Outpost difficulty ratings should be adjusted based on performance statistics

TheShindig
TheShindig Member Posts: 13

Let me start by saying that I love this game and I want other people to be able to love it to.

It seems to me that a lot of frustration comes from raiders not knowing what to expect from the outposts they're choosing to raid. For example, a new player might complete a few "brutal" outposts easily and then die dozens of times in a "dangerous" outpost that's punching way above its weight class. On the other hand, someone might try to step up to "brutal" difficulty and instead find a free XP loot farm.

By adjusting the outposts difficulty rating periodically, based on victories/kills, the readability of the game would be dramatically improved. Most importantly, this would better allow new players to set their expectations. As it stands right now, the difficulties in all 3 categories vary wildly, to the extent that there are "brutal" outposts easier than "normal" ones and vice versa. I think an ideal UX would be that "normal" outposts should be consistently beatable in a couple tries for beginners, "dangerous" outposts should be suitable for people who like a bit of spice, and "brutal" should be for people who really want a no-holds-barred extreme challenge with tons of interweaving traps and guards. If the game can consistently make the 3 difficulties feel distinct, then I don't think nearly as many people will feel cheated by "impossible" outposts.

The only way I can see to really achieve this is by moving outposts up a difficulty if too many people are losing and down a difficulty if too many people are winning. The current system by quantity and variety of traps could still be used for a provisional difficulty rating, but without some kind of statistical adjustments it will always get too many wrong.

Comments

  • MadMoeZel
    MadMoeZel Member Posts: 685

    how do people who quit measure into the calculation?

    how heavily do old stats affect it's placement, will editing the outpost reset the difficulty, if so what stops people from just changing a single block constantly to difficulty juggle?

    how do we tell if people who quit did so because of difficulty or bias against something like the builder using deadmans switch to kill the HRV or someone who is just hunting tombs for loot?

    do co-op players killing one another count towards difficulty changes?

    ----

    the issue of readability is real but it comes from the allowed variance in difficulty levels. i can make a dungeon say Brutal at less than 1500 capacity, but my Brutal can go to about 6300. that variance created a huge problem when trying to find consistent play. in the lower difficulties that variance is much smaller. for example you won't see a normal that has over 2000 capacity in traps. so that means from 1-2000 could potentially be normal, compared to the 4800 that could be from the bottom to top of brutal and you get a lot of confusion on what outpost should be how hard. we really need to add 2 more difficulty tiers. in addition to that i've been talking for a while about how it would be nice for builders to have some kind of way to flag their creation so people can find specific types of maps. i've suggested doing this via the accolade system where a builder selects the type of accolades they wish their outpost to earn based on their build, so a killbox would be after probably brutal, fun, ingenious, so they would mark that. then a raider would say i'm looking for a brutal difficulty map with brutal fun intentions. and that map would be more likely to come up. then the raiders accolades given over time could shift the map to another search category if the builder trolled/lied and say put a killbox in the "artistic" section when it's just a brick of death. this would allow people who played the map to help regulate it to the proper section of play while also allowing the builder to declare intention and find their target audience.

  • TheShindig
    TheShindig Member Posts: 13

    The weight of quitting vs dying would need to be approximated experimentally.

    Editing the outpost would not reset the difficulty rating. If you've made it easier or harder, the stats will sort that out soon enough.

    To address issues of taste, I would give quits lower weight if they were preceded by less deaths. To me this would indicate a lack of enjoyment rather than a struggle. It would still slightly increase the difficulty rating, but in practice I don't think any outpost would be adjusted all the way to "brutal" just because people don't like it. Accolades could also be a factor in this.

    I would disregard co-op play entirely for these assessments.


    The exact weights could be tweaked and tuned, or they could only address the most severe outliers. For example, a "brutal" outpost with 100% victory rate and no kills, or a "normal" outpost with a 10% victory rate and a ton of kills.

    Bottom line, there will always be people trying to game the system for whatever reason, but statistical evaluation will be vastly more accurate than any hard-and-fast formula like they use currently.

  • MadMoeZel
    MadMoeZel Member Posts: 685

    what do we do about people thinking they should be in difficult when they should be in normal? how likely are players like that to flip a difficult into brutal. how do we handle the speed runners playing actually difficult dungeons and completing them quickly with few deaths? will this shift things into normal and punish newer players?

