Suggestions to improve biter experience for both new and experienced players

[DragonMudd]

TL;DR

The biter defense experience as it currently stands rewards working harder (building a barrier around your entire factory), but does not reward working smarter (placing defenses efficiently and strategically).

Preface

(Feel free to skip to suggestions)

First, I want to say I've been playing for a long time (2000+ hours, and been playing since version 0.14) and I love all of the changes that have been made to biters and combat in all of that time. Recently, there was an interesting FF that raised some ideas about biters that has given me a lot to think about and focus my thoughts. In particular, "Currently, you can get by with only walling off part of your base, responding to attacks as they come, and leaving the rest fairly open. This runs opposite to what "most new players" [citation needed] would expect."

I definitely feel that 1) that's only true at the very early stages of the game and 2) when I first started playing, I definitely expected that biters would be making coordinated attacks to the parts that upset them the most and that if I were smart enough, I wouldn't need to put walls and turrets around my entire base. The second point led to me quitting my first factory attempts and restarting with pacifist biters, because I couldn't understand biter logic once they started to get more aggressive.

Now, I get that's just a single anecdote, but regardless of what "most" new players experience it led me to thinking about the biter experience, and the game experience as a whole. The aspects that I think most of us enjoy are that there are puzzles to be solved, and that there is no single way to solve them. Often times, they break down into two categories: make more things, or make things more efficient. Which is where I get the concept of working harder (more things) or working smarter (more efficient).

I would love to see more opportunities for biter combat to have the choice of working smarter instead of working harder, and the below suggestions all attempt to improve that. Some of these suggestions might be simple to implement, and some might not be. But, at the very least I don't think any of them take away from those who do want to always have the option to simply wall off their entire base.

As examples of opportunities for "work harder" that have been implemented that I love: the change in water generation that creates outstanding defensive choke points (it didn't used to be that way) and the ability to make long continuous cliffs, again for the creation of great choke points.

What ?

1. Biter Early Detection System - I'm thinking a researchable upgrade to Radar that you would toggle on/off for each Radar dish, but whatever it is, some system that allows a warning (sound?) when biters enter in range of the factory.
2. Show Biter Paths in a user friendly way - I always play with the Show Paths debug on, entirely so that if biters did attack somewhere I can figure out where they came from. I would love to see something done where a player can see what path biters took to get to where they got. One idea: biters leave behind a visible "pheromone trail" that dissipates over time when they walk. Or just something similar to the current Show Paths debug, but perhaps with a little less noise.
3. Biters shouldn't be able to path outside of the pollution cloud to get to the factory.
4. Biter target selection should first pick direction pollution is coming from, then once they get near something that makes them angry they should attack that - As long as we're talking about pathing... I suspect this might be the hard one. Make it straightforward which things my biters will prioritize and which they'll avoid. This might be a very difficult and/or fundamental to change, and since it's not clear how it currently works, it's hard to make suggestions, but this seems a) intuitive and b) less likely to have weird things happen.

Why ?

