Laser Turrets need upkeep

[Rahjital]

One reason laser turrets do so much damage is because there are so many rate of fire and damage upgrades.

This applies to every single weapon in the game, though (well, with the exception of the flamethrower). Gun turrets benefit from the bullet damage research, so in total, guns get 240% bonus from research, lasers get 280% bonus; not that great difference.

The thing that makes laser turrets clearly better is density. Gun turrets are 2 tiles in size while laser turrets only take 1 tile, meaning you can build 2 lasers for each gun; this gets even worse when you try to rearm the turrets automatically because the chests and inserters (and belts if you don't use logistic networks) take additional space that could otherwise be used by more turrets. This prevents massing gun turrets and massively hurts their performance when compared to lasers. Gun turrets are balanced without much regard for this density problem, they have 100 hp per tile while laser turrets have 250 hp per tile, meaning they also get destroyed much quicker should the biters get through the walls.

I think these problems could be easily be solved with just a few changes:

- Gun turrets should be made stronger or laser turrets weaker so that a single gun turret is at least equivalent to two laser turrets
- Same with health, gun turrets should have at least twice as much health as laser turrets to make them equal
- Gun turrets should be able to request ammo from logistic networks

[Marconos]

how many people have tried using the Gun Turrets late game? At max research Gun turrets eat through the biters. The only reason to use lasers over them is the upkeep cost. I can defend areas easily with only gun turrets and no lasers. The issue becomes the amount of copper / iron I need for the ammo to feed these things. That is a good thing IMO, we just need something on lasers to tone them down. If they don't get upkeep then increase their size, decrease their damage / range and make it that 4(or more) lasers == the power of a single gun turret. Of course balance this around upgrade levels.

I have been watching quite a few people doing lets plays and they all dive to lasers and complain about gun turret damage. What is sad is they are all using 1st level ammo with virtually no upgrades. Of course that doesn't work well. Lasers are just too good for the cost IMO.

[Kazuar]

[...]guns get 240% bonus from research[...]

Guns get 384% bonus from research - the bullet damage ugrade raises the ammo from 5 to 11(+120%), the turret damage upgrade then raises the damage from 11 to 24.2 (+120%) (2.2 * 2.2 = 4.84). In addition, even a fully upgraded laser turret has half the firing rate, compared to a gun turret. And, lets not forget that, in the factorio universe, bullets travel instanteously, where as amplified light travels slowly

Applied to the enemies themselves, a gun turret compares to a laser turret like this

• - Small biter's killed in one shot (vs. lasers 2 shots), using 0.5 iron and 0.5 copper in ammo, the very moment it enters range (vs. 0.2s* + projectile travel time)
  - Medium Biter's killed in four shots (vs. lasers 7 shots), using 2 iron and 2 copper in ammo, after 0.3s* (vs. 1.2s* + projectile travel time)
  - Big biter's killed in 24 shots (vs. lasers 35 shots), using 12 iron and 12 copper in ammo, after 2.3 s* (vs. 6.8s* + projetile travel time)

The problem with massed lasers is that projectile travel time is more or less constant: even if you have, for example, 35 laser turrets all firing at once at a big biter, the laser still has to travel; even though his death is instantly guaranteed, he still has time to bite things. If you have 24 gun turrets all firing at once on a big biter, he dies instantly. Yeah, if you take the density advantage into consideration, you can place three or four times as many lasers than gun-turrets into the area the latter take for themselves and their inserters/belts - but you also need three or four times as many lasers just to match the gun-turrets killing speed.

I'd actually love to know how a lasers energy consumption works. According to game file, it's 200kJ per shot, but the game files also say "damage = 2" (which is not what we end up with!), so.... I don't understand their entity-entry, at all.

* = the first shot is fired @ 0.0 seconds, second @ 0.1s(laser 0.2s), third @ 0.2s(laser 0.4s) and so on.


Thanks to Marconos for causing me to reconsider gun turrets. I did not want to believe you, Marconos, I was preparing to prove you wrong in another thread - and when I actually did the math, guns turned out to be even more powerful than you've said (big biters ignore only 8 points of bullet damage, not 50% of bullet damage), and I've turned out to be a complete tool. I even went so far to assume "this must be a display bug; there's no way these upgrades stack", then I made a test scenario in the map editor - and the displayed damage (5+19.2) is indeed the damage inflicted.

