http://dbd.game/killswitch
Regression event limits criticism
Hello Devs,
I come to you today to speak on the regression event limits that we've had for awhile now. These limits by themselves are not bad, however, the way it builds up on a generator is, when affected by multiple perks.
While these perks are doing their intended tasks, its also rapidly building the regression events which will eventually block out the generator as other survivors come back to repair it and I try to bring it back down a bit.
An example of this is if I tried Pain res and Eruption and triggered both in a situation that would cause 30% potential loss in progress if timed correctly, however, it also adds 2 separate regression events to a generator that is being focused by the survivor team and if repeated, mainly by eruption, can isolate the gen from the killer altogether, not forgetting that kicking a gen for 5% is counted towards this tally as well.
Kicks are almost not worth it, 5% loss is not worth giving it another event towards being blocked in this current state of the game. I'm not calling for a buff but maybe some change towards how it counts towards regression events.
With survivor repair speed perks being a constant and just general teamwork of focusing a generator, I don't see how it would be a bad thing if a killer could at least rely on a kick to start regression a few more times outside of perks, when eventually you shut off perks affecting them altogether on the killer side.
I get no one wants to be stuck in a game trading and even getting caught in loops with killers over a generator, its not fun, however, seeing a generator become isolated and basically theirs for the taking is not fun either due to the current system.
Seeing it improved upon would be amazing and I think the blocking is too much for doing a basic killer action with kicking while survivors can do healing and repairing to a much higher degree of repetition over the course of a game. You want to prevent perks from being looped, good. Preventing a killer from just slightly damaging a generator is negative to the overall game play loop.
it would also be nice to see perks outside of the meta pain res get buffed or new perks added that cause significant damage to gens since they will do nothing at all as the game goes on.
As always a hand clasp from across the miles,
Triglav
Comments
-
Essentially TLDR is that basic non perk kicks shouldn't count towards the kick limit. I agree. They do very little and yeah it does hurt Eruption unfairly.
-3 -
Agreed
-3 -
I agree that procing a regression event should count as 1 regression event
-2 -
Just put on the irruption, kick it, and explode it with Surge three times already. I don't want that obnoxious sound to be generated by the beeping about 3 minutes into the match. Is it abnormal gameplay to not let the Gen advance in killer play in the first place?
-1 -
My issue with the 8 regression limit is that it unfairly punishes weak m1 killers with no mobility.
Killers with no mobility cannot afford to travel very far to gens at the outer edges of the map. Survivors also know this, so survivors leave those gens last. That means the only targets (survivors) available are at the center of the map AND the killer can't afford the lack of pressure to travel to non-central gens anyway. That means the no mobility killers are forced to stay at the center which means the only gens they will be interacting with ARE those center gens. This means they are FAR more likely to be unfairly hit with the 8 regression limit thanks to survivors hammering out those center gens - especially if perks like surge are used hitting multiple gens or if eruption hits which takes up TWO regression hits. This doesn't affect killers with mobility because they can easily cross the map. This means they can patrol ALL gens, and considering their mobility, they can guard those center gens far easier which means survivors will want to spread out to draw the killer away from the center.
Long story short. The 8 regression limit really hurts no mobility killers with no rangefar more often than it hurts a couple of problematic killers who are good at 3 gen camping. They need to make the 8 regression limit only begin once there's only 3 gens left to ensure it's only affecting 3 gen campers and stop hurting weak m1 killers who weren't even 3 gen camping.
-3 -
This content has been removed.
-
I agree it does hurt the already weak killers way more while basically not bothering the already strong ones.
My issue is the double standard. Like a 3 gen lock down is too strong supposedly because it puts the survivors in a very hard to win scenario, but what about when the survivors split the gens with the last 3 on opposite sides of the map? Unless you're one of the few high mobility killers this is basically an auto loss for the killer. Is that not too strong and very hard to win in the survivors favor? What's the difference between that and the 3 gen in favor of the killer? I'd even argue that pre gen kick limit the survivors chances of winning a 3 gen are better than a killer winning with the gens on opposite sides of the map. Like significantly.
I don't mind the weakening of 3 genning, I just want balance consistency between both outliers and it feels like there's a double standard.
0 -
Yep, that or keep it the way it is.