  • TheShindig
    TheShindig Member Posts: 13

    So, your first question is actually why I'm suggesting this. I think some new players raided a few outposts that were only given the "brutal" designation because they had a lot of traps, then they thought they were good, so when they played outposts that were actually brutal they felt that they had been cheated in some way. I believe that by letting new players get utterly destroyed by a brutal outpost right off the bat we would be setting expectations properly and avoiding bitterness and confusion later on.

    If an outpost is consistently being aced by speedrunners, it might actually deserve to be adjusted down to dangerous. There is a deeper answer to this question that also takes into account the skill ratings of the people doing the speedrunning. For example, if a master aces an outpost it doesn't mean much because it isn't very surprising. If bronze ranks are routinely breezing in an out of an outpost, it might belong in the normal category.

    To answer your last question, imagine the difficulties of outposts as being on something like a bell curve. In the case of a bunch of aces, it should be easier for a brutal outpost to be reassessed down to dangerous than for a dangerous outpost to be reassessed down to normal. On the flip side, it wouldn't take many deaths to push a normal outpost into dangerous, but it would take some serious carnage to push an outpost up to brutal. This way the 3 difficulties should feel very distinct, and I believe that will help new players learn the ropes faster.

  • MadMoeZel
    MadMoeZel Member Posts: 685

    at what point of lethality does a dungeon stop being considered normal or dangerous?

  • TheShindig
    TheShindig Member Posts: 13

    Once there's some sort of (probably hidden) underlying difficulty score in place, the devs could put the cutoffs at 1st and 3rd quartile or whatever they think makes most sense. The important thing is that we wouldn't have outposts 2 entire difficulty levels away from where they should be.

  • Seraphor
    Seraphor Member Posts: 8,707
    edited May 2023

    While I'd love the difficulty calculation to somehow actually account for the difficulty of the outpost, rather than just the amount of traps. I can't see using performance metrics working out. You have no control over what raiders do, they can ragequit when they're missing the obvious strategy, or just because they don't like the format of your outpost, some can take 40 minutes and 30 deaths, while the next person speedruns it in 30 seconds. And with only about 5 raids per cycle, you don't have enough data there to derive averages.

    As long as difficulty categories exist for raiders to choose from, there will be bias in the selection process, and a reason to tailor the difficulty of your outpost to a given difficulty. As it stands, you can fill it right up to the line, get it to the cusp of Dangerous/Brutal and delete one trap, to keep it in the Dangerous category while still having some lethality. Because the moment it ticks over into Brutal, your raids are halved and everyone speedruns it. You don't want your outpost just over the line into Brutal, because you're wasting so much difficulty allowance, as soon as it does, you want to invest everything into traps and pile as much lethality into it as your capacity will allow. A max capacity outpost that has just barely made Brutal is a lost cause.

    So, if you were to make a Dangerous outpost, then come back to see that RNG has pushed it into Brutal... how irritating would it be to get screwed over like that? You'd have to be constantly monitoring your outposts, adjusting them as raiders complete/die to it.

  • MadMoeZel
    MadMoeZel Member Posts: 685

    Seraphor is bringing a good point which is that the wide bracket on brutal where it's from 1500-6300 is gonna still cause issues with readability in brutal. it also rasies the question in me, if i as a builder want to make a normal, but idiots die a ton and it becomes brutal, your system won't re-calculate until people have played, so i could have a 1 trap brutal map getting no raids because it's in brutal. getting bumped brutal by rage quitters would more or less equate in outpost raid death.

  • TheShindig
    TheShindig Member Posts: 13
    edited May 2023

    Monitoring and adjusting outposts is the fun part of building, so I hope you're doing that anyway.

    The current way the game works with its hard cutoffs encourages builders to push every outpost right up to the limit of its difficulty class. This is actually limiting creativity. Personally I would much rather see a game where people just build whatever experiences they're excited about, and let the difficulty fall where it may. If you truly do intend for an outpost to be easier, you literally have to watch replays to see what raiders are messing up. It's an iterative process. That's the only way.