[Rahjital]

I just checked too and whoa, you are indeed right, the turret gun damage is applied sequentially, which makes for a rather huge difference (124% !). Together with the way biter resistance works, it means gun turrets are good early on, rather useless at mid game, and then absolutely kick ass late game. That's some strange balance... I'm glad to have been proven wrong here, though, I'm not very fond of laser turrets. The pesky slow lasers you mentioned mean that biters almost always manage to damage the walls, even despite the turrets themselves having longer range than gun turrets.

I still believe that all factors considered, lasers are the superior kind of base defense. Unless there's a design I haven't seen yet, you can place 7 laser turrets for each 2 gun turrets, which levels the playing field, the lasers have total health of 1750 compared to 400 of gun turrets (630 if you count the chest and inserters too), they don't require any supply lines to be built (if you don't mind manually refilling a roboport with repair packs once in a blue moon) and most importantly, they are more powerful during mid-game when it actually matters because when fully upgraded, both kinds cut through biters like through butter when massed.

[Schmendrick]

I'd actually love to know how a lasers energy consumption works. According to game file, it's 200kJ per shot, but the game files also say "damage = 2" (which is not what we end up with!), so.... I don't understand their entity-entry, at all.

Damage = 2? Where do you read that?

The only display inconsistency I'm aware of is that laser turret energy consumption isn't updated to reflect firing speed increases.

[Marconos]

• - Small biter's killed in one shot (vs. lasers 2 shots), using 0.5 iron and 0.5 copper in ammo, the very moment it enters range (vs. 0.2s* + projectile travel time)
  - Medium Biter's killed in four shots (vs. lasers 7 shots), using 2 iron and 2 copper in ammo, after 0.3s* (vs. 1.2s* + projectile travel time)
  - Big biter's killed in 24 shots (vs. lasers 35 shots), using 12 iron and 12 copper in ammo, after 2.3 s* (vs. 6.8s* + projetile travel time)

Right here is the problem. Yes, they do more damage and kill enemies quicker. BUT they cost ALOT more to setup and supply and when you have massive waves coming at you the drain on iron / copper / power becomes quite large. Smelting ore to copper / iron then iron to steel. Not to mention the consumption of the copper / iron. With that massive drain on power and resources lasers become the "easy" option. Throw down some solar panels, couple of laser and bam you have a defense that maybe is a little slower but is SOOOO cheap why even look at the other option.

This again takes me back to my original point ... lasers need some type of upkeep to balance them out.

Glad I got people thinking and actually looking at numbers. So many just read guns r bad lasers be good and never try the other route. My last play through (that I failed on for being a moron and running 100% out of coal), I was using 100% coal power, 80% gun turrets and no solar power (accumulators were okay with in reason though imbalanced as well). What I noticed in this game is as it got later and later the rate that I was burning iron / copper to keep the guns running was getting pretty high. It was great having these massive iron / copper veins and running out of resources. (Just wish I had paid closer attention to my coal usage ... 100 steam engines eat a bit of coal).

[jakobeng1303]

laser turrets should not be nerfed but the could be stationary rocket launschers oder much better gun towers

[Rahjital]

[quote="Schmendrick"]Damage = 2? Where do you read that?

It says so in attack_parameters table of the laser turret (in turrets.lua). However, the laser in projectiles.lua has its damage set to 5, so it seems the value of 2 is either not used or it affects the turrets parametres in a rather cryptic way.

Laser turrets also have 47% longer range than gun turrets (25 tiles vs the 17 tiles of gun turrets) which pretty much erases the damage advantage of gun turrets, as it means you can have up to 22 rows (!) of turrets firing even if you use 2 tiles thick walls and an additional tile of free space to prevent medium and big biters from reaching the turrets directly.

[Kazuar]

[...]Laser turrets also have 47% longer range than gun turrets (25 tiles vs the 17 tiles of gun turrets) which pretty much erases the damage advantage of gun turrets, as it means you can have up to 22 rows (!) of turrets firing even if you use 2 tiles thick walls and an additional tile of free space to prevent medium and big biters from reaching the turrets directly.