-2 -
An example of this is if I tried Pain res and Eruption and triggered both in a situation that would cause 30% potential loss in progress if timed correctly, however, it also adds 2 separate regression events to a generator that is being focused by the survivor team and if repeated, mainly by eruption, can isolate the gen from the killer altogether, not forgetting that kicking a gen for 5% is counted towards this tally as well.
If you can shave 110% (35 * 2 from pain res eruption and kick, 40 from pain res for a total of 8 events) of a generator off with the regression events youre given and you still need more to keep the gen from popping, you need to work on your micro to take chases short enough to win games with. You don't need the devs breaking a system that has largely worked to forestall 3 genning outside of one notable exception to give you even more opportunities.
2 -
What's the difference between that and the 3 gen in favor of the killer?
The difference is that the "power role" (killer) gets to choose which one they prefer. Yes, as a low-mobility killer, you might have to protect a central 3-gen and let the outer ones go. You might not get a hook at all until a gen is done, but once you get them centralized, you win, so…
1 -
I don't see how this can be true, personally.
Eight regression events per generator is high enough that you're not going to hit it on the gens you're interacting with regardless of which gens they are, unless you're focusing really hard on stalling generators by constantly kicking them yourself.
The more immobile, weaker killers don't have the time to spare messing with generators over doing something important, so if anything, they're less likely to see this system fire. To me it seems like the system affects mobile and immobile killers equally- which is to say, it only affects players that focus on damaging generators over progressing their objective or practising good macro play.
-1 -
I found 8 kicks allready so much that it almost never comes into play and killer is free to destroy gen progress so much the limit is almost pointless. I would lower it honestly to 7/6 max.
-4 -
Its not a give or take here. Adding gen tapping that was proven to be detrimental to the game because I suggested more kicks to be allowed before the block kicks in is not even close to a fair adjustment.
-5 -
Fair question, so here is an example. I play a lot of survivor, and it's quite easy to weaponize the whole system to permanently lock a central gen within the first 3-5 minutes of a match. The moment the killer leaves, you simply repair it beyond auto regress line. Either the killer doesn't touch the gen (you win the gen) or he kicks it (you get a free locked up gen within minutes).
0 -
I've seen this video before, and the third option is "the killer actually prioritises the survivors they know are on this gen".
This isn't an example of the system being badly designed, it's an example of bad play accidentally tripping the system. It's also exactly what I'm referring to when I say it affects players that heavily prioritise attacking generators over doing something important.
The argument could be made that a system should be designed to not accidentally punish bad players too if that isn't their intention, and I'd be sympathetic to that, but I don't think this specifically is particularly problematic. It just teaches players not to focus on generators over their objective.
2 -
The problem here is the killer doesn't know that there's a survivor nearby the gen. He's literally running off to chase survivors he thought was on the gen. The answer is for him to just sit and camp the gen and don't do chases? Thats clearly even more problematic.
If he stops to kick the gen, he gets gen locked. If he chases survivors off, he loses the gen by followup repair survivors. It's a guaranteed lose lose situation that any half competent survivors can use against the killer early in the match.
In the end, bad play or not, this shows the system is absolutely affecting killers who are not 3 gen camping despite BHVRs claims that it would never happen.
0 -
If they kick the generator, and when they check again they're capable of kicking the generator another time, they know someone came back to undo the regression. In this particular video, the correct play - and the difference between a mediocre player and a good one - would be to send the guard out on patrol to catch whoever observably keeps coming back to repair.
For other killers without a tool like that, you'd expect them to at least check the section you're hiding in, that's a no brainer.
To the last part, it's worth mentioning that BHVR's phrasing wasn't that killers who are not 3 gen camping would never be affected. Their phrasing is that eight regression limits aren't reached in the average match and players who aren't excessively stalling the match (their actual target, 3-genning being the most egregious example but not the only one) wouldn't see it.
Which, most people aren't. I can count on the fingers of one hand how many times I've even seen the warning spikes pop up in either killer or survivor games, and I've never seen a gen be actually blocked.
Hearing that it shouldn't affect the average match doesn't mean there's something wrong if it is affecting you, necessarily. It could also mean you're playing in the way they wanted to discourage.
2 -
8 events rarely ever even happen in a game on the same gen. Its totally fine the way it is.
