Some thoughts on 0.16 -> 0.17

[CDarklock]

Let me begin by saying that I am entering this with a basic mindset: the direction of the game is correct.

It doesn't matter whether I think it should be going some other direction or not, so I haven't invested any time or energy in what direction I think it should be going. The game has changed the way it has changed, and will continue changing the same way it is already changing. If I don't like it, I can quit playing, and nobody else will care.

So what is meaningfully different from 0.16 to 0.17, and how does that alter the way I need to play the game? Maybe that will help someone else who is frustrated and struggling. The changelist isn't that helpful, I don't think, because of one major observation I've made.

The change in how biter bases absorb pollution (covered in FFF #283) is a much bigger deal than people seem to expect. It's just one line item in a big list, but it matters more than anything else I've found.

Now, I only have a vague understanding of it, but the way I interpret the change is this: in 0.16 and before, when pollution got to a biter base, it stopped. The base just sucked it all up like a vacuum and then trickled out a unit every few minutes, forever, because it had enough pollution to fuel producing hundreds of units. So for basically the entire game, that one closest base was the only one doing anything about your pollution.

Now, in 0.17, the base only sucks up enough for three units. I don't know all the details here, but lets think of it like a factory.

Imagine that you have a whole bunch of assembly machines, sitting right in a row, and you send a belt full of materials at them. If the first machine takes all your materials and throws them in the machine, your production is crap, because only one machine can produce anything. You need to throttle your consumption, so materials get to the rest of the machines, and then you double and triple and further multiply your production.

This is why inserters only stick enough materials into your machines to make two or three items: it's a waste of materials to stuff every machine completely full all the time. And this "just works." You don't have to think about it. You stick your inserters on the machines and they automagically throttle their performance to something reasonable, because that's probably what you want. If a machine backs up with two or three items, it stops producing, because clearly you do not need them or you would be taking them out.

So think of a spawner like a machine that converts pollution into enemies. We've taken a machine that consumed all the resources you threw at it, and we've optimised it - just like your assemblers - to only collect three enemies worth of pollution at a time. And the rest of the materials (pollution) move down the belt (spreading through chunks) to reach the rest of the machines (spawners).

This multiplies their production. Just like it does in your factory.

That means attack groups are bigger, and they show up more often. Plus there are other differences, namely that they're faster and hit harder and splash damage is a Big Deal. (There's a great video about these other differences on YouTube.) And that means defending your factory has gotten a lot more important. Military research is a lot more important. And you need to do it sooner.

Base hunting is also a bit harder, but... you still get to a point that it's easy relatively quickly. It will be a bit farther down the tech tree, but not too far. Base hunting on foot is a lot harder, if there are worms. Acid pools are a big deal. Splash damage is a big deal. Things are different, and they'll take adjustment.

Some people are used to similar things from other games, and will just go "oh, this is just like that other game" and their existing skill will easily transfer and they'll probably be all smug about it. Some people (like me) are paranoid by nature and were already gearing up for biter attacks that just never happened in 0.16, but we still did it because you never know, man.

I've always ringed my bases with walls and turrets from the early stages, before I even have decent furnace arrays. In 0.16 this made me feel stupid, because I would expand my base and when I took up the old turrets not one of them had fired a single shot. When I do that in 0.17, I often find little damaged areas of wall and the turrets are actually running low. I might be a little smug about that. A bit. Just a bit.

But it's different, and as I watch people streaming and uploading gameplay videos, I find a lot of them frustrated - or at least surprised - by the differences. So I thought I'd just jump in here and tell people what I've noticed, and maybe it will be helpful.

And if you've noticed things yourself, you're welcome to add your own thoughts. I'm still early-game in my own 0.17 run, and I know some people are already launching rockets in it. Maybe this can be a useful "how to migrate painlessly into 0.17" resource.

[vedrit]

So many +1

Some people (like me) are paranoid by nature and were already gearing up for biter attacks that just never happened in 0.16, but we still did it because you never know, man.

I am one of those people. But worse. I had my perimeter wall, and then I also had outposts further out to soften up what I thought would be hordes of spitters/biters, only to find those outposts were more than capable of handling most attacks, so my main perimeter never saw any action at all.

[Light]

This was how 0.15 and 0.16 were for me, but only through the support of mods. Defense was top priority when possible, or it meant being surrounded by nests with constant swarms.

The 0.17 update shakes up the experience for vanilla players and that's something that should be praised, since the research priorities and cookie cutter builds were second nature to many. Something to break the cycle was required, or it would just be 0.16 prettified with the same priorities and cookie cutter builds.

