I just finished streaming my very first Rimworld victory on max difficulty Randy with no savescum in the past week and I'll be typing out my thoughts on the adventure both positive and negative in the hopes that the game can become even better than it already is. So to start off, I'll just state my opinions on many things and then go into detail about them in this post. Keep in mind my opinions will be based off of someone who plays games on the hardest difficulty all the time and tries his hardest to find everything possible to win games. Just recently played about 300 hours of the game in the past 2 weeks to give an idea of how much of the game I've experienced. And to make it clear, I love this game. My focus on harsh input is because I really feel like there are improvements that could allow the game to be even better.
So I believe the game is too easy on Merciless Randy. Raids come by too little, and when they do, they aren't exactly very strong. When I first started playing, the raids were brutal and I'd easily lose all my colonists and get them kidnapped. After learning the mechanics of the game a little, the game suddenly became a cake walk. This says something very important about the game's difficulty; the game is artificially difficult until you learn how to abuse the AI. One of the most popular things to abuse that I discovered while playing is door peeking. I discovered this while trying to deal with manhunters, and then found out from viewers that it can be used on raiders. The raiders upon seeing you will scramble to cover while you get free shots, and by the time they got to cover the door closed and they started wandering out of cover again. This comes with the cost of everything outside of a room/wall getting lit on fire, but its a small price to pay for a perfectly safe colony. Especailly because you can make sure all coolers and power generators are safe behind a wall. AI should definitely react better to that, perhaps they just become determined enough to attack the door until its open despite being shot at or something, but that definitely needs to change. I believe even Warren333 (a big rimworld streamer) feels similarly. During this entire run I lost more colonists to uranium turret friendly fire than raids. In fact, the only 2 people I ever lost to a raid was when low shooting skill enemies ( 2 and 3 skill) head shot through the brain with a bolt action, minus the RNG everything else was easy to handle. During midgame, once the walls were built and we could door strat everything, sieges and sappers would be the next problem. I dealt with sieges and waiting raiders by range. Distance is an incredibly powerful mechanic in rimworld. I know a lot of people talk about how crazy melee is in 1.0 with guns being unable to shoot at point blank, but to force enemies to cross a distance before they can get to you is very powerful. Bolt action rifles outrange everything except sniper rifles. Most siege and waiting raiders have only one or two bolt action guys on their team. I gave everyone I could a bolt action and just set up a firing squad right at the edge of our range and focused down each bolt action guy until none remained and abused the rest of their team. Whenever they ran at us, we backed off until the raiders lost interest. In a game called XCOM (specifically a difficult mod called Long War), one of the most powerful things a player could do was break line of sight with the enemies. Its a problem because the AI just didn't handle it very well. The AI should really learn to handle losing line of sight (this also ties in with the door strats). Especially because all the player had to do was deal with half the enemies before they fleed. On that note, raiders fleeing shouldn't be a condition as simple as half their numbers going down. I once sent 3 people to fight 8 enemies at an outpost. This was when a guy had his brain shot out from safe cover and things looked grim. Another got downed and I had 1 guy left, but I figured all I had to do was kill one more guy. Thus 4 enemies armed with good weapons ran for their lives away from 1 guy trying to stabilize his downed teammate. Perhaps there should be a number proportioned with the amount of downed colonists during a fight to make them run. Back on the topic of the base, to deal with sappers we eventually set up uranium slug turrets and auto cannons around the entire base. I never had enemies breach the walls besides from drop ins after that. I can't say for sure how op uranium slug turrets are just because I've only done one run, but during this run, almost everything was 1 or 2 shot. Including my own power armor colonists lol. I had about 10 of them surrounding my base and was able to replace them when they broken and had plenty of ammo for the 15 days waiting for ship reactor on year 6 to offer some perspective. I'll let the devs decide if that's a good number. Perhaps the costs or capabilities of the turrets need to be adjusted. As for drop in enemies, I put a whole bunch of mini turrets and auto cannons into the hallways and that allowed us to win all our fights there. I personally find the internal drop raids most exciting because there's the most at stake when this happens.