-1 -
so since we have an 8 kick limit can we have an 8 touch limit nion survivor or more likely an 8 heal limit on survivors?
-2 -
Other skull merchants kick gens to slow the game
I kick gens to zoom with Machine Learning :(
1 -
Okay, how about this. The survivors are blatantly making mistakes at the beginning of the match. Within 20 seconds, a gen has already reached almost HALF of it's allocated regression hits - if other gens were being worked on, this would have affected THREE gens. Due to the survivors being out of position and blatantly making mistakes, the killer is punished heavily. Naturally, all of these survivors had unbreakable, so this only netted in a single hook. (sorry, external linking is broken in the forum, so will have to copy and paste)
and here's a match where one of the center gens is locked up despite giving chases and not just sitting at a 3 gen camping. The moment ghostie chases a survivor off it, 2 more survivors come swarming to immediately do repairs. The option is to give chase and lose the gen, or just sit at the gen and camp. Both options will result in a failed match. You'll notice that any time ghostie leaves the center of the map, his pressure drops to absolutely nothing, and the center gens get a LOT of progress.
EDIT - Better example - not sick in this one
It's a lose-lose for m1 killers with no mobility with the 8 regression limit. They can't afford to leave the center without losing a lot of pressure, and you'll also notice the survivors aren't even bothering to work on the far away gens - just throwing everything at the center. This is using only ONE regression perk and the WEAKEST one, at that. With eruption, this would have blocked even faster.
Again, this is clearly an issue specifically for m1 killers with no mobility. For killers with mobility, they can patrol the gens (ESPECIALLY the center gens) so fast, survivors naturally need to spread out and apply pressure globally across all gens. For killers without mobility, this need to spread out isn't needed, so survivors can just safely stay pressuring the middle of the map easily. This means the weakest killers in the game specifically are far more susceptible to the 8 regression limit than all other killers - even if a 3 gen camp isn't happening. Those weak m1 no mobility killers don't need the extra weakness. Instead, the 8 regression limit should only occur when 3 gen camping is happening - ie they aren't chasing or hooking - just sitting and camping.
Please note, I'm not asking to remove something to stop 3 gen camping (which is indeed lame). I'm just asking for it not to affect anyone not actually camping. I don't think that's an unfair request. Either move the 8 regression limit to starting when there's 3 gens left, or have the counter reduced but reset when someone is hooked…because if someone is being chased and hooked - clearly they aren't just sitting at 3 gens refusing to push the objective which is what the 8 limit was supposed to stop.
Post edited by RpTheHotrod on1 -
8 should be 10 at very least. Or buff gen regression
-1 -
For the record, I am not contesting that less mobile killers are likely to stick around the centre gens, or whatever gens are actively being worked on to force a gen split if it's not literally the centre. I agree with that part, that part makes perfect sense to me.
What I'm contesting is that the situation described above means the killer needs to kick generators that often. I've played M1 killers, I've played them in scenarios like this, and I've still never seen the regression limit system kick in. It implies that you're actually attacking generators that often, which my position on is that it's a bad play if you're an immobile killer, you don't have time to be paying attention to the generators. It's not a luxury that you can afford if you can't make up for that time with mobility or fast chases afterwards.
I watched the youtube video - not following reddit links, apologies - and there are two broad points to make about it.
First, despite what the video is titled, you did not lose this match because of the regression limit, you lost because you didn't play very well and were outplayed at several points. That's not me dismissing this video outright, to be clear, I still have points to make, and it's also not me trying to insult you- especially if you were sick and on some kinda cold meds, nobody's gonna judge you for losing a match of DBD, it happens.
The second is that even accounting for this, you still stopped to kick generators thoughtlessly with a perk equipped that removes any necessity to do that. While it's not 3-gen camping, it is - perhaps accidentally - still what BHVR were attempting to address with the regression limit, which is excessive generator stall.
It's not about whether or not you're camping, it's about how much generators should be stalled within a match. At the end of the day, you still need to actually progress your objective and play your macro game, not just stall out your opponent's. You specifically may not have been attempting to grind the match fully to a halt… but you were still trying to rely on attrition, and if that worked, then someone who is trying to grind the match to a halt would be able to do it. That's why this system wasn't just exclusively designed around 3-gen camping, that was just the main problem that necessitated it.