1. Currently, unless you happen to be watching your map (usually we're building our factory) you don't learn about biters who are entering your factory until after they're already destroying things. Then, by the time you get over there, so much has been destroyed that 9 times out of 10 we all just reload our last autosave. It's not until late game where the loss of time and resources is minimal enough to let it go. Adding a feature like this will give players a chance to be able to react to the biters and attempt to race them to wherever they are and actually engage in combat attempting to defend your precious factory. Bonus: with some sort of signal there might be very creative circuit conditions players could come up with to automate defenses in fun and interesting ways as well.
2. One of the most common stumbling points I see from new players commenting on reddit, and what I find frustrating for myself, is figuring out where the biters are coming from. Especially if, say, the biters took a bizarre path around a large sea or something. Similarly, when they do sneak through, it really helps to try to figure out what hole I might have in my defenses.
3. As above, I see new players confused where the biters are coming from when they began on the opposite side of a lake and took some lengthy, circuitous path way out of the pollution cloud and then back in in order to attack.
4. From what it appears to me, the biters choose some object in my factory and then make a path to that. But it's not clear what objects they may or may not choose, and weird stuff happens often. Here's an example that I want to see NOT happen. A couple days ago I was playing and some biters got into my base (I intentionally tore down my complete walls because it was too boring and I wanted more excitement). They were in a weird spot in the middle of my base attacking some solar panels. After I killed them, I looked at the path they took to get there: they had originally targeted some assemblers and chose a path there. This path took them past some rails and Big electric poles, past some Roboports, past more rails, in between some miners and the electric furnaces they were directly feeding to (about a 50 tile wide gap that had lots of belts and balancers), past a Radar, past a bunch of solar arrays, past more Roboports, and rails, along a long narrow bridge I built for myself to cross some water without getting hit by a train, past another set of miners and another Radar, and then at this point they gave up their original goal and starting attacking some other solar array of mine. That behavior really doesn't make sense to me. I definitely would've expected them to attack any of the things along the way that they walked past. This example might be a somewhat extreme example, but it definitely isn't a unique one.
   My suggestion should clear all of that up. Once they decide to launch an attack party, they first determine what direction the pollution is coming from (exercise left to the reader), and then the start marching in that direction. Once they encounter something they want to attack within their zone of awareness or whatever, they attack it. This also has the benefit of feeling more natural; why would the biters know where every assembler and solar panel and electric pole is in my base?
   This is definitely a major piece of the "work smarter not harder" opportunities. I could simply build a wall around everything. That would be working harder. But, what if I was able to figure out where biters are going to attack? Then I can support those areas with my defenses. I can improve my factory by being efficient with my resources, if I'm clever enough. Solving those puzzles is just as rewarding (if not moreso) for me as solving the "How do I make a whole lot more of this" type of puzzle in this game. The idea of only defending specific areas works a little bit in the early game, but any factory generating any significant amount of pollution can't do this in the mid and late game; no matter how aggressive you build defenses, if there's a gap anywhere the biters will go through. Similarly, even if you do keep pollution down somehow, simply expanding your borders means more gaps and more points you need to defend "smartly". (Sidenote: I agree with the sentiment in that FF I linked about about the over-effectiveness of clearing out of your pollution cloud. I recently converted my megabase to Nuclear Fuel, and the pollution it generates has made biters a bit more exciting again.)

Summary

Right now, biter defense is very uninteractive. The result is that the only way to defend against them for a majority of the game is simply to wall off your entire base. I want to make sure that is still a viable playstyle, but I would like to see other playstyles be viable as well. Additionally, I think these suggestions can improve the experience for newer members.

p.s. Yes I know you should only put one idea/suggestion per thread, but these really are each suggestions to address what I see as the main design problem. Addressing that design problem in some way is the suggestion.

[Honktown]

I don't have as many hours, but I've played on mostly high-biter games (to the point of everything maxed in .16) and felt to comment.

In my experience biters almost always attack the same things. Almost. It is random sometimes what they decide to go for, but with large or small bases, I could tell from where they'd be coming in, and knew where my gaps were, but not often exactly what they'd attack. I admit I've reloaded saves a lot, and seen them pick different targets, but the targets they'd pick made sense (miners and furnaces first).

Walling off your entire base is, in a sense, a feature not a bug. If enemies are everywhere, and they come at you, expect them from everywhere. If you're big enough to reach areas that are largely water, you've either got water turned up or should be able to have defenses everywhere (it's expensive, but your factory is worth more). With laser turrets, you shouldn't need walls, and flamethrower turrets need them because they can't kill the first guys. I've only used walls with gun turrets for affecting the flow of guys, because once medium, or worse, big biters should up, it's panic time and I need to have better turrets and/or less pollution. Medium biters can be managed with AP rounds, but it's very very expensive.

With regards to some of the "what"s

1) Radars are a bit power expensive, but they let you easily see guys forming groups. I've repeatedly gone out to harass them and lure them back, because I don't want another nest or 50 dudes coming at once (and when they chase you they prioritize you and the turrets get free shots - a lovely feature).

2) See 1. Use radars and look for big red blobs moving in. If your base is really big and you can't see all of it, use lasers (you always have to transition to them eventually anyway).

3) That sorta makes sense, but that's only relevant if your pollution is over a big blob of water, because enemies will not attack you unless they've absorbed pollution or are within spitting distance of your base. Probably more relevant with how water gen has changed, but then you know enemies will be coming from side A or B of the water edge.

4) That's normally sorta what they do, although they can pick different areas/things.

On the "why"s.

1) Yeah... you had a gap in your defense. Wtf were you thinking would happen? And that's a serious question. You didn't defend your base. Surprise.

2) See "what" 3. I've played many games where I engulfed an area far out-sizing my factory (we're talking 20:1 area or more, even after launching 200 rockets), and enemies always pathed around water closely. It was more annoying that if you were on the "convex" side of a water body, the turrets closest to the water would often take damage from guys hugging the coast.

