, up to their current level. This is similar to veterancy in other strategy games. If they are deconstructed or destroyed, the calibration data is lost. This would effectively end turret creep, but would be such a major gameplay change I could see it introducing other problems. Still, it would definitely be INTERESTING (not necessarily GOOD, though) to have some sort of incentive to keep turrets in tip-top shape, since especially in the early game, it is currently so easy to just plonk them down as an easy defensive line without really considering their longevity at all.

[GregoriusT]

But overall, yeah, stopping creep through a deploy timer is probably a bad idea.

There is a Mod that adds a Deploy Timer, it is not bad. However what you could really do is have the Deploy Timer ONLY happen for buildings when you are close to an Enemy.

Do note this would be much less fun if it was forced into vanilla though, there is a few fun things that a deploy timer would ruin.

[sarge945]

Do note this would be much less fun if it was forced into vanilla though, there is a few fun things that a deploy timer would ruin.

Why would anyone want to have fun in a videogame?

I say that in jest but there's a certain truth to it, sometimes you have to put aside something that's fun on the surface for the sake of balance, to increase the overall depth and fun of the resulting game. It's just like with the PLD. It can be fun steamrolling biters with them, but at some point we have to step back and say "no, that's too powerful", otherwise the rest of the game suffers.

Potentially all of this could be fixed easily if turrets just required power. You could still turret creep but it would take a lot longer to pull off and your power lines would leave you somewhat exposed.

[GregoriusT]

Potentially all of this could be fixed easily if turrets just required power. You could still turret creep but it would take a lot longer to pull off and your power lines would leave you somewhat exposed.

That would take away the abilities of Gun Turrets to work as a Backup in case of a Power Outage caused by Biters eating a Power Pole.

This would also incentivize using Lasers even earlier because the Effort of supplying Power to the Gun Turrets was already taken, so why not just use Lasers, which dont need Ammo. (once you got the Lasers)

The unique Ability of being able to just go to some Cliff Gap or commonly taken Path (especially with the terrain-generation FFF), and put a Gun Turret there to prevent Biter Expansions into an Area, would also be taken away due to the Power Requirement. Sure you could do the simple Solar Setup for this, but Batteries are far too late in the Game for that to not ruin the Gun Turret.

[XT-248]

We do not know how quality works with personal laser defense modules (increase PLD's range? increase PLD's shooting speed? increase PLD base damage? Or some other benefits? Only one stat? Or all?). Unless I missed a dev comment regarding the quality effect on PLDs, feel free to correct me with a direct quotation.

I have also seen various claims from random people on the internet on how effective the post-nerf PLD would be relatively after the release of the 2.0 patch. Some would claim that Q5 PLD DPS will be slightly under the pre-2.0 PLD DPS baseline value. Some would claim that the new Q5 PLD DPS will be above the pre-2.0 PLD DPS baseline value.

I would not put too much stock in any combat-related numbers before October 21, 2024.


I would be okay with a modest decrease to the PLD range, which is already really short. There are already six enemies (most are worms, not counting new creatures coming with Space Age) enemies that can match or outrange PLD.

[sarge945]

That would take away the abilities of Gun Turrets to work as a Backup in case of a Power Outage caused by Biters eating a Power Pole.

Honestly, this might be a welcome change. One of the big problems of Factorio's base defense in my opinion is that once you get to a point where you've got your defences in place, you're basically invulnerable. Having more ways to become vulnerable through mismanagement of resources is welcome. Of course, flame turrets would also need power in order to make this remotely fair for the poor-little early game turret.

One feature I am divided on from other defensive games (mostly tower defense games) is that they will often have random events where you lose power due to a "solar flare" or whatever. Riftbreaker does this, among many others. While I like the idea in concept, because it forces you to be more adaptable regarding your defensive setup, it can also randomly screw you over for no fault of your own and it's never fun to be randomly punished at exactly the wrong time. If I was going to take inspiration from these systems, I would try to find a way to make it more dynamic and controllable by the player, like a looming threat of a solar storm with a countdown that you can account for and plan around. Again, these are not serious suggestions that I think should be actually added to the game, as they would be extremely divisive and likely have knock on effects, but they could be fun ideas for mods, especially for players who have already mastered Factorio's combat and who are at the point where their base is practically indestructible.

This would also incentivize using Lasers even earlier because the Effort of supplying Power to the Gun Turrets was already taken, so why not just use Lasers, which dont need Ammo. (once you got the Lasers)

Maybe not. The power requirements for lasers are MUCH higher, and they tend to drain power at an inconsistent rate, meaning power outages are far more likely. Gun Turrets having a constant (but small) power would still be significantly easier to manage overall, and with the different ammo types available, retain their utility even late into the game.

The unique Ability of being able to just go to some Cliff Gap or commonly taken Path (especially with the terrain-generation FFF), and put a Gun Turret there to prevent Biter Expansions into an Area, would also be taken away due to the Power Requirement. Sure you could do the simple Solar Setup for this, but Batteries are far too late in the Game for that to not ruin the Gun Turret.

You'd still be able to do this by building a long line of power poles, which is admittedly kind of annoying, but definitely workable.

I fully understand people don't really want this in the base game, which makes sense, but I will definitely be creating a mod for this as I believe the potential downsides are absolutely worth it for making defensive play more interesting overall and preventing "cheesy" turret play.

[mmmPI]

Why not, instead of tying it to quality, it's restricted to late-game through research instead.
The proposed idea is something like:
PLD: -66% damage of now, but gains up to 2.5x through quality.
What if it was instead
PLD: -66% damage of now, but gains up to 2.5x through successively more expensive researches.

There are already 2 researchs to increase the damage output of the PLD's in the game. I think it's more interesting to have another optionnal mechanic rather than another research wich would quite overlap with the existing laser damage and laser shooting speed.

[sarge945]

There are already 2 researchs to increase the damage output of the PLD's in the game. I think it's more interesting to have another optionnal mechanic rather than another research wich would quite overlap with the existing laser damage and laser shooting speed.

Right, I forgot about those.

But if it's a problem, why not just add new levels to the existing techs and tweak the numbers accordingly instead?

[XT-248]

Right, I forgot about those.

But if it's a problem, why not just add new levels to the existing techs and tweak the numbers accordingly instead?

Laser damage is infinite repeatable technology—no need for more levels. Although the exponential growth in technology costs could use some tweaking as the modifier scales off base laser damage.

The number of Laser shooting speed tech is limited. Otherwise, the DPS would be absurdly high after attaining-max-quality/multiple-laser-damage-repeatable-levels/etc.

[mmmPI]

Right, I forgot about those.

But if it's a problem, why not just add new levels to the existing techs and tweak the numbers accordingly instead?

That will end up almost the same as what the devs did, if you play without quality, you have the same PLD as before except you need more research for the same damage.

Also i would like to correct a misunderstanding, more shooting laser speed doesn't allow for higher damage output if the fight is prolonged, like clearing a nest, because 1 shot of laser cost the same energy, for the same damage, and what limits you is the power of your armor, shooting faster but not having more power generation is useless. If you were to have absurdly high level of shooting speed you would also have absurdly short period of shooting.

This is not the case with things that require ammunition, like the submachine gun. Since their capacity of shooting is limited by the stock of ammo, which is not necessarily limited like the equipment grid of an armor, and as such is quite different regarding balacing.

Laser shooting speed seem much more useful for laser turrets or destroyer capsule than PLDs.