The absolute best you could conclude from that video is… it's possible to accidentally trip the limit because of poor play. That's a pretty reasonable edge case, in my opinion.
0 -
Always good to have a discussion with you, Jester.
I already mentioned in the original post with the video that I played poorly - I was very sick and heavily medicated, haha (you can see me entirely stopping mid-match on multiple occasions to take a quick "nose check". The quality of the match vs me being well is vastly different. My apologies for having such a piss poor video example while I was very sick (I'm cringing on some of the decisions I made in that match), but it was simply the video I had on hand. I'll see if I can hunt down a "legitimate" match. EDIT - I found a better video where I wasn't sick. I updated the original post. Now granted, I do some gen kicks, but "well, just don't kick the gens!" is a silly solution because not regressing gens simply means they'll just be finished instead of locked but not yet finished.
That being said, even if I never stopped to kick, surge alone would have still caused the same result. You'll notice several surges didn't do anything due to the gen being locked up, so even without stopping to kick, it would have happened. Also the title of the video of course is absolutely debatable, and I played poorly being sick enough that I'd probably lose regardless, but that's not the issue - the issue is that the video shows that m1 killers are getting gens locked up unfairly against them in situations where 3 gen camping is not happening. Regardless, I'd argue that "the killer kicks a gen 3 times, he should be punished!" is not a good way to balance the game. If kicks are so evil and terrible, just remove them from the game entirely and readjust balance accordingly. Kicks "existing" shouldn't be the cause for regression limits. Killers trolling by camping a 3 gen the entire camp and refusing to do chases and hooks should be the cause for regression limits (ie artificially extending the match).
No worries on reddit links - I totally understand. To describe the situation, at the beginning of the match, 3 survivors were caught off guard. I hooked one, and injured 2 in the process. They went for the save, and I downed the rescuer. This triggered a SURGE hit. The person who just got unhooked instead of running simply dipped around the corner then stopped because they wanted to do a flashlight save. I instead simply turned the corner and downed the flashlight attempter. Second SURGE hit. I then go to pick up the first downed survivor (I prefer not double-hooking survivors), but I spot the THIRD survivor just outside the room healing themselves due to A Nurse's Calling. I down them. That was 3 SURGE hits all in a spam of 20 seconds at the very beginning of the match. Why should I be punished for the entirety of the rest of the match of having nearly half of my allocated gen regressions already evaporated just because the survivors are throwing themselves in harm's way? 8 regression limits isn't sustainable if you can literally lose 3 in 20 seconds. I get your point that it's probably bad that m1 killers stop to kick and it's a bad call, and you're probably absolutely correct. However, the regression limit can still easily kick in with things such as SURGE where the killer is NOT making mistakes and merely progressing the objective by downing people (and hooking after, of course) such as the example I provided above. I wouldn't say the killer downing 3 people in 20 seconds is the killer making "bad calls". Regardless, it affects the gen limit WORSE than kicking!
I'll have to disagree the match was about attrition. You can clearly see that I had entirely left the center area on multiple occasions to chase a survivor, but there's a LOT of safe loops on that map, especially when the majority of the time they just ran to killer shack.
Anyway, the point is the 8 regression limit sucks for killers with no mobility - especially ones without range. They give up ALL of their pressure every time they chase. That means they are far more vulnerable to gen rushes. They typically aren't even allowed the time to stop and kick gens which is why surge is a popular perk among them. Even surge alone though can trigger the 8 regression limit against killers who are already designed to be weak.
As you mentioned, this can happen due to poor play. Okay, so if a killer plays poorly, let's kick them in the gut while they are down by forcing their gens to lock up? They are already playing poorly and likely are going to lose - why make it worst for them? I don't think the concept of "8 regression limits only only affect killers who don't play perfectly" is a good end result. Killers and both survivors are allowed to make mistakes in the game.
My whole point is simple - let's stop having the anti-3 gen camping mechanic affecting killers who aren't 3 gen camping. I don't think that's an unreasonable ask. It's supposed to stop people who, per BHVR's own statements, camp the gens the entire game with no intention on progressing the objective (ie chases and hooking). If this system ever kicks in where that isn't happening, then it's flawed and needs to be re-thought.