Now, this is a rather hypothetical value - I'd presume this many lasers would actually make the defence weaker than it'd be with less lasers*. The reason being, the last, let's say 3 rows, would only ever fire if a biter came very, very close. If one of the last lasers in the line fire at him, no one else will, since he's already "marked for death", yet due to travel time, he'll still be able to connect two or three attacks** - it stands to reason that, if one of the other (closer) turrets had taken the shot, he'd be dealt with more swiftly.
In any case, I think this part of the discussion be best "solved" with a testing scenario from the map editor - a small-ish fortress, lasers on one side, turrets on the other, and place spawners in front of each, until one of the sides allows for damage to be received.


To return to topic:

[...]lasers need some type of upkeep to balance them out.[...]

I agree with the general notion, but am unsure if this should be applied directly to the lasers, themselves. My point being, lasers do consume energy, and if Schmendrick's right, it probably remains at 200kJ per shot, regardless of firing speed. That would be a whole MW per turret firing, which is not a trivial amount. Applied to enemies, this would mean:

• - Small biters needed 2 shots = 0.4MJ, or 0.1 coal
  - Medium biters needed 7 shots = 1.4MJ, 0.35 coal
  - Big biters needed 35 shots = 7MJ, 1.75 coal

This is already some amount of upkeep (not saying it's enough), and frankly, the discrepancy in upkeep cost could also be "solved" if, for example, bullet magazines where buffed from 10 to 30 rounds each (imho, it's just bullets who are overpriced, even for use by smg - they're (arguably) more expensive than piercing shells!), putting the two turrets closer together in terms of upkeep cost. There's also the idle drain of lasers - for each 85 lasers, you need one steam engine running all the time at maximum capacity, in turn consuming 459 coal/hour just to pay for doing nothing, and really, 85 turrets is not that many (had 1.1k in my last game), but it is another way through which one can balance them out.

Of course, none of this matters once you power them through solar - it is the one-time investment of 40iron/27.5 copper that gives you 37.8coal/h in worth***, forever and for free, which imho is unbalanced, and where (again, imho) either an upkeep cost should be applied, or a cost increase be in order.



* = Of course, this depends on what metric you use to compare defences - I prefer 'averagely received wall damage/minute'.
** = every big biter "chomp" on the walls costs you 0.9 iron/0.45 copper in repair, if I recall all numbers correctly.
*** = This doesn't seem huge, but you generally need more than one panel - if you assume 500 panels (for a daily average of a meagre 21MW), you'd be at 20k iron/ 13.75k copper, but you "earn" 18.9k coal per hour, for ever.

[Rahjital]

Now, this is a rather hypothetical value - I'd presume this many lasers would actually make the defence weaker than it'd be with less lasers*.

So is the majority of other numbers in this thread. There are many factors that haven't even be considered so far, such as the time spent unfolding and targeting new enemies (the values for gun turret are higher, but I have no idea whether it means gun turrets rotate faster or that they takes more time for the same rotation), biter pathfinding (a nice forest can stall a large biter attack considerably), the shape of the attacking biter swarm itself, etc.

As for the problem with slow projectile travel from the far turret rows, I think that would only happen in a situation when a lone biter manages to get next to the wall without getting shot at. In an actual attack, the biter would either get vapourized long before reaching the range of the last rows or it would come with a lot of other biters, in which case the back turret shooting would allow the front rows to attack a new target long before it dies, which makes up for it partially.

Indeed, a test scenario would be the best for this, although I would propose a more rigid test than letting the game's randomness affect the results. I think a custom map where enemy waves are spawned by a script would be much better for testing purposes.

I agree with the general notion, but am unsure if this should be applied directly to the lasers, themselves. [...] either an upkeep cost should be applied, or a cost increase be in order.

I completely agree. Idle laser turrets can drain megawatts of idle power and more when they are shooting, but solar panels make this completely irrelevant. That's a problem that applies to everything powered by solar panels, not just laser turrets, so it should be fixed at the source. Just raising the cost of solar panels won't help with this as once built, the panels will still produce infinite energy for free, with no pollution and no need for logistics. I think the path of diminishing returns would be the best here: solar panels need accumulators to get through the night, so if accumulators wasted x% of the energy they stored, each additional accumulator would mean more energy lost. That in turn would mean one can't just build massive arrays of solar panels and accumulators and be set forever.