4) I think there's been discussion over whether radar is a building they may actively target (especially with solar outposts) but I haven't seen anything official and mostly people said no. It does make sense they should attack the first polluting things they come across, under "active" circumstances.

With .17 I don't know how the game plays time-wise, but it's been feeling a lot longer before I can get to laser turrets, since blue science takes more oil products. In .16 with the maxed settings the evolution would get ahead of me (not a big surprise) and I'd be absolutely burning ammunition while I got a few laser turrets up here and there and turned the tide, or it was a game where I had to use flamethrower turrets because I didn't get time to reach laser turrets (that scenario is more likely, I didn't play too many vanilla games on the hardest settings).

Another comment: depending on how you're playing, your base doesn't actually need to be that large to get anywhere. I can imagine you're like a lot of megabasers and try to have a main bus and all that crap. Better organization takes more space, and you have to make a trade-off of how much area you have and how well you can build things. A tighter base means fewer defenses, and if you don't have the defenses, you have to build smaller/cleaner and take your time. In some games I've stopped researching while I expand my defenses and built up stock of ammo/walls/infrastructure, in another game I've literally shut down my base and had to move to a forested area because the amount of pollution I produced under normal circumstances caused more enemies than I could handle (stupid Overhaul+Prospector mods dropped my starting resources to a few thousand per pile and fractions per second of oil). That same game I had to strategically place single miners on patches with a turret, otherwise the pollution would be bad enough to lure groups in. Took a stupidly long time to make progress.

[Honktown]

Had to load up my "maxed enemies" game and check some things. 344 steam engines with steam tank accumulators while I was transitioning to lasers, and probably 225 flamethrower turrets were being used before I switched to nuclear power. Only 600 solar panels. You gotta do what you gotta do . You don't want to see all the spaghetti.

[DragonMudd]

I've made a video showing a biter path as it currently works. Obviously this is extra egregious on my megabase, and it is unclear where they would've changed their mind and attacked something else, because they happened to path right by some turrets. But the point is they pick a specific target within the base first, and then make some weird path towards it. And experience has shown that what may or may not trigger their anger along the way is highly variable.

The main point of the entire post, though, is not that I can't find a solution to biters. It's that there is only one effective solution; there is no choice awarded to the players to try out different strategies. And in my opinion, that is a weakness in the game design.

[DragonMudd]

Just happened to catch another one of those bizarre situations where biters scrape past turrets, ignore a bunch of polluters and other objects, only to attack some solar panels. I reloaded the autosave to see how it started. They were marching up the gap when a logistic bot wandered past, they attacked it and hit the solar panel and then began just fighting the panels until repair bots showed up.

My point with all of this, however, is that because they don't follow any predictable logic, there is no opportunity for players who are interested in trying to outsmart the biters in being able to outsmart them. The only way to prevent this is to completely close off every possible path into one's base, which is the "work harder" brute force method, and is fun, but so would be a "work smarter" option. Having only one available strategy reduces gameplay.

[Honktown]

The enemies are everywhere. It's already been it pointed out different places that the enemies largely do follow predictable paths, which is "confusing" for new people because they think the biters will attack from anywhere - and here you are expecting the opposite. You megabased and are somehow surprised you have huuuuuuuuuuuuge gaps they don't care about. It's true that it's odd they go for stuff deep in the base and don't hit nearer stuff, but that's how they work. They go for a high pollution target.

Tip for you: most people stop at 1, a few, or at most a few hundred rockets. They don't launch thousands. You're in screwing-around territory, way outside the normal difficulty of gameplay. You won. You won long ago, many times over. If you don't defend yourself, yes the biters will get you, but the game is not difficult any more unless you purposefully make it so, which you did by leaving crap open.

Watching your video a few times, the path the enemy took was completely predictable, it was the shortest path to its destination. If you look at your pollution cloud it was probably heading towards a dense area, as they do. I already mentioned enemies can only go around side A or B of water, so if you don't want them coming in by that water, put things there already. If you don't want them coming from that nest, PUT DEFENSES THERE.

Play a game like this and you won't have to worry about pathing:

ouch.png (661.27 KiB) Viewed 6401 times

[DragonMudd]

You're in screwing-around territory

Oh definitely. That's the point. With this base I can try out wacky things like leaving gaps and determining exactly what "logic" the biters take.