In case anyone needs a reminder, here is BHVR's actual words of the intent of the 8 regression limit:
"…3 gen scenarios are those where the killer aims to defend 3 generators that are close by in particular, and does not participate in normal gameplay - not chasing survivors, etc, in this instance survivors are not able to make any meaningful progress to those generators."
"…In the overwhelming majority of games, this will not come into play. However, in scenarios where the Killer is defending a set of Generators and refusing to chase Survivors, there will eventually come a point where they can no longer damage the Generator…"
Clearly in my examples, this never happened, and yet the system still kicks in - mostly against weak no mobility m1 killers. I've watched a lot of streamers, a lot of yt videos, and have had a lot of matches, and I've very often seen at least one generator lock up when no one was stalling the game. In fact, most of their normal matches where there isn't a lockup only lasts 10-15 minutes. How is 10-15 minute matches considered "excessively long matches"?
In the end, I simply want us to stop punishing m1 killers with no mobility simply for existing. "We're adding in an 8 regression limit to combat trolls, but if you're an m1 killer with no mobility, you're also likely to get punished too just because."
Post edited by RpTheHotrod on1 -
I personally think, that it should not be regression events by rather totall regression amount.
Like, gen is blocked for regression if it's total regression reaches 270 charges, or 180, or anything.
Some perks would acknowledge this like adding additional regression before block occurs, or allow gen to regress below 0…
1 -
The fix to this is simple and should've been the obvious choice to begin with: only have the kick limit start when the last 3 gens are the 3 closest gens to one another or within a set radius.
Surely there are other things to consider but this just seems like it always was no-brainer.
-1 -
I feel like this is a separate perk issue more than anything.
More specifically, Surge, Eruption and non-regression gen kick perks.
0 -
@jesterkind Okay, thanks for your patience. I went and and pulled the video showing how charges can be lost very quickly out of the reddit link. Pardon sound quality, but video from video caused degradation. This is only 22 seconds long, so not a long watch.
The end result after all of them having unbreakable was a single hook. You'll see here that at the beginning of the match, the killer is penalized for survivor mistakes by having one of the gens get nearly half of its charged lost in a matter of 10 seconds. This isn't at the fault of the killer. This isn't the killer making "bad plays". Imagine if the survivors had progress on more gens? That's 3 MAYBE 4 gens within surge range…all would have had nearly half of their allotted 8 regression limits evaporated within 10 seconds. The killer is then significantly punished for this for the remainder of the match, as those gens will perma-lock up FAST.
Now remember, the whole intent of the 8 regression limit is to try to shut down trolls:
BHVR quotes:
"…3 gen scenarios are those where the killer aims to defend 3 generators that are close by in particular, anddoes not participate in normal gameplay - not chasing survivors, etc, in this instance survivors are not able to make any meaningful progress to those generators.""…In the overwhelming majority of games, this will
notcome into play. However, in scenarios where the Killer is defending a set of Generatorsand refusing to chase Survivors, there will eventually come a point where they can no longer damage the Generator…"This means it's specifically designed to shut down killers essentially holding the game hostage by defending and refusing to participate in chases\downs\hooks.
So essentially, I've provided a variety of ways that the 8 regression lockup can occur. The first video earlier (knight video) had shown where survivors can weaponize the system to permanently lock a gen within minutes of the match beginning. Your suggestion was the knight simply not regress the gen, but then the gen is simply lost (it's a vital gen). Lose-Lose situation.
The second video (I replaced it) showed you were a killer making bad plays can trigger permanent lockdown even if not 3 gen camping and running around chasing and hooking.
The third video (replaced a few comments above) shows a standard match where killer and survivors just play normally - I'm sure both making mistakes occasionally, but that's literally part of the game.
The fourth video here (the 22 seconds one) shows that by zero fault of the killer, survivors can make mistakes which can eliminate nearly half the charges of multiple gens within 10 seconds.
On average, ALL of my matches are 10-15 minutes before and after the 8 regression limit, so the whole "it prevented an extended match from happening" isn't valid.
That's a mix of bad killer plays, neutral killer and survivor plays, and bad survivor plays - all of them triggering unfairly on regression hits as no 3-gen camping was occurring.
