Friday Facts #329 - Campaign reassessment

[FactorioBot]

Here it is! (beep boop)

https://factorio.com/blog/post/fff-329

[chris13524]

What about old/abandoned/ruined factories?

Edit: lol wut? My comment here originally read "First!". Did a mod edit it to what it says currently? Am I trippin? wtf

[doktorstick]

I'm genuinely curious--what methods did you use to determine your conclusions about NPE?

[raiguard]

I'm not a level designer or a storywriter, and I appreciate all that has been done for the good of the campaign.

But no story?

The premise of Factorio is simple: you have crash-landed on an alien world, and your ship is completely destroyed. You need to build up a Factory that can build a rocket to get you home.

So why not make the story as simple as that? Have a cutscene at the beginning of the ship crashing, and the player emerging from the mangled wreckage. After that, it's basically like a segmented freeplay (this is going with the expanding campaign). The player achieves each step with some minor help from the scenario, and progresses to the rocket. When they build the rocket, have a special "escape capsule" payload that they put in, then in an ending cutscene they board the rocket and fly into orbit.

The rocket is definitely not interstellar-capable, so perhaps there is already an interstellar ship in orbit, and the ship that you crashed in was simply a lander ship? That would explain why the rocket is so small, because you only need to get to a low orbit. You dock with your interstellar ship, and return home in an epic triumph.

The story doesn't have to be a complicated drawn-out thing - just go with the premise that has already been established! Have story at the beginning and at the end, with the middle being completely up to the player. After finishing the campaign, the player can progress into freeplay, with the only difference being an infinite world and non-segmented gameplay, and launching satellites to get space science instead of an escape capsule.

It's probably a lot more complicated than I'm making it sound, but I thought I'd give my two cents.

Thanks again for all that you guys do! Best game ever.

[ElCapitan1701]

I always loved the old RTS games like Age of Empires, Command and Conquer because of the fact, that you had to build your base again and again. I also restarted Factorio multiple times playing the same content - just to do it better. So, I think the level based campaign would be better and I encourage you not to give "some starting items" and making "the gameplay (...) about short term problem solving" ( I always hated the mission with a half build base) because I think factorio players love to iterate and bring things to perfection. Rebuild a base in the next mission is a good way to try a new better way and avoid some mistakes. There is nothing more painfull then dragging an old, bad build base from mission1 to mission5. Sure, you can`t do that five times. But if the middle mission (the third) would be something different I think people wont be frustrated.

[OvermindDL1]

This is the main prototype that we have been working on so far. A single map which starts small but grows after each objective is complete. This would emulate Freeplay gameplay in that the player can build very large bases and expand in the directions they want, but with quest objectives to steer the player towards building the rocket.

The RTS game Warzone2100 (open sourced a long while back if you want to give it a try, it's a lot of fun, though careful as it might give you idea on user-designable units for Factorio. ^.^) used this style.

Essentially there were 3 campaigns played back to back. Each campaign you had your main base area along with the rest of the map. Each time you completed a 'level' of the map then the next map was the same map but expanded bigger, and by the same map I mean everything is the same, your researches, what and where you built things, etc... When you finished a campaign you were called to help save one of the other 3 bases and this started the next campaign (usually in the middle of a major battle to save them, bringing your units from your old base via blimps again). It was a very nicely cohesive style of map.

However, expanding a bigger map would get boring after a while (things would get so far away!), so what Warzone 2100 did was you started on your base on some map levels but you'd load units into a carrier blimp thing to take to a new area to do raids, get new technology to research, etc..., and you could go back to your main map when done. You had a remote interface while on the raiding map to load more units into the blimp thing, and it took travel time based on the distance you were raiding to and the blimp had a limited cargo capacity, which was your main throttling aspect of those maps (always bring a couple trucks to start with is my goto style, defend up the landing site!). Like imagine when you crashed the ship in orbit split up and rained down on the planet, you had to recover the pieces for key technical recoveries to be able to research things via such raids.

In short though, I *LOVED* this style of expanding map. I loved being able to raid far off areas to acquire new technology to research and get as my own. Although you only mentioned the expanding map part, having some raiding areas with some form of transport adds a lot to a campaign without it just feeling like the map is bigger. I'm unsure how precisely to integrate that idea into a Factorio campaign but I do have a few ideas that I think would work. However the expanding map style is definitely loved and I highly recommend it, just if there are level changes between them, keep what the user built, where they built it, etc... etc... It both has you expanding your base, and basically having to rebuild it twice more with the technologies you've already acquired so you can do it better.

And play through Warzone2100's campaign as well, swipe some of the design ideas please.

https://wz2100.net/

[MrGrim]

That video was great! "Thank you, Klonan."

[BlueTemplar]

Kalon the barbarian !
That was a bloody spiffing game ! (even with all the server drops...)