I do find your analogy of biter assembly machines to be amusing, because I'm picturing queens in the nests just pumping out eggs like an ant colony. I guess she goes to sleep when the air is clean, then wakes up when she can't breathe and starts creating her army.

[Koub]

For people who build small, and with low pollution footprint (as I usually do, always being reluctant to attract biters), I guess the difference will ne be as surprising as for all the people building like there's no tomorrow - and no biters - claiming that biters will never be more than a non issue.
I'm really looking forward to have the time to test 0.17 .

[FunMaker]

I did not play 0.17 yet. But what about the easy way - killing all biters in pollution range. For me this was the easiest way to come by with the biters in 0.16. Indeed it is boring as hell and i think the devs should adress this.

[CDarklock]

But what about the easy way - killing all biters in pollution range.

Good luck with that

Seriously, it's harder than it used to be. I was starting to think this was better than setting up a perimetre, in 0.16, but it's so much harder in 0.17 that I'm back to walls and turrets. Until you get heavy armour and grenades, base attacks are way too risky for my taste... although I'm told, once you have a car, they're trivially easy again. And, of course, turret creep still works.

[Ranakastrasz]

But what about the easy way - killing all biters in pollution range.

Heh. I used to do that, way back when. Eventually I was confident enough in my defenses that I only really needed to clear them out to avoid constant attacks from proximity aggro.

I was especially worried that if my pollution got bad, it would spread deeper and send more waves. But that never happened because of the bug.

Now, well, anything in the cloud is a threat. Cut your pollution back, and clear enemy bases. I wonder how many people will stop considering efficiency modules worthless now?

[bluerock]

But what about the easy way - killing all biters in pollution range.

Heh. I used to do that, way back when. Eventually I was confident enough in my defenses that I only really needed to clear them out to avoid constant attacks from proximity aggro.

I was especially worried that if my pollution got bad, it would spread deeper and send more waves. But that never happened because of the bug.

Now, well, anything in the cloud is a threat. Cut your pollution back, and clear enemy bases. I wonder how many people will stop considering efficiency modules worthless now?

Since the first oil patch you get to in 0.17 is outside the starting area (and likely to be close to an enemy base), I just checked to see how much pollution could be reduced with two cheap lvl 1 efficiency in a pumpjack. It reduced the pollution from 9 to 3.6. This is quite a reduction, which seems really good considering how much the biters hate pollution now. Researching them can be done early with red/green science and is pretty cheap, but unfortunately the module still takes plastic to build it so can't be used when the pumpjacks are first set up. But shortly after. I'll try it.

[CDarklock]

I wonder how many people will stop considering efficiency modules worthless now?

I certainly never saw the point of them in 0.16, but I sure as hell do now.

[Ranakastrasz]

Personally I found them useful to cut the massive power consumption of electric furnaces in the first game I got that far. Helped a lot. That said, once you can casually sprawl solar panels make power irrelevant, and nuclear power is kinda op due to the advance refining thinggy.

[CDarklock]

once you can casually sprawl solar panels make power irrelevant, and nuclear power is kinda op due to the advance refining thinggy.

I am still overly confused by solar and nuclear. I feel like the ratio of panels and accumulators is complicated, so I don't want to do the maths to work it out, but I also don't want to just grab someone else's blueprint and use their design. Nuclear utterly confounds me, I haven't even tried to work with it yet.

[xfir01]

I feel like the ratio of panels and accumulators is complicated, so I don't want to do the maths to work it out, but I also don't want to just grab someone else's blueprint and use their design.

25 solar panels to 21 Accumulators or thereabouts. Could probably go 5:4 and throw in a random area of accumulators get close enough for most purposes.

Nuclear gets complicated with the adjacency bonus. A standard 2x2 reactor will produce 480 MW with 4 reactors, 5 pumps, 48 HE and 83 turbines. Though a lot of designs have steam tanks and save fuel as well.


https://deniszholob.github.io/factorio-cheat-sheet/

[MoleOnDope]

once you can casually sprawl solar panels make power irrelevant, and nuclear power is kinda op due to the advance refining thinggy.

I am still overly confused by solar and nuclear. I feel like the ratio of panels and accumulators is complicated, so I don't want to do the maths to work it out, but I also don't want to just grab someone else's blueprint and use their design. Nuclear utterly confounds me, I haven't even tried to work with it yet.

Well you absolutely should then, it's fun

Solar is rather simple. The perfect ratio accumulators to panels is 21:25 or 0,84 with no rounding and one solar panel produces 42kW on average over one day and night cycle factoring in accumulators. So you just make your own blueprint that keeps the ratio or at least close to it and has a specific average output. From then on you just stamp it down however many times you need it.

