1.0 changing the game too much?

[Tynan]

Wow East, thanks for the videos! I'll review them carefully.

For changing animal zones, did you know you can click and drag on the animals tab to set their zones all at once? Click and hold the mouse button and just drag it.

[Aerial]

I guess I'll speak as someone on the opposite end of the play style spectrum.  I enjoy Rimworld primarily for the colony builder/survival aspects.  I like that there are raids because a colony that is never threatened would require far less infrastructure, but I don't want to play an RTS or turn-based strategy game.   "Door popping", watching raider statuses to see when they reset, etc are all things that sound horribly un-fun to me.  I want to be able to strategically position my pawns and let the battle play out, with pauses only occasionally to give more specific directions (there's an opportunity to flank, focus fire on *that* guy, etc).  So I was perfectly happy with the killbox model because it meant I didn't have to play a game type (RTS/turn-based strategy) that I don't like and am not very good at.

I generally accept that I won't have colonies survive more than about a year because the raids just get too hard for me to manage and colony death spiral inevitably sets in.  Unfortunately, going to easier difficulties to reduce the raid challenge ALSO reduces a lot of the other challenges (like mood penalties, type/frequency of events, etc) which makes the colony building aspects too easy.  So I hover around the middle difficulties, using Randy Random because I like not knowing how the game is going to escalate, exactly, and see how long I can make it.

[East]

of course. tynan. It's only worth less than 20!

The 2-click and 30-drag are not the same.

[Headshotkill]

About raids it might be usefull to see why historically killzones became outdated and switched to lighter temporary defenses.

-Cannons became big enough to blow up forts, so a well executed artillery/mortar barrage would fit this criteria ingame.

-Mobile armored gun platforms in the form of assault guns and tanks shook up the battlefield being able to sustain fire and return devastating counter fire on stationary positions. Some sort of armored tank unit combined with raiders could fill this roll.


The main reason killboxes work still however is simple, the raider ai is unable to recognise a killbox for what it is, I would think a system where raiders remember where most of their team was killed and what killed them could help them but I have no knowledge of coding ai.

[Sheodon]

Regarding the question how to make colonists engaging in combat a great idea, I have got some suggestions:
- Make a wall type that has shooting slits (sry, dont know the correct term). These walls drastically decrease the chance for colonists behind them to be hit, but allow them to shoot within a limited cone in front of the structure. So one could build a defense perimeter in the style of a medieval fortress and defend the base using colonists and weapons against a large number of raiders, while not having to worry too much about everyone loosing their head... Or arms or anything else.
Sounds OP? Then here is the list of drawbacks and counter measures: first of all they dont insulate at all, so you cant just integrate them into your normal wall. Also, there should be blind spots at the sides of the wall, so attackers can come close and destroy the wall, while being untargetable when moving at the right angle. These walls can not be used by turrets.
When players want to raid a fortress with shooting holes, then there are two main options, which i think are both viable: one is as mentioned walking in the blind spots, out of vision of the defenders behind the shooting slits. The other method is using siege equipment as mortars or grenades to destroy the defensiv structures, and most likely every human behind it as well.

- What is also tactical play that can only be pulled off with colonists is kitinig, or running out of the enemies range while shooting him. For this purpose almost everything exists: weapons with long range and ways to imporve the walking speed of colonists, but I would like to have something that slows or damages enemies approaching the base, without giving them cover, as razor wire fences and water or mud lakes.

- The last thing that I wanted to suggest is including an "aim for" feature. I know this is probably quite some work, but it would fot perfectly into the world of Rimworld. The feature could work like this: after researching "combat tactics" colonists that are drafted gain an extra menu point that says "aim for", which is off per default. When clicked a drop down menu opens, giving opportunity to aim for arms, legs, torso or head. Whatever you choose, the chance to hit would go down by some percentage (maybe depending on the shooting skill), but would strongly increase the chance to hit that body part, but would not guarantee it. This would lead to strategies like aiming for the head to kill, aiming for the legs to slow enemies and kite them, or the stop them from fleeing, or aiming for the arms to reduce their damage. Whether the ai should be able to use these tactics and how to balance it I cant really tell, but I think it would be a really fun and tactical feature to add.