    Regarding the outpost with a single trap making it all the way to brutal, this is completely absurd. It would never happen. If you don't believe me, we could even construct rigorous simulations with the game in its current state. You could literally make an outpost with a single trap, then hide as many other traps as you need away from the harvester path to adjust the difficulty up and down. We could use a simple kill ratio cutoff as a proof of concept. If it gets above to 2.0 (which will never happen) hide more traps to move it up to dangerous. If the kill ratio falls back down (which it would immediately) adjust it back to normal.


    The game needs a statistical sanity check for fixing outliers. There are way too many outposts that are only in "brutal" because they are big. This is frustrating to builders who just want to make something fancy and moderately challenging, and it's confusing to new raiders who are getting mixed signals about how hard brutal outposts should be.

  • MadMoeZel
    MadMoeZel Member Posts: 685
    edited May 2023

    "Regarding the outpost with a single trap making it all the way to brutal, this is completely absurd."

    you misunderstood me. i said what happens if you make a map you personally think is normal, and while you're offline sleeping or working, it gets enough murder to push it brutal. it can't be lowered back to difficult or normal without more plays. but to lower the lethality and adjust it back down you'd have to reduce the trap count. to quickly move it down you'd have to have no lethality, so you'd minimize your trap count to get back to the bracket you intended. but now you're in brutal, which has less total raids statistically, so you can't get your dungeon back down. death by brutal regulation.


    We could use a simple kill ratio cutoff as a proof of concept. If it gets above to 2.0

    how do you handle this?


    i'm intentionally providing an exterme example because we need to consider them if we are to make a truly balance system. i'm not against a reorganizing of difficulty. just want to make sure we're fully investigating it's effects before it gets 'mis-implemented'

  • Seraphor
    Seraphor Member Posts: 8,707
    edited May 2023

    Personally I would much rather see a game where people just build whatever experiences they're excited about, and let the difficulty fall where it may.

    I agree, but this isn't possible as long as:

    1. Raiders can choose their difficulty.

    2. Outposts get far too few raids with which to calculate actual difficulty.


    If instead of 3 difficulty tiers, we had some kind of matchmaking spectrum, like your outpost difficulty is scored from 0 to 2000, and raiders 'skill' is scored similarly, with raiders being allocated outposts within their range +/- 200 or something along those lines. If it was a hidden mechanic that couldn't be gamed, maybe then you could allow dynamic difficulty calculations.


    Monitoring and adjusting outposts is the fun part of building, so I hope you're doing that anyway.

    After every prestige cycle yes. Not after every raid, I don't play 24/7.

  • Seraphor
    Seraphor Member Posts: 8,707

    To back this up a little. I've found it very common that my outposts initially get a high kill ratio, around 5-7, for the first couple prestiges, then it levels out at around 3-4 as it gains prestige levels.

    So this would likely result in most outposts shooting up in difficulty initially, then dropping down at higher prestige levels.

    This is the opposite of what should be happening, as you gain more capacity for traps and refine your design.

  • Samannoyed
    Samannoyed Member Posts: 6

    I think you make a good point, I previously thought and posted about bases having another difficulty added to them within their current categories based on kd. I completely get where you are coming at when you can have a brutal base anywhere from 1500 to 4500 capacity.

  • Seraphor
    Seraphor Member Posts: 8,707
    edited May 2023

    There definitely needs to be some work to refine the difficulty rating system either way.

    One thing I think they should do, is add an additional multiplier for same traps within proximity of other traps of the same type.

    There's already a multiplier for trap proximity in general, but I mean if you stack dozens of corrosive cubes together, or line a room with sentinels.

    This would make "kill boxes" more costly in terms of difficulty rating, making them more likely to get into Brutal, and high Brutal at that.

    Then the threshold for Brutal could be raised, allowing more leeway for high capacity Dangerous outposts that are built more sparsely.

  • MadMoeZel
    MadMoeZel Member Posts: 685

    "One thing I think they should do, is add an additional multiplier for same traps within proximity of other traps of the same type."

    I can't quite figure out why, i'll have to think on it more but this doesn't sit right with me


    "There's already a multiplier for trap proximity in general, but I mean if you stack dozens of corrosive cubes together, or line a room with sentinels."

    "This would make "kill boxes" more costly in terms of difficulty rating, making them more likely to get into Brutal, and high Brutal at that."