I also agree about gun turret ammo. 10 shots in a magazine is ridiculously low amount, a turret uses 1 magazine per second and a fully upgraded SMG chews through 4 and half magazines per second! That's crazy. Shotgun shells can last much longer as they have the same amount of shots but deal much more damage per shot. Raising the number of shots in bullet magazines to 30 sounds like a very good idea.

[Robbedem]

I think the unlimited power from solar panels and accumulators has a simple solution:
- accumulators need power to run when they are used OR they don't return the full amount of energy back (perhaps 75%)
- the addition of weather would influence the solar power production. One of the main disadvantages IRL for solar power is the variation in production, but in Factorio it's always exactly the same cycle.

[ssilk]

The first one was mentioned in some forms:

https://forums.factorio.com/forum/vie ... urve#p9894
https://forums.factorio.com/forum/vie ... rve#p15051
https://forums.factorio.com/forum/vie ... rve#p11674

I think it is clear, that taking the loading curve of an accu into account then that rises it's efficiency?
In other words: loading/ unloading with full power heats the accu up. A hot accu is not so efficient, than a cold.

The question is, what diodes it bring for the gameplay? Very questionable...

[Rahjital]

The question is, what diodes it bring for the gameplay? Very questionable...

Not being able to run huge factories on free infinite solar power alone. I think that's a good enough reason to at least consider it

It doesn't have to be a loading curve itself, it could be a simple efficiency mechanism like the rest of the game already uses. If efficiency is 0.8, the accumulator store 80 watts of every 100 wats you give it. Simple to implement, transparent for player.

[Kazuar]

[...]I think the path of diminishing returns would be the best here[...]
[...]It doesn't have to be a loading curve itself, it could be a simple efficiency mechanism like the rest of the game already uses. If efficiency is 0.8, the accumulator store 80 watts of every 100 wats you give it. Simple to implement, transparent for player.[...]

Boilers have an efficiency of 0.5, but there is no indication of this in-game - the game even misleads slightly by calling their thermal output "burner consumption".
Also, putting accumulator efficiency to 0.8 would not lead to diminishing returns, by itself - they would consume more power then they store as energy, and as a result, assuming equal power surplus in both cases (efficiency = 1.0 vs. efficiency = 0.8), the most prominent effect would either be an 25% increase in charging time, or a 25% 'higher peak' during charge-up (both by extension of a 25% increase in energy required for a full charge).

For purposes of a "solar only" power supply strategy, this would be a flat increase in the minimum amount of 'basic accumulators' needed for each X solar panels, making it a cost increase, basically - reducing a panels peak power production from 60kW to 48kW would have a similiar effect, but without simultaneously making accumulators less attractive for boiler-oriented power supply strategies.


I could imagine something like a degradation of panels work*, where their power output is modified by their health (e.g. health/maxHealth), and once every noon, an arbitrary amount of damage is applied to solar panels - effectivly making solar powers run on repair packs**.
Now, not every kind of upkeep has to be the 'straight, recurring cost' type of thing - making them cost safety, for example, would be another possible vector of balance***.



* = From a gameplay standpoint, not so much from a realism standpoint
** = We went full circle back to the OP, here
*** = Yes, I just proposed solar panels to generate pollution - no one ever said in-game pollution equals co2-emission or somthing like that; there are a lot of other kinds of pollution imaginable (sound, light, concussion, heat etc.).

[syneris]

We just need a larger variety of enemies with different resistances. I don't think anything has any armor or resistance to laser damage. While paper, rock, scissors damage schemas are fairly common, it is still one of the best ways to balance out the superiority of certain weapons.

Strengths and weakness are in the game, but not really taken advantage of yet. Once this changes, all the different weapons will find their place. A biter strong against energy, but weak to bullets would certainly change laser towers being the all out best. (And I still think loading combat bots in towers to help fight airborne enemies that try to avoid stationary defenses would be awesome)

[Schmendrick]

Boilers have an efficiency of 0.5, but there is no indication of this in-game - the game even misleads slightly by calling their thermal output "burner consumption".

boilerefffectivity.jpg (42.74 KiB) Viewed 12400 times

While I agree it is unclear whether the "effectivity" occurs before or after the 390kW, overall the functionality is not unclear, and it's 0.5 is quite explicit, at least under the crafting menu and on the item (in inventory) itself. The only place it isn't displayed is on the actual built structure.