In the end, this means that the 8 regression limit is ABSOLUTELY affecting games where there is no 3-gen camping going on, at all, and when it happens, it's MOSTLY happening due to m1 killers being forced by the survivors to focus in one small area of the map due to their inability to quickly move across large distances. One might say, "Well it doesn't happen enough to warrant a fix", but I see this happening on streams very commonly, and besides, the last thing weak m1 killers with no mobility needs are to be unfairly harmed by a system that's designed to punish trolls.
-1 -
Sorry about the late response, yesterday was a busy one for me. Happy to get back into things now!
So, there's a couple things I want to clarify about my position as we move forward.
First, I really want to hammer home one big thing - when you say that a killer playing badly shouldn't be harmed more by a generator being locked up, I want to make it clear that the only thing I'm talking about, in general, is that if you play badly your chances of winning are lower. I'm not saying a player should be outright punished for playing badly with a heavy blow, just that if you repeatedly make mistakes, the game will get harder in ways that are specific to those mistakes.
Before this system, overly kicking/damaging gens made the match slightly harder to win because you were giving survivors too much time to make distance and/or evade you. Gen kicks aren't and never have been actually impactful, after all.
After the system, it's broadly the same but there are some edge cases where you might also get a gen locked up midway into the match. I'm not saying those edge cases are super niche, one of them is just having Surge equipped, but it's also not just kicking the gens that causes it, there has to be more going on.
Second, I think it's important that we take a step back and remember what the punishment we're discussing actually is. If you play poorly by focusing on stalling gens, you don't just automatically lose, that one generator becomes impossible for you to continue damaging. Considering that your job as killer is not to stop the gens from being completed, they're merely a timer for you to complete your actual job, and considering that the bulk of a killer's slowdown broadly comes from survivors being too occupied to work on gens to begin with, that isn't actually that big of a deal for any player that isn't trying to stall the gens out.
It can be a relatively noticeable deal, there are centre gens where you get a harder endgame if they're completed early, but even that isn't unwinnable. You have to make a play in the endgame and broadly forgo regressing the last gen, but it's salvageable. I think this is worth mentioning because we're not talking about making one mistake (or a series of instances of the same mistake, even) and that's it, game over. We're talking about that mistake leading to a very specific unfavoured position that you can still come back from. It's just that you don't come back from it by continuing to focus on generators.
Don't get me wrong, the gen locking up IS a gigantic deal if you're trying to stall, 3-gen or otherwise, this system does work for the cases we both agree it should, but if you aren't… at worst it's a knock and something you adapt to, it isn't inherently game-losing.
Third, I do find the video you posted interesting, because you're immediately put into a favourable position - less favourable than outright winning because of the Unbreakable, but still favoured - and you really don't have any reason to continue attacking that generator. If it does get locked up, are you even really affected? It's one generator in this clip, and while I'd grant that you'd notice the situation more if it were more than one… you've still put the survivors into a hard reset, pressing that advantage means they're not even doing that much gen repair to begin with.
That's the broad pattern, and that's where my ultimate position is. Unless you view stopping the generators from being done as a fundamental necessity for the killer role, even the worst case scenario of the gen actually reaching the eighth event limit is like… just sort of that gen no longer being a resource for you after a delay, right? It's still a place survivors congregate, it still counts as one of their objectives that you can prevent them from being able to repair, but you no longer get to kick it.
Fourth and probably finally unless I remember something while I'm typing, I do want to concede something. I went and looked at the dev update in question again, and I'll admit that I remembered the phrasing being more concrete that it actually is. The devs DO specifically frame the system as being more broad than JUST 3-gen camping, having it in brackets as an "in particular", but what I recalled as "excessive stall" was phrased instead as "excessively long matches".
I still think that's relatively ambiguous and definitely not "this system should only affect the most extreme forms of 3-gen camping", but even if it isn't, I would maintain that the extra scenarios this system does affect aren't a problem. You're affected a little if you fall into the bad habit of excessive gen kicking, but it isn't game over unless you're also only excessively kicking those gens. It is, at worst, a soft pressure to learn better habits and improve your gameplay, because you just lose the ability to keep doing that habit after enough attention.