    I guess i just need to accept that i play in a way that what people call "killbox" in dangerous or normal doesn't register to me. i've never seen a "killbox" outside of brutal, the games difficulty calculations more or less force it to be brutal if there are enough overlapping angles. i think the most threats i was able to fit in a single room even trying to game the difficulty lower was 9 enforcers with maybe 12 or 15 acid cubes scattered around a room that was like 9x9x9

    thats 24 threats, over half of which have no form of targeting and are stationary. that's a "killbox" in dangerous. after playing brutal dungons and dealing with 20+ things that TARGET me in a single room, i'm just not phased by things like the dangerous "killbox" if it really is an issue of stacking too many things in a single space then i guess sure, yeah increase the difficulty for duplicates more. but i already find it immensely difficult to make a normal dungeon that i think will be a challenge, fun, and artistic without it pushing over to dangerous OR leaving me with a disgusting amount of empty space which is boring.


    "Then the threshold for Brutal could be raised, allowing more leeway for high capacity Dangerous outposts that are built more sparsely."

    Raising the threshold for Brutal will help with the confusion caused by the wide margins on capacity, but it would then increase the availability for killboxes in lower difficulties, because any outpost that was in the low range of brutal before would be considered dangerous after that.

  • Seraphor
    Seraphor Member Posts: 8,707
    edited May 2023

    i've never seen a "killbox" outside of brutal,


    Raising the threshold for Brutal will help with the confusion caused by the wide margins on capacity, but it would then increase the availability for killboxes in lower difficulties

    This is why I suggest these changes together. Individually they don't serve much of a purpose.

    • There's currently little need to push killboxes into brutal, most of them are already there.
    • Raising the threshold for Brutal would allow more killboxes to slip into Dangerous.

    But change both at the same time, and you get more generous ratings for non-killboxes, while keeping the killboxes where they should be.


    Oh and for another suggestion; reduce the impact of HRV path length. It's largely inconsequential for actual difficulty, killboxes often have short paths, while longer paths tend to be easier because the traps are more spread out, but a long HRV path seems to be pretty impactful for difficulty rating.

    My current outpost is on the cusp, and adding one wedge block so that HRV skirted around a corner a little wider than if the block wasn't there, was enough to push it over into Brutal.

  • MadMoeZel
    MadMoeZel Member Posts: 685

    did that edge block maybe change the los of some traps so they overlap?

  • Seraphor
    Seraphor Member Posts: 8,707

    Nowhere near any traps, it was on the outside of the outpost.

  • TheShindig
    TheShindig Member Posts: 13

    The actual function wouldn't use kill ratio directly. It would be much more appropriate to weigh kills asymptotically. For instance, 20 deaths already indicates the raider is struggling significantly. If they decide to stick around and die another 1000 times, there would be no reason to increase the weight of their run much further.

  • MadMoeZel
    MadMoeZel Member Posts: 685

    anything routinely getting 20 deaths is most definitely brutal, you'll have no argument here. but where do we put the lower breakpoint?

  • TheShindig
    TheShindig Member Posts: 13

    As a general vibe to target, I'd say anything bronze ranks are consistently beating in less than 3 deaths can go to the normal category.

    As a side note: I just found out some players die repeatedly on purpose in loot farm levels to break the traps multiple times. So that would need to be addressed by the devs in some other way, otherwise their difficulty will always be overrated. I don't think that's a huge problem though, because farms seem to be a relatively small % of all maps, and people know what they're getting from the thumbnail, so they're kind of outside the scope of the problem I'm interested in.

  • MadMoeZel
    MadMoeZel Member Posts: 685
  • hastarkis
    hastarkis Member, Alpha Surveyor Posts: 580

    As a general vibe to target, I'd say anything bronze ranks

    I must add that's pointless because every season everybody becomes bronze ranks and have to climb again. Deep in the core that's a battlepass for rewards, not a rating system.

    Of course experienced people climb faster but - still - they have to climb. And while they are climbing they are "bronze", "silver" and so on which will affect this system in a very wrong way at the start of every new season (at the very least).

    Also, it is possible (and pretty much intended) to just grind your way into masters if you play enough. And, in addition, it is possible to be great raider with bronze rank (social raids don't give rank points). How can you take them into account along with skilled masters and inexperienced "bronze" players and make it accurate? It's an actual meaningless mess.