----

- No story.
- No exploration gameplay.

This makes me sad, as the previous campaign and the 0.17 NPE (and Factorio Abandoned Colony/Cities, though the scale of that is obviously out of scope) have shown how much potential there is there !
Why ? So you can be sure to release on 2020 September 25 ?

----

Am I only one to find it weird how this claims to emulate "freeplay gameplay", and yet we see car & train before MilSci and train before oil ?

[benkor1]

I like the expanding map idea, but I worry it will lead to situations where players will have to rebuild bits of their factory in order to get stuff to the new section, as they've built a lot of spaghetti in the way or something. Why not have a system, where you build a set level of production in the previous level - i.e. for the first level, you need to make 15 iron and 15 copper per second - and then belt it to the edge of the map, at which point the map expands in that direction. That way, it teaches you a little about organising a factory, which may help point players towards busses later on in the freeplay, while also ensuring the player can jump straight into the next section.

Of course, this somewhat breaches the idea that players learn better the hard way, but imo forcing them to possibly rebuild a whole section is too hard. I'm sure there are other ways around this, this is just one idea.

[ledow]

General rules for tutorials:

Players like to know what to do: now, next and in the future.

Players want to do that themselves. They don't want you having to spoonfeed them and limiting the speed at which they can do it (if I want to wait for things to scroll and read the whole tooltip, I will, but don't steal my control and force me to click through ten messages about the thing I KNOW I need to do).

Players want to be able to literally say "What was I supposed to do?" either through a repeat button or something, to remind them what they're trying to do, what they need to do now, or literally what button needs to be clicked.

Players don't like invisible walls that block progress without reason. This applies to maps, and to greying-out 99% of the options rather than let them play and find out what they do and how they work.

Players don't want to have to follow *your* route, they want to spend all their money on that sentry gun that shoots cool things, and not wait for money to build back up so they can put down that second coal miner you want them to. In general, this means that the start should let them have plenty of resources in hand/chest so that while they're waiting for that bit of coal to get all the way to the furnace, they can mess about with other options and do things while they're waiting.

Players don't want to die. This generally means never letting them be in a position where they're outgunned, or where they HAVE to craft all this stuff before it's safe to step outside.

Ironically, this means tutorials needs to give players lots of stuff to make things really easy, then get out of the way, without leaving them completely alone, then gradually make the same game *harder*. This is the flip of how games like Factorio start (where they have nothing, it's hard/takes a long time to get anything, and when they get things like bots and guns the game suddenly gets much easier... at least for a while).

Any tutorial for a game, however, follows the exact opposite of this. You have nothing, can't go anywhere, have to learn everything, can do nothing, sit around getting bored, then get attacked even though you had tons of time in which you *could* have seen / prepared against the threat, and then get baby-sat on how to build the next item exactly when, where and how even though it's basically the same as the previous three whatevers.

The guy crash-landed on a planet. Put in a huge damn shipwreck with tons of already-gathered resources - show them what they all look like, how to manipulate them, let them build how and where they like. Let them secure themselves and *then* show them how to extend those defences for the future threat that stays put until they are ready for it. It's a perfectly reasonable storyline and method of teaching.

Honestly, if game tutorial people worked in schools, we'd set every kid back 5 years. The kid is screaming "I just want to put it in the box, I know the answer!" and the game is going "Step 2: To get the next achievement click here to select this..... now click here to select that.... now click here to put it in this EXACT square... Look isn't it cool?!?!?!!" and not letting the player do anything themselves.

Hell, the number of tutorials where there is literally only way one to succeed and you have to place units/towers EXACTLY and you have EXACTLY that amount of money and not a penny more. Dull, dull, dull.

I didn't even bother with Factorio tutorials whatsoever. Hell, I spent my first game on an island wondering where all these aliens things were and got to (was it blue science? It was a while ago) before realising I had to kill aliens for the purple science and there were none and I couldn't build land. My second game, I got the rocket.

In fact, most games I've ever played, I've never bothered with the tutorial (unless forced).

Gimme a basic instruction manual, and let me at it in a safe sandbox with an obvious goal and a way to discover the route to that goal without guesswork.

[bman212121]

Honestly, why can't we have both?


#1 is likely better suited for players who have played the game before, but if someone wants to use their "shoe string and bubblegum" map to help get them to the end, that seems fine to me as well. Personally I would prefer this option for a game like Factorio, because it has such a long build compared to other RTS style games, it would be annoying to perfect something only to throw it away on the next map. Something that I always have done personally is make a different save after every campaign step. Old games like Warcraft 2 gave you the ability to save when you finished. So if you use separate saves, then you can always go back a step or two if you decide what you did for the last part was so bad you want to do it over, but you don't have to go all the way to the beginning of the game.