And the "pollution cloud" for a base like this is monstrous and literally just covers everything. But that doesn't mean individual entities still don't pollute. The point I was hoping people would see is that the biters pathing routes do what you and I described: they somehow have knowledge of every object in a player's base, choose one point, and focus on it, and maybe or maybe not attack something else along the way depending on some mysterious internal values. This odd behavior happens in all factories, but it's very plain to see in this factory I'm playing around with at the moment.

The original discussion I started isn't a result of what's happening in this factory; they're results of all my experiences in all of the factories I've built since 0.14 plus online commentary I've read from frustrated newer players. I suggested a slightly different set of rules the biters could follow instead that would allow for opportunities for a larger variety of playstyles. It would still allow players to play the way you want to play and enjoy the experience just fine.

[Honktown]

Idk man, it was dev notes or reviews or something, people usually turbo wall-off everything, when enemies don't actually meander much. Most I've seen is a mod I'm currently running where the enemies usually pick off corner/edge turrets. Most vanilla enemies aren't ranged or do much damage so I don't know if that's supposed to be normal behavior.

[ssilk]

This suggestion is in my eyes a description of how to learn enemy behavior.

I can say this:
- if you leave your base open there is a chance, that they will find a way
- there is no need to build walls everywhere (only on high frequented parts), but you should have a wall of defense, because that holds them off for a while, so that you are able to react.
- you can build walls from other “cheap” stuff like pipes, or old furnaces/miners...
- learn, what buildings attract biters: not only weapons
- cover the most important parts with a second shield, leave some random turrets around. Most important are the power production (water), oilfields, mines, smelters....
- in my eyes, there is no need to watch the path they take: put some random turrets out and look which one gets destroyed first.

[DragonMudd]

Only the 4th suggestion is about biter path selection. The rest are far more straightforward, and really geared towards the new user experience.

And the attached video clips demonstrate that there's still too much variation in biter "logic". Logical biters would have chosen any of the polluters nearer to them, and/or never attacked the solar panels. They aren't truly predictable, which means that there is no "work smarter" play, only the "work harder" play.

[Honktown]

And the attached video clips demonstrate that there's still too much variation in biter "logic". Logical biters would have chosen any of the polluters nearer to them, and/or never attacked the solar panels. They aren't truly predictable, which means that there is no "work smarter" play, only the "work harder" play.

What you're suggesting for "work smarter" is absolutely trash for gameplay. If I know what the enemies will always attack, it makes for a terrible game. Put turrets around miners/pumpjacks, there, literally done, unless they hit other things first, then I just look at the straight line from their nest to me and put turrets in the way. This is dumb. Extremely dumb. If I could put productivity modules in miners and lure every single enemy on the map to that spot I wouldn't consider it "playing smarter", I'd consider the AI broken.

Enemies attack semi-random targets, and don't instantly veer off course if an assembler farted upwind? I have to defend myself. The enemies are dangerous. I have to play smarter, managing my defense production and pollution, or else I will have problems. Defending yourself IS playing smarter. It takes more work and makes things more complicated, for the reward of not having your factory smashed to bits.

Efficient pathfinding and making it interesting is actually a very difficult thing to do, so I'm a bit surprised the pathfinding in Factorio is actually good over long distances. See AAA companies screwing it up constantly when all the enemies take exactly the same path or do exactly the same things. I don't want Factorio to be made worse. Yes, bigger numbers are normally what people focus on, but there are other aspects of the game.

[BlueTemplar]

But biters are predictable, as you say :

They were marching up the gap when a logistic bot wandered past, they attacked it and hit the solar panel and then began just fighting the panels until repair bots showed up.

So, they were going to attack the most polluting emitters "nearby" - (I bet it's the local pollution maximum ?) - but were distracted by a logibot, then went on a rampage. The only thing not predictable about that was when/if exactly they were going to meet a bot, and you have control over your bot coverage...

[DragonMudd]

Good lord, you all are really missing the forest for the trees here. Forget the fact that the way biters are implemented causes many players to feel bad and turn off biters. Obviously, in my 2000 hours of playing, I must not know what I'm talking about.

[Honktown]

Good lord, you all are really missing the forest for the trees here. Forget the fact that the way biters are implemented causes many players to feel bad and turn off biters. Obviously, in my 2000 hours of playing, I must not know what I'm talking about.

1sps is quite far into the game, so it seems they didn't understand the enemies will get worse and you need to switch to flamethrowers/lasers, and those have their own potent difficulties. This isn't a problem with biters, it's a problem with offensive capability and knowing what to do about biters. There isn't a gradual shift in military tech, and within a few hours you can go from killing every wave with bullets to bullets being completely worthless. Much of this is related to having any idea how the damage/armor system works (and you'll need a calculator if you don't want to destroy all your resources).