Since I am also more into base building and growing my colony peacefully, I would be also glad to have the option to try to convince raiders with a very socially skilled person to go home, or even to join the colony, maybe also giving them a bribe, be it money, animals, resources or even prisoners, colonists or their body parts.

[I Am Testing This Game]

How to make colonists engaging in combat a less bad idea:

Armor with reliable, guaranteed mitigation, so that combat downs colonists before killing them. Very low chance of permanent injury for well armored pawns, essentially zero chance of a squirrel tearing off your toes through power armor. Downed raiders are essentially dead, while downed colonists live to fight another day.

Maybe pro players would still use extreme door cheese, but people who enjoy the gun fights would be allowed to participate in them without extreme negative consequences.

This goes against the story telling idea though, as it means many fights will be consequence free.

There is a fundamental contradiction here. If we're in a consequences-heavy, story telling environment, where any injury can kill or cripple, then we would never fight fair, that's a death sentence. In real life, people don't usually fight fair, everything is based on cheesy exploits to kill the enemy without letting him hit back, because getting hit is just too costly.

If we're making a game where people can fight fair, surviving hundreds of battles, then it needs fewer permanent injuries and becomes less of a storytelling permadeath simulator and more of a "game".

[bbqftw]

I would also point out re: East's videos (the Molotov manip is pretty neat btw) that many of them are essentially against classic/"dumb" raids. These are the freest of raids, basically any non brain dead strat will make them look bad.

The reason the tactics like door peeking mk2 can look so powerful in such contexts has a lot to do with dumb raids not offering any time pressure - if you want to wait until they do something else, they can't really punish you for it. Unless your crops are outside, there is rarely stuff outside they can threaten.

Against drop / sap you have people setting fires or attacking your stuff you have to make the calculation of pawn health vs property damage. Keep in mind that with door peeking you can have 3x covered shooters per door, putting more exposes them to no cover shots - so effectively this is a sacrifice of damage output.  Of course most property can be replaced but a brain shot is not so easily recoverable, so you'll tend to wait and let them do more property damage and whittle them down using doors.

You could increase this pressure by increasing property destruction values, but setting fires and breaking objects would also expose enemy AIs to more aggressive lines and can also be cheesed.

I will also echo that moving pawns in a consistent manner is still extremely difficult. Last night I had two separate instances where I end up with pawns walking into the enemy with no cover since they decide to take a different path than the one originally drawn.

In theory I could probably catch this redrawn path, but when controlling the actions of 5 pawns its easy to slip up. Luckily these were just tribal sappers and not a scyther drop, otherwise I eat a death and not just a down.

It's a nice illustration of unintended consequences, an attempt to kill pawn stacking exploits (I say attempt as under limited circumstances its still possible) managed to make pawn control much more miserable.

[Tynan]

I appreciate the points made all around, some of these are great and non-obvious. Especially about the role of time pressure.

Still love to hear more suggestions about how to make tactical pawn vs pawn combat more attractive.

In thinking about this it occurred to me how restrictive it is to be fighting humans with humans, since rule changes have to be kind of symmetrical. E.g. in DF you fight goblins, so the rules can be different on both sides (and you can't use their clothes).

bbq FYI pawn stacking prevention was to protect the player from 20 stacked manhunting boars, and from combat that's incomprehensible because pawns are stacked and you can't see them. It's not an exploit fix.

[DariusWolfe]

I think the first thought here is to make the raiders smarter in a human-like way so that they're not trivially defeated by such tactics. E.g. if it was a human controlling them, he'd not have them leave the door and attack the wall over and over, he'd probably have them attack the door specifically, right?

Or do players like this strategy? Should it be left alone? After all, consider the topic of this thread... All opinions welcome.

I think, regarding Firestonezz door peeking, the problem is definitely in how the raider reacts. If shooting the door is an option, that's absolutely what I'd have them do; increased "hold" time (i.e. waiting with their weapons 'charged' for the colonists to reopen the doors) would also do a bit to make this a less vital strategy; The longer the battle, the more such other problems rise, such as hunger, sleep deprivation, injured pawns, etc. So increasing the tactical patience of the raiders will do some of the work for you. Especially if the raiders come on to the map well-rested and well-fed.