At best I'd say Surge makes that habit particularly prone to backfiring. I feel like even with the bad habit of excessively kicking generators, if you're not running Surge, you're not losing events even close to as much.
1 -
"First, I really want to hammer home one big thing - when you say that a killer playing badly shouldn't be harmed more by a generator being locked up, I want to make it clear that the only thing I'm talking about, in general, is that if you play badly your chances of winning are lower. I'm not saying a player should be outright punished for playing badly with a heavy blow, just that if you repeatedly make mistakes, the game will get harder in ways that are specific to those mistakes."
100% agree
"Before this system, overly kicking/damaging gens made the match slightly harder to win because you were giving survivors too much time to make distance and/or evade you. Gen kicks aren't and never have been actually impactful, after all."
GENerally agree (no pun intended…okay maybe a little). However, for gens survivors aren't paying attention to, natural regression is somewhat helpful - and kicking gens can trigger useful perks such as nowhere to hide or Trail of Torment.
"After the system, it's broadly the same but there are some edge cases where you might also get a gen locked up midway into the match. I'm not saying those edge cases are super niche, one of them is just having Surge equipped, but it's also not just kicking the gens that causes it, there has to be more going on."
A fair statement. Regardless, my point is that it happens, and happens worst to M1 killers. A system designed to punish trolls shouldn't be punishing legitimate gameplay (whether that gameplay is effective or poor gameplay).
"Second, I think it's important that we take a step back and remember what the punishment we're discussing actually is. If you play poorly by focusing on stalling gens, you don't just automatically lose, that one generator becomes impossible for you to continue damaging. Considering that your job as killer is not to stop the gens from being completed, they're merely a timer for you to complete your actual job, and considering that the bulk of a killer's slowdown broadly comes from survivors being too occupied to work on gens to begin with, that isn't actually that big of a deal for any player that isn't trying to stall the gens out."
According to BHVR, it's not a timer. See their comments - it's specifically designed to stop killers from holding the game hostage by refusing to participate in normal gameplay such as chases and such. The 8 regression system is for anti-troll. How, it DOES have a secondary benefit that if a match is just taking too long overall, it will naturally start taking place over time - we're talking this is great when a match starts getting 30+ minutes long. A happy side effect. This is a good thing!
"It can be a relatively noticeable deal, there are centre gens where you get a harder endgame if they're completed early, but even that isn't unwinnable. You have to make a play in the endgame and broadly forgo regressing the last gen, but it's salvageable. I think this is worth mentioning because we're not talking about making one mistake (or a series of instances of the same mistake, even) and that's it, game over. We're talking about that mistake leading to a very specific unfavoured position that you can still come back from. It's just that you don't come back from it by continuing to focus on generators."
Can't disagree, but it's rather off topic. The discussion here is how the 8 regression system has a much higher likelyhood of unfairly affecting weak m1 killers with no map mobility when that killer is not 3-gen camping and refusing to participate in normal gameplay per BHVR's definition of the intent of the system.
"Don't get me wrong, the gen locking up IS a gigantic deal if you're trying to stall, 3-gen or otherwise, this system does work for the cases we both agree it should, but if you aren't… at worst it's a knock and something you adapt to, it isn't inherently game-losing."
It can be game losing depending on the circumstance, but that's far too much of a variable situation to say it absolutely will be or absolutely never will be game-defining. Losing a central gen in a map like RPD can absolutely make or break the match, especially for m1 killers who can't effectively patrol the very far gens. Once the center is broken in RPD, the odds swing heavily towards the survivors. There's a reason lobby is usually the first to get focused on.
"Third, I do find the video you posted interesting, because you're immediately put into a favourable position - less favourable than outright winning because of the Unbreakable, but still favoured - and you really don't have any reason to continue attacking that generator. If it does get locked up, are you even really affected? It's one generator in this clip, and while I'd grant that you'd notice the situation more if it were more than one… you've still put the survivors into a hard reset, pressing that advantage means they're not even doing that much gen repair to begin with."