On some other notes, there has been much discussion on how kill boxes has been getting nerfed, and how traps have been getting nerfed. I believe this is a very good thing, and many people might disagree. "Why nerf anything in a single player game" is one of the biggest arguments against reducing op strats in a game. From a development standpoint, its very bad for a game to have one easily abusable strat. newcomers to the game will be quickly turned away when the viable strats are to build a box where enemies die in a maze before reaching your people. "Play your own way" isn't a good argument here because you're literally telling people who want a challenge within the game's boundaries to just not do something that the game obviously allows. If anything the game should just be tuned to a lower difficulty if people wish not to fight raids. Something I think would be cool is if the AI adapted to the players base. If raiders dropped from one side of the map saw no easy way in, they'd make their strat to be sapper instead of just walk in etc. I'm not too knowledgable about how AI strats are determined, but AI adapting to counter your base is a very fair thing. Thus the player is forced to constantly change and evolve his base layout to avoid a campaign stagnating into a resource gathering run. In a popular game called slay the spire, the player builds card decks to clear 3 levels in a rogue like fashion. They could make a quick low cost card deck where they spam attack cards for a win or power up using power cards. The 3 final bosses each countered a type of deck really hard so a player couldn't just make what they wanted, but they had to adapt their deck based on the challenges ahead. Of course a super strong speed deck could still beat the boss that countered that deck, but they'd have to make it really strong etc. Same thing goes for Rimworld, players should be under pressure to adapt their builds to face different incoming threats as time goes on. This is one of the best ways to get past a boring lategame and prevent players from not wanting to finish a campaign. Anyone who wants to be able to relax and do what they want regardless of the game still has the option of a lower difficulty. If they get bored of the difficulty they can change the story teller and difficulty mid way through.
Overall this was a really fun experience. Moments like trying to decide who to use a resurrection serum on or whether or not to take care of a guy with a broken pelvis are great. My favorite part is how this is a story generator that creates a story regardless of victory or loss. I really just wished the end game story wasn't just little small stories of how our colonists got into fights and mentally broke while sitting in the safety of the base. Regardless of how much input from this post is taken, continue making this awesome game better!
hi - you improved a lot and fast! Gz on the successful run.
With respect to ease - well, you are an xcom player and pretty good one at that, that's almost unfair compared to this community. You already have the desire and attitude to learn, and have innate talent at games, I think a player like this should be able to graduate to merciless within 2 weeks (I have maybe 2/3rds of that, so it took me an extra week).
I think that especially on merciless its always going to be an arms race between balance changes and new ways to asymmetrically demolish the AIs in somewhat unintended fashion. The major axis on this fight is tactical wealth destruction, but there are other things in development as well.
Both a congrats to the OP and a long quite agreement with his post.
The "wealth axis" has issues, not just in end game. It is a central mechanic that would be hard to remove. It can certainly be "smoothed" out a bit. As is, it mostly seems to create additional micro tasks, which I don't think adds "fun". I poke at it often, because I think it can be improved. That said, I don't think it is the most important thing.
I suspect success in "balance" is making sure that there is more then one way to play the end game. If the answer is always a killbox plus turrets in your interior, that is fail.
If success requires you to choose a few methods out of turrets, traps, terrain, animal hordes, drugged out berserkers, killboxes, controlled fires, IED's, laboriously crafted masterwork+ gear, clever walls, caches of close combat weapons in your compound, competent micro, allies, terminators, massed mortars, cultivated hives as a distraction, and/or a one shot weapon arsenal and then pray to the RNG Gods it goes your way, that sounds pretty epic to me.
I suspect that is where Ty is aiming. The challenge is making sure all these things are useful tools in end game AND don't break the early/mid game ... while listening to us whine that it isn't realistic or the story we want to experience or the way it was that other patch where I felt so clever after I came up with ... I am not telling because you might nerf it ...
It's a hell of a game.
Quote from: bbqftw on July 30, 2018, 08:54:44 PM
hi - you improved a lot and fast! Gz on the successful run.
With respect to ease - well, you are an xcom player and pretty good one at that, that's almost unfair compared to this community. You already have the desire and attitude to learn, and have innate talent at games, I think a player like this should be able to graduate to merciless within 2 weeks (I have maybe 2/3rds of that, so it took me an extra week).
I think that especially on merciless its always going to be an arms race between balance changes and new ways to asymmetrically demolish the AIs in somewhat unintended fashion. The major axis on this fight is tactical wealth destruction, but there are other things in development as well.
Hey again didn't know you frequented these forums! While perhaps I do have a background in harder games, I do also believe that this game has the potential to cater to players like me. That's what the difficulty should be for I believe. I've only played merciless so I can only speak for this difficulty, but I feel like this is what I'd expect from rough difficulty. I can definitely see how on this difficulty it is an arms race between the balance changes and how much a player is willing to do to win. I do think the AI definitely suffers in a way that can be fixed largely to provide a much more engaging experience to the player rather than the optimal strat be to avoid contact with the enemy ever through outranging/door strating. To be fair trying to fix this can be difficult; if they make a pawn dedicated to destroying a door after getting peeked, the players will find new ways to abuse this behavior. One thing that will become a problem is that since 1.0 is the final supported release, the arms race eventually is going to end with the players finding one solution to getting a 100% winrate in the game and that's where it'll start staling the game once this method starts circulating. Whether its some sort of killbox or some method of abusing AI, as it stands, the method to do this is present in the game.
Quote from: Copperwire on July 30, 2018, 10:29:01 PM
Both a congrats to the OP and a long quite agreement with his post.