  • TheShindig
    TheShindig Member Posts: 13

    Well, yeah, I would have preferred the ranks to be a skill rating too, but ok, fair point. How about if most players beat the outpost in less than 3 deaths?

    The devs can set the exact cutoffs however they want, as long as it does a better job of sorting the outposts than the current system, and that's setting the bar pretty low.

  • Viktoriusiii
    Viktoriusiii Member Posts: 36

    dont bother. I tried to tell them noths ago.

    But most ppl on here and on reddit are hardcore fans that do not grasp the concept of balancing... so the devs wont fix anything, no matter how much the steamrating drops. (its at 60% now :D)

    Sadly... I was VERY hopeful at first because they had the feedback thing in the beta... but it feels like they just threw that in to pretend like they gave a crap. The main problems are not adressed whatsoever:

    Map difficulty calculation

    Map rewards (instead of only kills, give rewards with accolades)

    Harvey as the only way to verify maps

    As long as these are not fixed, I can not see this games reviews getting better again.

  • MadMoeZel
    MadMoeZel Member Posts: 685

    the 3 death count is a bad metric in a game that promotes dying to learn. it needs to be set higher. i think having the max penalty for deaths being at 3 is what promoted quitting and moving on to new stuff, the sharp decline in reward early on discourages persistence and rewards moving on. persistence will lead to higher skill player base, where as moving on i 'guess' promotes the high octane 'fun' player base. but incorporating both will allow for dynamic raiders

  • noggieB
    noggieB Alpha Surveyor Posts: 63

    Just gonna agree that no stats metric is going to be perfect, but most of the suggestions made here would be better than what we have at the moment! Devs can of course tweak the metric over time.

    I think the current calculation is a good starting point, but then metrics could bump the outpost up or down one tier.

    One concern raised by Loys in https://forums.bhvr.com/Meet-Your-Maker/discussion/364123/be-friendly-to-newbies-save-them-from-mix-maxxed-normal-outposts was that adjusting the difficulty of an outpost would burn the builder who had thought it was in a different tier. Although I guess that mostly bothers min-max builders (who deserve it!) and isn't really a problem if all tiers are getting raided equally anyway.

    Another thought is that Normal, Difficult, Brutal could be different for different players. I.e. a low ELO player would see fewer truly brutal outposts, and a high ELO player would see more. This would also allow the devs to distribute outposts to get an equal number of raids, regardless of their calculated tier.

    Having three distinct tiers is somewhat artificial anyway, since it's clearly a spectrum from easy to murderous.

    Finally, do we even need tiers? How about instead, at the end of a raid, the raider can press "Too easy", "Too difficult" or "Just right" and then the game will give them the next raid, without them even needing to go back to the Sanctuary?

    TLDR: Yes, keep the existing rating, but update it slightly with raid metrics, to catch outliers.

  • MadMoeZel
    MadMoeZel Member Posts: 685

    "One concern raised by Loys in https://forums.bhvr.com/Meet-Your-Maker/discussion/364123/be-friendly-to-newbies-save-them-from-mix-maxxed-normal-outposts was that adjusting the difficulty of an outpost would burn the builder who had thought it was in a different tier. Although I guess that mostly bothers min-max builders (who deserve it!) and isn't really a problem if all tiers are getting raided equally anyway."


    why does a min-max builder deserve it?

    do raiders not min-max? are they somehow different? not a big fan of double standards, please do explain why builders desrve to be bothered by having their stats adjusted after hours of building and meticulous calculation to full utilize the tools they are given WITHIN A POINT RESTRICTION. why is it that someone with wit should be punished on the builder side? do co-op players not raid in co-op to reduce difficulty and get access to that juicy infinite revive so they don't reset traps? both sides are equally guilty and everything the builders do is in RESPONSE to raiders actions. they pull ammo through the walls, we have to multi layer, they kill ALL the guards, we have to make ammo traps (and one of them just got removed by the devs cause of crying), builders are constantly forced to adapt and have things removed from them. raiders haven't had a single removal since release. all the bug fixes, make the game easier for the raider, not a single bug fix that helps the builder, like the fact that ammo can be picked up and reloaded through solid objects, or that you can revive a player through the walls and floors.


    so why does the min-maxing builder deserve to be bothered by having their outpost they spent hours on suddenly change difficulty?