I'm put into a favorable position due to survivor mistakes at first glance, aye. However, the end result is that I get one hook at the cost of nearly half of that generator reaching it's allocated charges within a mere 10 seconds. The survivors are indeed pushed into a hard reset, but once the dust settles, we are essentially back to square one where the survivors made no gen progress to hard reset, all the survivors are healed back up, and the only positive thing I came out on top of was one hook but at the cost of the endgame being significantly weakened with one gen getting locked up by the 3 minute mark. That gen was unable to be regressed all the way to the end of the match. That is a long term penalty that's intended to punish killers that hold the game hostage. Regardless of who has the upperhand or not, the point remains - the system is unfairly affecting killers who are NOT holding the game hostage, and survivor mistakes themselves (or intentionally!) can trigger the system itself. This simply shouldn't be. You ask if the inability to regress a gen would affect me. What if I was using kick-perks like Nowhere to Hide? That can be a game changing perk - remember…there are multiple reasons to kick a gen when perks are involved - it's not just to regress. Also, try playing an m1 killer and never regressing gens from the beginning of the match. You'll find really quick that regression does actually play a role in matches and are not "unimportant".
"That's the broad pattern, and that's where my ultimate position is. Unless you view stopping the generators from being done as a fundamental necessity for the killer role, even the worst case scenario of the gen actually reaching the eighth event limit is like… just sort of that gen no longer being a resource for you after a delay, right? It's still a place survivors congregate, it still counts as one of their objectives that you can prevent them from being able to repair, but you no longer get to kick it."
I don't feel stopping the generators are necessary (though obviously a big help, especially with certain perks). I actually don't kick gens all that often (my original gameplay you saw I was very sick and on meds, so I just didn't care to not kick, ha). That's typically why I run surge, so I don't need to stop and kick. As an m1 killer, you've gotta be putting CONSTANT chase pressure, and stopping to kick gens halts that pressure. That being said, part of the game balance is INDEED having the ability to regress gens, and it was that way for years, but due to some bad apple killers, a punishment for trolls was engaged. Whether kicks are mandatory or optional doesn't matter - again, the point simply is this system to prevent killers from being held hostage is being unfairly forced on killers - especially weak m1 killers with no mobility - where no hostage taking is happening.
"Fourth and probably finally unless I remember something while I'm typing, I do want to concede something. I went and looked at the dev update in question again, and I'll admit that I remembered the phrasing being more concrete that it actually is. The devs DO specifically frame the system as being more broad than JUST 3-gen camping, having it in brackets as an "in particular", but what I recalled as "excessive stall" was phrased instead as "excessively long matches"."
Memory is silly that way. Yay for getting older! Jokes aside, in all of the examples, none of these matches were approaching "excessively long matches", so the system should have never kicked in - ESPECIALLY kicking in after only 3 or 4 minutes (knight video) of the match starting.
"At best I'd say Surge makes that habit particularly prone to backfiring. I feel like even with the bad habit of excessively kicking generators, if you're not running Surge, you're not losing events even close to as much."
We shouldn't be associating killers who happened to one to run Surge as trolls. Originally, Surge had a cooldown. However, the devs LOWERED the damage output on Surge in return for a cooldown removal. Their reasoning? To allow the perk to PUNISH SURVIVORS who are out of position. However, with the 8 regression system, that "compensation" for the damage (no cooldown) now can actually help the survivor since it can rapid fire trigger multiple regression events at once. "Well that's what you get for running Surge - the game has flagged you as a troll." isn't something I think we should even remotely consider.
My ultimate suggestion is this:
REDUCE the regression maximum down from 8 to 4 "tokens". Have it ONLY affect gens that are in close proximity to each other. Have the tokens refill after someone is hooked - as if the killer has chased, downed, and hooked someone, they clearly are not refusing to participate in normal gameplay of chases and hooks.
This would allow the system to kick in MUCH faster against people who are just defending gens and not chase\hooking. This is a survivor buff. However, if the killer is chasing and hooking, then the anti-troll system won't kick in against them, at all. It's a win win.
and let's not forget, there ARE good reasons to kick gens - lots of kick related perks that are not related to gen regression. This whole system that can affect everyone really harms being able to use those perks. Before, you could always rely on kicking a gen to put your TR on it and you be undetectable. With the 8 regression limit system, that absolutely can be a point where ALL of your kick perks just simply no longer function for the rest of the match. That sucks.
Always a pleasure having a respectable discussion.
-2 -
This content has been removed.
-
This content has been removed.