The "wealth axis" has issues, not just in end game. It is a central mechanic that would be hard to remove. It can certainly be "smoothed" out a bit. As is, it mostly seems to create additional micro tasks, which I don't think adds "fun". I poke at it often, because I think it can be improved. That said, I don't think it is the most important thing.
I suspect success in "balance" is making sure that there is more then one way to play the end game. If the answer is always a killbox plus turrets in your interior, that is fail.
If success requires you to choose a few methods out of turrets, traps, terrain, animal hordes, drugged out berserkers, killboxes, controlled fires, IED's, laboriously crafted masterwork+ gear, clever walls, caches of close combat weapons in your compound, competent micro, allies, terminators, massed mortars, cultivated hives as a distraction, and/or a one shot weapon arsenal and then pray to the RNG Gods it goes your way, that sounds pretty epic to me.
I suspect that is where Ty is aiming. The challenge is making sure all these things are useful tools in end game AND don't break the early/mid game ... while listening to us whine that it isn't realistic or the story we want to experience or the way it was that other patch where I felt so clever after I came up with ... I am not telling because you might nerf it ...
It's a hell of a game.
One of the things I love a bunch about this game is how raids scale with your wealth. If you're rich and armed to the teeth, you'll draw the attention of stronger pirates, I love that. I'd love to see some pirate strongholds and outposts armed really well also. Currently its mostly just some thugs with a solar panel and a room full of beds. It would be cool to see them with a fortress and auto cannons/uranium slugs down the line. I do agree 100% on the fact that multiple strats should be viable. Killboxes should be a thing, but not the only way. Personally, I think killboxes should work to some degree, but not allow the player to avoid all interaction. Spacing out the traps and requiring replacement is a good start. But the problem is when I was just starting out playing rimworld and getting destroyed, I was making one big mistake: Setting up combat vs enemies. The moment I started avoiding contact and just outranging/peeking doors, the game trivialized so quickly. I think this does show a lot about the game when there is such a huge discrepancy in results due to a difference in playstyle of that magnitude. Even the few times I did decide to meet the enemies in combat, it resulted in brain headshots twice in the colony. I definitely undestand there's difficult balance decisions to make, but one of the first and best places to start is the AI abuse.
Door peeking is absolutely OP and breaks the AI.
And then there are cheesy tactics. Most experence players/ streamers know them all and try not to abuse them, see warren, disnof, rhadamant, xwynns, eviltrick... it often includes no turrets, no traps
If you're proactive on merciless, you should have a few losses. Though i agree mid-late game on 1.0 is 'easy' once you've space marines. The difficulty curve has been nerfed in the balancing process to allow late game mega bases. And there's the ship. If you didn't feel the heat during the ship sequence, for sure your defenses were an overkill.
QuoteAnd then there are cheesy tactics. Most experence players/ streamers know them all and try not to abuse them, see warren, disnof, rhadamant, xwynns, eviltrick... it often includes no turrets, no traps
They are thoroughly unaware of the most degenerate stuff, for both non-combat and combat aspects. And it is quite clear when they do decide to cheese it is typically executed so poorly it doesn't even look OP. This suggests they don't actually know how any of these tactics work.
But their goal is to entertain, not to break the game. So it is not that surprising.
(frankly the level of tech development in rimworld is glacial compared to other strategy games, but its unsurprising since dev is very active in purging optimal strats, and for some reason people accept defeat much more readily in this game. That said, there are certain strategies that are so absurdly broken right now that I am shocked that they are not publicized.)
I'll throw out stuff that was in B18 and yet was severely under-utilized - a quick test in dev mode would show that it was insanely more optimal to rely on 40% intercept shots at long range for most shooters than to rely on direct aiming. But the amount of people that even manually aimed to try to maximize intercept shots.. was practically zero.
@ wanderer_joins well, any successful strategy is going to use a large amount of metagaming right now because the raid system is super wonky, things they probably aren't aware they are cheesing the game.
For instance, most people who play "fair" like those guys are highly selective of their pawns. This is just personal taste in b18, but not so in 1.0. In b18 pawns add a flat 40 raid points. In 1.0 they scale rather quickly to 110 points. Having 40 pawns in b18 isn't a huge deal: 1600 extra raid points is manageable. In 1.0 that's 4400 points, or another 15 centipedes or so on top of your usual wealth.
I used to just recruit everyone because the chaos is fun, but knowing this, I'm shooting every old, chemical fascinated, or depressive on sight.
Do you have a guide for idiots like me for how raid point generation works now?
All I can see is that fun points is practical mirror of wealth, surprised to see that pawn contribution is such a high factor for it.
It looks like that because the raid calculation is a little crazy lol. Your pawn weighting scales with wealth ::)
At 300,000 wealth your pawns are worth 110 raid points. In the beginning they're worth 25 raid points. So wealth is crazy important and so is pawns (but not when you're starting out). So you can recruit everyone you like at first, but make sure to shoot the bad ones in the back in the mid game.