The more I think about #1, the more it really seems like the campaign should be dynamic rather than linear. If you wanted to do this type of campaign rather than having a strict linear progression, you would essentially just have a player in free play mode, but allow them to decide when they want to start the next phase of the campaign. So in that essence the game would be much more like an RPG game where you talk to an NPC and that starts the next "quest". This would allow the player to naturally progress when they were ready to, rather than what they were forced to. They could just decide to run the gameplay back to back as fast as possible, or they could complete the quest and then spend another 5 hours just messing about before they start the next one.


#2 This is likely suited for NPE, where the players are really just given a small task and once they complete it, they move on to the next one. In a game like Factorio there is so much build time I don't think it would work as well to start a player over at square one, since it would take far too long to get going again. Most other RTS games this is like a 30 minute time to get going, it might be several hours in Factorio to get to that point. So really you're still kind of doing #1, you're just giving the player a pre-canned save. The main advantage of this mode is if a player gets stuck and can't complete the previous step, they could potentially just skip it if they couldn't figure it out. For me personally if I were going to make #2, I would have a menu that just lists all of the scenarios. By default scenario 1 would be unlocked. As long as you attempted scenario 1, that would trigger an unlock of scenario 2. So on and so forth until all scenarios are unlocked. That way a player doesn't get walled and can't progress, but it also means that if a player really liked one scenario, they can just pop into the menu and play that one over and over again. (It also would allow people to speedrun them since all of the starting conditions would be identical)


So for me #1 is the true campaign mode, but #2 would be a great scenario mode for the game. #1 is ideal because it allows you to not have to rebuild each section, and you get to learn how doing something poorly will cause you more problems later on. (Which is an important lesson in Factorio) You could be doing the main campaign, and hit a point where you're stuck with progression. You could pause that save, then go into #2 and play the matching scenario for what you're trying to do. The scenario might help you realize where your mistake is at, and then you can continue your campaign. I think that both of these would really complement each other, they just need to be handled a tad bit differently.

The unlock suggestion I did for #2 could also be mirrored during the campaign mode. Once you complete the same objective in campaign it will automatically unlock the scenarios as well. So it doesn't matter if you played through scenario mode or campaign mode, the scenarios would be unlocked. They both would have identical objectives, but in scenario mode you could make it do more hand holding, like putting ghosts down for a player so all they have to do is build the item. With a simple toggle of ghosts on or off, the scenario would function as both an scenario and a tutorial at the same time. Both would help reinforce how to play the main campaign and offer different levels of guidance when a player gets stuck. Really, you'd end up with 3 difficulty levels, 100% guided with ghosts for players who never played the genre, pre-build bases so the player doesn't need to rely on their previous work, and then completely hands off for people who want to figure it out on their own. All 3 of these could be on exactly the same map, with the same NPCs, tooltips etc. The only difference is that the easiest mode is literally someone from Wube playing the campaign, and it showing ghosts of where you placed everything on the map, one with your bases at the start but letting the player figure out the next step, and the final one is all on the player to make it through.

[melind]

Apologies for the long-winded post. With the recent FFFs I've been thinking a lot about my experience first playing the campaign and my first freeplay games. I'm overall very glad about the new direction of the campaign, and in my mind the goal should be to try to be hands off (let players make mistakes) but provide some guardrails around the freeplay experience. (Which is normally what I suggest to new players in terms of tuning down biters to help reduce the fail state new players can easily get into with default freeplay)

A few thoughts on elements I think will make the campaign fun:

• Allow the player to solve/meet requirements how they want in multiple ways to move forward. Rather then locking access to an area to specific technologies or setups, unlock areas at a natural point and allow the player to decide when they are ready to engage in it. I didn't use trains at all until my 3-4th game and it would be bad to stop the player from moving forward over something arbitrary like this.
• Place prebuilt structures in the world. When playing through the original campaign my mind was blown when I ran into that gigantic factory with dedicated smelting columns and train systems. I spent hours trying to understand how it worked and rebuild it and it was many times larger then anything I had even considered building at the time. Moments like this are helpful for players to understand the scale at which things can be built.
• Make it worthwhile to explore the map. Players can find ruins of factories behind optional biters bases. Islands with power grids set up on them. Cliffs blocking off value items (for that point of the game).

I also love the idea of a continually expanding map and would love it it the criteria to expand the map was largely self-regulating. With what you have it could be something like :

1. Starting area of the map, to unlock area to you have to put down radars.
2. You've scanned out additional resource patches, player needs to start using these to scale up their base. You notify the player they can continue to expand by crafting landfill.
3. Once the first landfill is crafted the map expands and the player can bridge over water to an area to a previously destroyed base, biter nests, and oil.
4. Once the player has started crafting blue science, they are informed that biters are getting restless and the final area of the map opens.
5. The last area of the map has biters with expansion enabled. Perhaps a few solar satellite setups so the player can see them expanding their way, but multiple natural chokepoints the player can make use of. To continue the player can make use of landfill or cliff explosives to open up areas of the map and/or to help fight the biters or get access to resources. This segment continues until they launch a rocket.
6. Once the rocket is launched the map opens in all directions. Perhaps an infinite map, but if thats not possible it just opens up really far (100s of chunks) in all directions and the player can continue to build on the map if they want.