Regarding an earlier comment, about buffing using colonists in combat/defense: For me, it all comes down to long-term risk. Raiders don't play by the same rules, because they've got an effectively infinite pool of pawns to throw at you. maimed or killed pawns don't hurt them in any fashion, so they can zerg rush you all day long; they'll always have more; I think literally the only way a bad raid can go against them is if the player fucks up a captured pawn (peg legs, removed eyes, etc.) and sends them back, because then they have the potential to be in a future raid. I want to use my pawns in combat, because then it's a matter of my tactics versus theirs; it's just rarely a good idea because any loss or permanent injury I suffer lingers, and can cascade into a total colony loss.

[East]

bonus. https://youtu.be/BGI-THl0iyY
The video is now hard to play because the warm-up is short and the door is closed late.
It's hard to play in real play.
Like the centipede, the minigun and charge blaster will stop shooting when the target disappears. It is a tactic that uses it. It only fires one or two shots, reducing the probability of extreme hit.

[Akyla]

As I mentioned before, I think it would be a good idea to have a way to train your pawns in combat.
Towns can then invest in training up a few pawns as a security team, something Raider towns would not do.
Tactical combat would be more interesting since you would have to protect the other pawns.

I also suggest adding another combat skill that reduces the chance of being hit with ranged weapons (by keeping in cover, by keeping a small profile).
That way trained pawns also have less risk of dying and it would make things more interesting.
Melee vs ranged could then only be successful if the melee pawn has sufficient "ranged Dodge".

[Ilya]

Still love to hear more suggestions about how to make tactical pawn vs pawn combat more attractive.

I find pawn vs pawn combat fun, and I tend to go for it, but I have one major problem with it: it's too unpredictable. The reason why it's unpredictable is that the gap in shooting skill between pawns that are supposedly good at shooting and pawns that are bad at shooting is way too small, in that everyone is too inaccurate, especially at short range. Just think about it; in Rimworld, it's perfectly possible for a pawn who is supposedly a planet-class shooter to miss several shots in a row on a downed elephant that is only a few meters away. This exact thing happened to me many times, but it shouldn't. In real life, not even a beginner would miss like that.

If the gap were wider, it would become more reasonable to specialize pawns into soldiers and workers rather than just draft everyone every time, because professional soldier pawns will have more chances to beat by a raid by themselves, unless the raiders also have very good soldiers. Another problem that inaccuracies cause is that it makes banzai charges from raiders too good, especially early on in the game. A few guys with knives and clubs charge at your pawns across an open field, your pawns miss every single shots at them, and then you either stand your ground and lose, or you do the door strategy and just wait until they turn their back and walk away before you go out and try again. It's lame, but it's necessary. And even if your pawn does manage to hit something, it most likely wont incapacitate the raiders, who can receive a direct shotgun blast and still go on to fight in melee.

On the other hand, you don't want everyone to be too accurate, otherwise all fights will be over in seconds. Maybe that aside from widening the gap in combat skills, all pawns should also be much more accurate at close range and less accurate at longer range. And while we're at it, the chances of suffering permanent injuries and especially of dying are way too high. In actual wars, most of the participants survive and come out without being crippled. They should have less permanent injuries, but also take longer to recover from their wounds, and maybe bleed out a bit faster to make battlefield medicine more important.

[mcduff]

Here's a question. What do raiders want?

It's a story generator. Well, I do the occasional bit of script doctoring for people, and one of the biggest and most common problems I see is stories that don't work because the characters are only doing things because the writer wanted them to.

Too often raids just don't make sense because the raiders don't want anything. They just want to do as much damage as possible until a bunch of them are dead.

Raids are generated because of wealth, to me it makes sense that a lot of raids should happen specifically because they want to steal stuff. The interaction you get in caravans, where they demand a whole bunch of your shit and if you give it to them they go away, makes sense.

Raiders might also want to free a prisoner that you've taken, or take revenge for a previous battle. Their motivation for the attack would dictate their tactics.

For eg, I just had a sapper raid that killed one of my foxes. He was a tough little guy with 3 peg legs (thanks to the A Dog Said... mod) and him going down felt really fucking enraging. What I wanted to do there was go over with a bunch of mortars and just pound their base down. I didn't want their stuff, I just wanted to go teach them a lesson.