Coming up with an "optimal" Nuclear setup has been a huge time sink for me, but even more fun.
For getting one reactor going (and keep it running) you don't need more than 4 uranium miners and 1 centrifuge constantly running on average if my math is correct. So maybe double that to make up for unlucky RNG at uranium processing.
Setting up kovarex and optimized reactor builds takes time, but boy does it feel good powering up your personal mean green monster

[DaveMcW]

For even easier math, peak accumulator charge should be 100 times peak solar production (click a power pole to check). Just add more solar panels or accumulators until it balances.

[torne]

Solar is rather simple. The perfect ratio accumulators to panels is 21:25 or 0,84 with no rounding and one solar panel produces 42kW on average over one day and night cycle factoring in accumulators. So you just make your own blueprint that keeps the ratio or at least close to it and has a specific average output. From then on you just stamp it down however many times you need it.

Or if you're not particularly interested in figuring out ratios and things: build a bunch of solar panels and a bunch of accumulators. If your accumulators run low before day, build more accumulators. If your accumulators don't fully charge by night, build more solar panels. Since the factory always grows, you're never really going to have *too many* of either.

Having the correct ratio only starts makes a significant difference if you're going to make a blueprint and stamp out dozens/hundreds of copies of it: the amount of resources you "waste" is the combination of "how far from the right ratio is your blueprint" and "how many copies did you make" - but even then, it's only actually wasted if your factory never grows large enough to need the extras (say, because you spent too many resources producing a really significant excess of accumulators and the biters destroyed everything...)

[LordWampus]

...gearing up for biter attacks that just never happened in 0.16, but we still did it because you never know, man.

Always be prepared!

[CDarklock]

Always be prepared!

Always be closing!

Second prize is a set of steak knives.

[Bauer]

Always be prepared!

Always be closing!

Second prize is a set of steak knives.

And I thought the new ABC rule is "always be charging"...

[Hiladdar]

Having "mucked" around with 0.17 in a good way, here are my comments.

Starting a new game is not that big of an issue. Likewise converting a pre-rocket launch is quite a bit harder, especially if it was developed with using a .16 science packs.

Here is where I found an issue, and that is with bases that have several thousand launches. During the conversion process a lot of the military research from military science, and post .17.5 from mining is wiped out. On a mega base from .16, it is fairly easy to re-research most of the, especially if the base keeps a buffer of 100-200k each science pack. That process can take a few minutes and not an issue.

But if that game had .16 military research in the low teens where the next level to research was over 100k, and mining productivity well over 250, that now becomes an issue with .17. The issue is that with the changes to military research being changed it is also reset, as is mining productivity. This reset can take many playing session hours to reset, resulting in a negative experience for the user.

I think this is ok, for the experimental version, but not ok for the sable version which is still .16.51. My recommendation is to have some sort of check and auto granting of military science / mining productivity when a game is upgraded from .16 to .17. My expectation is that it not be exact, since it due to the changes it can't, but once a game is loaded up from .16 to .17 it should be comparable to levels of research conducted in .16.

My basis for this is I enjoy developing mods about as much as playing the game, and have several different games I use to test the mods I develop for compatibility and play-ability with other mods. This is key for me since I need to have a fairly quick way test a research and new components of a module, as well as stress test the modules on a heavily modded mega base. If I am experiencing this as an issue, I am sure that someone who is still on .16 and is not as technically literate, yet has invested a significant amount of time on a base post launching the first rocket, might be slightly annoyed.

My second observation are cliffs, especially so close to the starting area. Although they do can be used for defense of a base by putting your turrets behind a cliff, and can be used to channel biters, they are a significant issue until the player has obtained cliff explosives which is locked behind sulfur, which needs oil. This means that there is no effective way to deal with cliffs. My recommendation is to not have any cliffs in the starting area, are at least a way to configure that option from within the game start up menu.

Regarding biters bases, that a number of others in this thread have posted. One technique is to set up what I call zones out side the main base as you are building to the first rocket launch. First this provides a buff for your pollution to expand and be absorbed by the environment. Second thing the various zones do is they can be used to channel the biters into an ambush kill zone for quick death. Finally if an exterior zone wall is breached, it gives you time to react buffer up defenses in a sector of the base. Finally work on integrating defense into you build, manage your pollution, and that means maybe turning stuff off and on, or using efficiency modules.

Hiladdar

[CDarklock]

Starting a new game is not that big of an issue. Likewise converting a pre-rocket launch is quite a bit harder, especially if it was developed with using a .16 science packs.

The detraction of "converting an old base is hard" necessarily decreases with time, of course - not only do fewer and fewer people need to do it, but fewer and fewer people expect it to work.