Compy could be redesigned to be an element the player can click on to see their objective, maybe some hints on their current predicament, and to access tutorials. Of course Compy should be enable-able in free play as well

[zebediah49]

A significantly less long-winded post, but based on

Main Problem: At the end of each 'chapter' the number of different situations the player could have gotten themselves into is near infinite. This makes it very difficult to predict the state the player is in, and construct an appropriate challenge.

I would like to point out a technique used by Rimworld in particular, but to a lesser extent in other games. Specifically, the game will add up the total value of all of your stuff (loose items, buildings, gear, etc.), and then use that to adjust difficulty.

For Factorio, it's less of a cure-all, but it could be a useful tool in the toolbox. When starting the next stage, scan the map, make a heuristic for player strength, and then use that to inform the difficulty.

[Cgeta]

Aw I was hoping there'd be stories and exploration
With different maps featuring different challenges and detailed landscapes. I've been playing around in the editor for a bit and it's amazing how varied you can make maps by combining different terrains with different decorations.

And even once you have "unlocked" everything, the campaign should continue at least for a bit so you can actually use your endgame techs for a while. Afterall, one of the big features of factorio is that even when you have researched everything, and sent a few rockets, there's still so much you can do, and expand with. Large complex train networks that actually have multiple trains going to the same stations aren't really needed much until postgame for example

Or it the campaign mostly just supposed to be a long tutorial that guides you through all the way to end?
I guess a half-half approach could work. Where you get maybe 6 missions, but each map can be expanded a few times, without really being strictly connected to progression, at least in the later missions. And it would be more about different story events or challenges

[TPRJones]

It may be possible to do a hybrid between types 1 and 2. Like, when the first part is done and red science is running have a "cutscene" sort of thing that results in the player being pushed away from their base for whatever story reason, there is now water in the way of their going back, and they have to start over for green science ... but just over there beyond where they can go the red science region is still existing and running. Perhaps Compilatron can be used to do something simple to give them a leg up on this new base like connecting a single power line to their old factory so they start with power. This happens again when they finish the green science quests, and maybe this time part of the end of that quest line was loading up a bunch of supplies into a cargo wagon which is how they get their leg up on the next part. Again the old factories exist and keep doing what they are able to do unsupervised and the player restarts a new base but without having to start from scratch. And once all the questing is over the waters recede and they regain access to everything they have built. They get to look at how their bases were different and compare what worked and what didn't. They get to salvage or revitalize their prior bases as they see fit. No work is actually lost in the long run but the player was still forced to restart and iterate, which seems to be a key to really learning this game.

[bluerock]

I know its just a dumb little thing for most, but the geek in me got a bit excited when noticing the design for the expanding campaign level areas is so close (but not quite) that of a fibonacci / golden spiral.

[Spiff]

One way to accommodate both new players and veteran players would be to have Compilatron give you advice that you can just ignore. It could monitor your factory's output and say things like "Your copper production is low. You should build more mining drills.", "You don't need to hand-craft red science. Try to automate its production.", or "If you eliminate that biter nest, you will gain access to lots more iron ore. You'll want more than a pistol for that.". Compilatron could even ask if you want a blueprint to build from.

[quyxkh]

veteran players know not to handcraft science packs for 30 minutes while standing still, but forcing a player to discover this by artificially not allowing them to handcraft, lessens the Factorio experience

Right: removing elementary design puzzles from the game reduces its value. Because it's a design puzzle game. People who like to cook don't stop liking fixing scrambled eggs. Even Olympic sprinters like running. They're not always setting world records, most of them aren't doing that at all. They like running. Take away the things people like, the result is something with less to like. To quote Billie Eilish: "duh."

Coming at the same point from another angle: removing bad choices removes choices. Even if there's nothing left but "good" ones, there's no pleasure in making a consequence-free choice, it feels fake. Even if you've made the same consequential choice a thousand times, if it's done right it feels like dancing. Factorio at its best feels like dancing.

[5thHorseman]

You have an infinite map at your disposal. Why not instead of opening specific areas, you unlock successive concentric squares around the player, each rich with its specific resources and whatnot?

Then the player can expand how they want but won't unknowingly go the "wrong" way.

[BlueTemplar]

The NPE used such an heuristic for player strength : pollution !

----

Ah, yes, branching campaigns !



(They can require a lot of work though...)