That's pretty much all there is to the raid calculation, aside from animals now also counting a lot. Additionally, at no adaptation, raid points are multiplied by 0.8, and at 60 days (days not being time spent in game, but time between downs/kills which lower the "days") it's multiplied by 1.2, so 50% higher than at 0 days. It also reaches 1.6 at some crazy amount but adaptation does curve off.
Some more interesting metagaming stuff: Installed bionics doesn't count, so make everyone a terminator, and flooring doesn't count, so make your workshop beautfiul with silver flooring rather than sculptures.
Quote from: Greep on July 31, 2018, 02:35:08 AM
In b18 pawns add a flat 40 raid points. In 1.0 they scale rather quickly to 110 points. Having 40 pawns in b18 isn't a huge deal: 1600 extra raid points is manageable.
But in b18 you had no bionic parts...
My first b18 vanilla run had to select 4 guys i would give bionic parts to among 25 (https://imgur.com/a/wfAWF), my first 1.0 vanilla run i had an army of 26 bionic pawns (https://imgur.com/a/3ymHvNF) with charge lances for an end game around 10 000 points. If anything the benefit for adding bionic parts in 1.0 is underrated.
Quote from: Greep on July 31, 2018, 02:35:08 AM
In 1.0 that's 4400 points, or another 15 centipedes or so on top of your usual wealth.
It's a bit disingenuous to present things like that, because the 'on top of your usual wealth' has changed too. Say 40 vs 110 at full wealth would be 1820 more pts, keep in mind wealth is no longer worth 10 pts per 1000, say 6 per 1000 at the stage of a 400k end game colony: 4000 -> 2400 is a 1600 discount....
Roughly points you get for colonists are given off on wealth, net results are roughly the same points with bionic pawns, and threat points following more closely your combat power.
And centipedes are worth 400 instead of 240 iirc.
Well in 1.0 the wealth calculation is complicated compared to b18. There's two curves in 1.0, and a division in b18. in b18, 100 silver is a raid point, and it gets divided by 4 late game, so 400 silver is a raid point. In late 1.0, it's ~250 silver per raid point (900,000 -> 3600), and it gets divided by about about 1.5 (6000->4000) so it ends up being 375 silver per raid point. Not much of a difference, although early game is easier. It's actually kind of interesting that after the dust has settled after a lot of updates, there's basically no difference in b18 raid points per wealth 8)
My point isn't so much b18 is easier/harder than 1.0 so much as knowing how pawns and things like bionics gear etc influence raids is kinda grossly changed and this knowledge dominates just "playing the game well". You mentioned making everyone a cyborg. That itself is seriously gaming the system since a cyborg is no different from a non-cyborg as far as raids are concerned.
But what about armor? That counts fully for raid points actually, so your 5000 wealth power armor is just a huge raid magnet. Might wanna gear yourself mostly with an excellent charge lance and a cloth duster. And maybe make only a few frontliners armored.
I appreciate the great feedback Ronar, thanks a ton. It's useful.
There are degenerate strategies in RimWorld, and there always will be to some degree. I want the game to support some level of skill exercise but at the end of the day it's not a purely skill-focused game; there's a lot of other stuff going on and given the level of complexity and the other design concerns people are always going to find weird ways of gaining advantage. And it's an SP game so if people want to do that it's not a fatal problem.
That said, as Ronar touched on, there is a degree of obviousness/game-breakiness/un-fun-ness whereafter a strat starts to degrade the experience. If a strat is super obvious, not fun to execute, and completely replaces large segments of gameplay, generally I'd like to make the AI less stupid to counter it in an intelligent-looking way, or otherwise rebalance to make it not a be-all-end-all strat any more (though it may still be a useful tool in the toolbox - see killboxes).
But this isn't a binary thing. Whether a strategy is worth addressing on the design/AI side is not a yes/no single-factor decision, it's based on a multi-factor analysis of how much it obscures other gameplay, how obvious it is, how fun it is to do, and other factors.
i played a lot of extrem randy hardcore lone survivor lately as well.. and i can tell you... its no fun. randy is bitch..
the raids are not the most dangerous. most of the time i loose, because of a fucking catatonic breakdown while starving or firefighting. and there are as well the infestations with 60 insects vs my 2 colonists.
i dont think at all its "too easy". especially not in the lone survivor mode..
Re: door peeking, lets just say there is already spicy tech ready to go for some of the more obvious solutions.
Quote from: bbqftw on July 31, 2018, 10:27:06 AM
Re: door peeking, lets just say there is already spicy tech ready to go for some of the more obvious solutions.
What is this spicy tech, specifically?
https://www.youtube.com/watch?v=Z0YIJQ1jgEI
;)