[Kazuar]

*snip*

Ah. Somehow, my memory only worked for the structure itself. My bad!

[Rahjital]

Also, putting accumulator efficiency to 0.8 would not lead to diminishing returns, by itself - they would consume more power then they store as energy, and as a result, assuming equal power surplus in both cases

And that's why you should always pay full attention when commenting I should have read thought twice before I said that. Thank you for calling me out on it! What I was thinking of at that moment was a system of diminishing returns where each additional accumulator would multiply the amount of energy stored in all accumulators on the grid by a certain "efficiency" value (if that value was 0.8, the first would limit the amount of energy stored to 80%, the second to 64%, the third to 51.2%, etc). However, this solution scales rather badly because it penalizes the first few accumulators much more than the rest of them, which would make them rather useless.

An actual solution to the problem could be having all accumulators leak a small percentage of their maximum charge whenever there is any energy stored in them. That would mean an additional penalty for each accumulator built, and unlike the previous example, it scales linearly. It's also much more transparent for players.

without [...] making accumulators less attractive for boiler-oriented power supply strategies

In my opinion, that would be actually desirable. I feel that accumulators should be there only as emergency power supply to prevent your laser turrets/rocket defense/other energy hungry things from stalling your power grid, but running a full featured factory on accumulators alone is a bit ridiculous. By themselves, solar panels aren't as powerful, as they take a lot of land and provide little power at night; if they were more expensive and perhaps provided slightly less power, it could easily offset the fact they provide free energy without pollution. Only when combined with accumulators they become a game-breaker, which partially applies even with the steam-accumulator combo - it does not allow to power everything for free, but it still makes shutdowns of power generation a complete non-issue.

[Kazuar]

[...] What I was thinking of at that moment was a system of diminishing returns where each additional accumulator would multiply the amount of energy stored in all accumulators on the grid by a certain "efficiency" value (if that value was 0.8, the first would limit the amount of energy stored to 80%, the second to 64%, the third to 51.2%, etc).[...]

I do kinda like the idea of this, so I've taken some free time to make a spreadsheet to see what value would feel sensible to myself (that, and I was bored at work). It turned out, that this would introduce a "ceiling" for the amount of accumulators one can build, after which every additional accumulator actually reduces the sum total of your capacity.

Warning - Here be numbers

Taking an efficiency of 80% as example, you'd have

• - 1 accumulator @ 80%: (1*5MJ*0.8) = 4MJ
  - 2 accus @ 64%: (2*5MJ*0.64) = 6.4MJ
  - 3 accus @ 51.2%: (3*5MJ*0.512) = 7.68MJ
  - 4 accus @ 40.96%: (4*5MJ*0.4096) = 8.192MJ
  - 5 accus @ 32.77%: (5*5MJ*0.3277) = 8.192MJ(again!)
  - 6 accus @ 26.21%: (6*5MJ*0.2621) = 7.864MJ(oops!)
  - 7 accus @ 20.97%: (7*5MJ*0.2097) = 7.340MJ(and it keeps dropping from here)

At about 50 accumulators, your total capacity has dropped to about 4 kJ - enough to power an idle laser for 2/3 of a second
This ceiling stays around, even at much higher efficiencies - taking 99% as start, you hit this ceiling somewhere between 100 and 125 accus, taking 99.5% as start, you hit the ceiling at 200-225 accus; even at 99.8%, you're stopped at around 500-600 accus.

No matter what base you'd pick for diminishing returns like this, that ceiling seems to stick around as a bad sideeffect, unfortunately. Don't know how to fix that, either.

[...]An actual solution to the problem could be having all accumulators leak a small percentage of their maximum charge whenever there is any energy stored in them. That would mean an additional penalty for each accumulator built, and unlike the previous example, it scales linearly. It's also much more transparent for players.[...]

This has some actually interesting side-effects! I played around with numbers, again, and found two interesting results:

• - It will create an hard limit on the duration your accus will last - if they would e.g. lose 1% of their charge each second, you could never have more than 100 seconds of your surplus stored as energy (solar-only takes 208,33 secs to charge/discharge, respectively). As a sideeffect, your accus would only ever be useful for sudden spikes in consumption - a slow erosion of your surplus (e.g. by expanding your factory) would cause accu charge to drop to reflect your new, lower surplus, until it comes to rest at about 100secs of worth. I did not do the math if sunrise/nightfall would count as slow or sudden, for this purpose.
  - As a result of the first point, using accus would either cause your power generation to be always at maximum, or your accus to consume 50kW each (if your surplus*100s is larger than your capacity).