The same time, the game doesn't let you control pollution until mid-game game with one specific thing indirectly affecting pollution (efficiency modules affect power, not pollution on the label), nor can one have an intuitive sense of how significant pollution is without many games of trial and error. I would easily expect newbie players to throw down as many miners and furnaces as possible to get plates/steel incoming as fast as they can (and not think it about until they run into throughput issues or many mines depleting or whatever). Obviously this will get them attacked constantly, by bigger enemies, in larger groups, AND cause evolution to advance much faster.

The damage/upgrade system and enemy health/armor needs a total rebalance, so upgrades are actually like, useful. Something players would immediately notice, like a base magazine kills a small biter in 5 shots, and the next upgrade brings it down to 4, then 3, etc. With 5 damage, it's not until a 50% damage boost you need any fewer bullets to kill a small biter. Only the fourth tier of physical projectile damage makes you need any fewer bullets for small biters, turrets at the third, to kill the first enemy in the game more effectively per bullet. Even with AP rounds, the player needs an 87.5% damage boost (yellow science, sixth upgrade) and turrets need the third upgrade still. Bullet upgrades are more effective against higher HP enemies, but A) it's going to be too late and B) while effective against medium biters, bullets are garbage for big biters (one should have better turrets by that time, but how is a newer player supposed to know?).

The way biters are implemented is very good in most ways, and very predictable. The problem is countering them correctly is unintuitive and even misleading (projectile damage is near worthless for a long time, bullets are awful, efficiency modules don't just affect power).

[BlueTemplar]

Against blue biters, damage dealt per bullet roughly doubles with each upgrade level... the issue is that you're still much better off rushing oil and flame turrets rather than getting all the pre-blue science damage upgrades... (also, AP ammo is pretty much wasted on small and medium biters and all spitters)

P.S.: Good point about biters having limited HP, I should try to integrate that in my graph somehow...

For pollution, finding out how it works should now be much easier in 0.17, now that we have the pollution graphs !

[mudcrabempire]

Please note that small biters/spitters have 15 resp. 10 hp, while basic ammo deals 5 damage per shot.

Which sounds like you need 3 resp. 2 shots to kill them.
Which is actually wrong, because biters regenerate health. So during the delay between the first shot and the "killing" shot they will regenerate just enough to barely survive, requiring an additional shot to finish them off.

So actually the first damage upgrade reduces the number of shots to kill small biters/spitters from 4 to 3 resp. from 3 to 2.

[BlueTemplar]

Hmm, does this mean that you might be able to save resources by using massive overkill - so that they get shot enough times in a single tick ? (Though that would require the biters to start out in the range of several turrets at once - a contradiction for more than two turrets unless you place the turrets with bots ?)

[Cribbit]

Against blue biters, damage dealt per bullet roughly doubles with each upgrade level... the issue is that you're still much better off rushing oil and flame turrets rather than getting all the pre-blue science damage upgrades... (also, AP ammo is pretty much wasted on small and medium biters and all spitters)

P.S.: Good point about biters having limited HP, I should try to integrate that in my graph somehow...

For pollution, finding out how it works should now be much easier in 0.17, now that we have the pollution graphs !

I made a spreadsheet back for .16, slowly upgrading for .17: https://docs.google.com/spreadsheets/d/ ... edit#gid=0 (only the first page, damage/shot, done so far)

But for gun turrets the damage upgrades are actually huge.

Flamethrower turrets miss the first bit of a wave, and in the early-mid game waves are too small to be effectively hit.

[ssilk]

I re-read this thread.

@DragonMudd: I think being more specific would help. So what I would suggest is to write 4 suggestions instead of one and making them as precise as possible.
For example:

Biter trails

Biters and other enemies should make traces, visibility and time of visibility should depend on underground. Hard underground lasts shorter than soft, hard is less visible. Sand for example depends on wind speed, how long it is visible. And there is room for much more...
Reason: to have a hint, where the enemies had been coming from, especially beginners have problems to find that out.

Etc.

Besides that I think from your description, that your playstyle is relaxed - you don’t like to be distracted from your current task - so it might be an option for you to turn enemies off (as me sometimes do).

Edit: and what we should not forget is how the new biter behavior and pathfinding will influence gameplay. That is currently not really known, so it might be also an option to wait a while and play the game instead, and watch how the game works now.