However, if I wanted to go raid them for things, I'm not going to take a bunch of incendiary mortars - because that would burn all the things I wanted to steal.

Stories and tactics are both products of motivations. Who wants what and why. Without that, you just end up with massive crowds of raiders grenading their own people into kibble.

[Rahzix]

I think a large part of what makes pawn vs pawn tactical combat fun/not fun is the tedium involved especially when dealing with larger numbers. There is currently no way to issue meaningful orders to more than one pawn at a time. At best you can select a group of pawns and tell them to move to one spot in a huge blob. In practice, once your blob of pawns gets to that spot you still want to select each pawn individually to move them behind cover.

The more tactical you want to be the more tedious it becomes to manage. When engaging an enemy you probably want your snipers positioned in the back with your melees up front and your assault rifle/LMG wielding pawns somewhere in between. This is already a little bit of a hassle to set up. It really becomes a problem when the enemy manages to break your formation. Lets say your front line gets bum rushed by melees and they have to retreat or accept heavy losses. That would be fine if you could easily issue a few commands and your squad would fall back while maintaining some semblance of the formation you set up. In the game as it exists right now though you end up selecting individual pawns and making a lot of minor adjustments (possibly pausing a lot) and it really breaks up the flow of the combat.

If you want people to embrace tactical combat then you have to give them some powerful tools for managing it. Being able to assign pawns to squads and have them move in a set formation when issued a command as a group would go a long way towards making larger scale battles more fun and less tedious. I have played with a few mods that add some of this functionality to great effect. Defensive Positions in particular is a mod that comes to mind.

TLDR: Micro-managing a lot of pawns in a combat situation is not fun to me and I would like to see tools added for the player that make managing multiple pawns in combat easier.

[Robc]

Still love to hear more suggestions about how to make tactical pawn vs pawn combat more attractive.

Here is an option worthy of some thought rather than immediate dismissal.  Ideally I would make this an optional combat strategy selectable in the options menu, not required, as it will not be to the liking of a reasonable number of your players.   I think it would please enough players though that it is worth considering.  I would certainly enjoy it better than the current game play if the AI was of at least reasonable quality, it allows me to concentrate on preparing my fighting force rather than micro'ing the combat itself.  I understand that having two different battle modes is not practical from a development resource standpoint but I'll share the idea anyway in the spirit of your request for thoughts.

Take the micromanagement out of the players hands.  When a raid arrives ask the player to create a defensive zone or to use a previously created defensive zone.    Ask the player to select which pawns will participate in the defense and what mortars or other equipment is available to them.    Are you sure?  Start.  Let the AI play both sides until the battle is resolved.  The player would be free to manipulate pawns not chosen for combat for collecting wounded, evacuating from combat areas, or normal colony duties, etc. but not to attack with those pawns. 

If the player is defeated and non-combatant pawns remain alive the play through should stop and the player can designate a new defensive team with the remaining pawns until the colony is lost or the raid is repelled. 

Multiple raids?  Allow multiple defense instances to run simultaneously on the map. 

It would also be nice for players on a fixed interval only to be able to withdraw or add pawns to a particular combat instance and/or change the defensive zone.   Take out the severely wounded and add reinforcements if the initial force is proving inadequate.

This approach guides the player more toward strategic decision making and reduces tactical decision making, the object is to optimize AI defense tactics instead.  With this approach door behaviors cease to be a problem, kill boxes only work if the AI sees value in using them, open field combat will become the norm rather than an exception.  I don't think it will work for mad animals though and I'm not sure about manhunter packs... testing would be required.

One thing to me is certain, high end raids now are of immense power because you are trying to overcome the profound advantage that humans (especially humans with time manipulation, ie a pause button) will always have over any AI algorithm.  If you switch to AI vs AI for the actual combat then high end raid power will have to be brought to a much lower level to balance.

This method will undoubtedly frustrate players especially in the initial development and AI shortcomings during defense will be amazingly clear... but exploitation becomes much harder.  Many problems are resolved and new ones will surface that need to be considered.  It might be worth a trial if for no other reason to just see how it plays out.  That's what testing is for.