If you'd take 0.833% as waste, your limit would be at 120 seconds (1/120 = 0.00833_)

In either case, the behavior of accus would change drastically, even for boiler power. I'm not sure whether this is good or bad, or if accus should be changed like this (just to throw this around: a single storage tank could hold more than 30,000 MJ, or 30GJ, of electric energy, assuming light oil to solid fuel processing - if making a stockpile for bad days is your intent, there are more space efficient ways than accus) - but it is an interesting behaviour.

[...]I feel that accumulators should be there only as emergency power supply to prevent your laser turrets/rocket defense/other energy hungry things from stalling your power grid, but running a full featured factory on accumulators alone is a bit ridiculous. By themselves, solar panels aren't as powerful, as they take a lot of land and provide little power at night; if they were more expensive and perhaps provided slightly less power, it could easily offset the fact they provide free energy without pollution. Only when combined with accumulators they become a game-breaker, which partially applies even with the steam-accumulator combo - it does not allow to power everything for free, but it still makes shutdowns of power generation a complete non-issue.

This is kinda the purpose of a battery, though - think how big the accumulator of a notebook is; the notebook can be run completly on battery power.
That being said, I do feel solar panels could stand to cost Advanced Circuits for production, though maybe not 15 each. Doing so would make their assembly dependant on oil (=plastic), which already limits the availability of a lot of other things (tech progression, armor equipment, modules, lasers, accus, logistics networks etc.). I've often found my oil reserves to be the harshest limiter to my progression, and I feel solar panels should be a part of this limit*.

* = The question of whether oil is abundant enough or not, is a completly different balancing question.

[Rahjital]

This has some actually interesting side-effects! I played around with numbers, again, and found two interesting results:

- It will create an hard limit on the duration your accus will last - if they would e.g. lose 1% of their charge each second, you could never have more than 100 seconds of your surplus stored as energy (solar-only takes 208,33 secs to charge/discharge, respectively). As a sideeffect, your accus would only ever be useful for sudden spikes in consumption - a slow erosion of your surplus (e.g. by expanding your factory) would cause accu charge to drop to reflect your new, lower surplus, until it comes to rest at about 100secs of worth. I did not do the math if sunrise/nightfall would count as slow or sudden, for this purpose.
- As a result of the first point, using accus would either cause your power generation to be always at maximum, or your accus to consume 50kW each (if your surplus*100s is larger than your capacity).

The first point was pretty much the entire reason for the suggestion - I am a firm believer that accus should be there for situations where your power generation spikes. For more gradual changes of consumption, expanding power generation has always been the better way of dealing with it.

As for the second point, I don't quite view it as a disadvantage. If you build so many accumulators that your power network can't handle them then you probably need to improve the network already. Of course, the number doesn't have to stay at 1% of capacity, it could be a lot lower, even below the treshold for preventing solar panels from charging the accus for night - as long as the energy losses are significant enough to mean you need a huge number of solars and accus to power a factory through night, I'm fine with it, especially if the costs for both increase. It would make getting a solar powered factory an achievement instead of the most efficient option.

This is kinda the purpose of a battery, though - think how big the accumulator of a notebook is; the notebook can be run completly on battery power.
That being said, I do feel solar panels could stand to cost Advanced Circuits for production, though maybe not 15 each. Doing so would make their assembly dependant on oil (=plastic), which already limits the availability of a lot of other things (tech progression, armor equipment, modules, lasers, accus, logistics networks etc.). I've often found my oil reserves to be the harshest limiter to my progression, and I feel solar panels should be a part of this limit*.

The power requirements of a notebook versus an electric smelter are a different beast, though. As far as I'm aware, we don't have a battery that can provide sufficient voltage and current to smelt alumina.

As for solar panels, I kinda like how they help reduce your coal consumption by providing power during the day, though I agree that they could use a cost increase and perhaps a slight reduction of their power output. Personally, I would add a cost increase to accumulators too, as right now they are very cheap themselves.