The second iteration of 2v8 is now LIVE - find out more information here: https://forums.bhvr.com/dead-by-daylight/kb/articles/480-2v8-developer-update

Does Anyone Have Video of a Killer Using 8 Gen Regression Events on a Single Gen in a Normal Game?

CodeDB
CodeDB Member Posts: 285

Obviously the 3 Gen change currently being tested in the PTB is a pretty big change. One of the major concerns I've been hearing have been players worrying that 8 gen regression events on a gen is too little. Personally, 8 seemed to be on the high side to me so I decided to check out some streamers to see and from everything I've seen, it would seem I'm right.

Just to illustrate the example, I actually re-watched and counted the events in a few random streams. It takes quite a bit of time to rewatch a stream so I tried to get some solid examples with players I knew are good killer players (Hens being one of the best all around players I know of and Carniveris being more of a specialized killer main as she mainly plays Artist though was doing a random perk streak at the time). I also tried to choose some recentish weekend dates as generally weekends mean more players online and better chance for MMR to actually work (questionable logic to trust MMR works at all, I know). Obviously, this isn't a definitive conclusion but I was shocked at how few actual gen regression events occurred.

Here were the stats:

All of the Gen Regression Events counted were total for the match, not per gen. I started counting it by gen but quickly realized 1. Pain Res made it harder to know which gen was hit and 2. It didn't really matter to count per gen because the numbers never came close. I also decided to track if a specific perk/add-on contributed to the gen regression event in case that could swing the numbers. I also included results and map just in case there was some correlation between result and number of gen regression events but I think it is inconclusive at best.

I had a few issues with my sample, mainly being a lack of Wraith (who can be quite the gen kicking machine) and lack of perk variety (mainly Jolt/Surge not coming up in any of the games) but I couldn't really find a good way to control for that.

So that was why I wanted to ask for others inputs, does anyone have gameplay evidence of 8 gen regression events being too little? Anyone have any recommendations of other streamers I should check out to get a better more holistic view on the issue?

«1

Answers

  • Chaosrider
    Chaosrider Member Posts: 489
    edited January 13

    But 8 Events are too much !!! Its affects normal gameplay !!!!!

  • LordGlint
    LordGlint Member Posts: 8,559

    Unfortunately this isn't something you can really test on PTB unless you do KYF. Bring a killer like Hag on PTB to try 3-genning and ppl will just DC since it's not Billy, Blight or Onryo.

  • baharuto48
    baharuto48 Member Posts: 124

    Players who want to take a game hostage. Players. A person. They'll still try their darndest.

  • Xyvielia
    Xyvielia Member Posts: 2,418

    0K = 2

    1K = 6

    2K = 2

    3K = 9

    4K = 11

    When you look at the outcome of 30 matches, the 3Ks and 4Ks between the 2 players are still dominant.

    It’s almost like theater or drama class was a prerequisite to playing DBD for most of the playerbase

  • Stroggz
    Stroggz Member Posts: 500

    Does oppression counts as an event on all generators effected or just one that is kicked?

  • crogers271
    crogers271 Member Posts: 1,842

    Now someone might say, "Oh, you won, so who cares?" Yeah, I won because the Survivors were making some serious blunders and also didn't have tool boxes or gen rushing perks. On any other match, this could have easily been a loss. They were more interested in playing altruistic than actually doing the objectives. They could have easily knocked out that center gen if they actually tried, and with the other gens spread out like they were, that may have been walking miles as an m1 killer with no map traversal power.

    But if they were actually gen rushing, would you have ever had time to get to 8?

    That's the loop everyone has been going in since the 3 gen mechanic was announced. As a killer to trigger it through normal gameplay requires the survivors to play really poorly, in which case the killer has the game in hand already.

  • Ayodam
    Ayodam Member Posts: 3,146

    Can you all explain to me why these killer mains insist that survivors are somehow in a good spot if the killer is kicking the same gen 8 times? It just seems like the survivors are throwing the game from my perspective, and I can’t imagine losing a game as any killer under such a circumstance. But it appears to be a major concern for the killer mains here and on Reddit. Why?

  • Firellius
    Firellius Member Posts: 4,429


    There's this insistence that any anti-X mechanism must not benefit survivors under any circumstance outside of the absolute most egregious examples of X. This is in disregard of whether this'll make survivors actually win, or indeed, if it is even beneficial. Bit of an issue of scope, I think.

    Reminds me of this discussion on 'pausing the hook timer if the killer is nearby', and someone was insistent that survivors would abuse this by standing on opposite sides of the zone limit, and if the killer starts chase with one, the other would go for the unhook. In that scenario, the paused timer would benefit the survivors, compared to that scenario without the hook timer pause.

    But that didn't take into account that without the paused hook timer, that scenario wouldn't play out, and those two survivors would instead both be on gens making progress and forcing the killer to spend more time searching, so the very existence of the scenario is a killer boon in the first place.

  • crogers271
    crogers271 Member Posts: 1,842

    I think the concerns have dropped a little over the past few days at least. I'd give four possibilities that I don't think are unique to killers:

    1: People don't like math. This is why discussions tended to dead end when people actually tried to lay the scenarios where all of the regression events would be eaten up. Conceptually it was easy to say 'well, if I have eruption I need only this many downs' without actually looking at how much time would be spent between activations.

    2: People treat new game mechanics like they've violated a set of sacred rules. This was like the AFC, even if triggers once by accident every couple of months, its not like the killer's account gets banned and the game is immediately forfeit. As much as we say we want new things, people become apocalyptic when a new thing is proposed.

    3: Even before the change was announced, many killers had gotten the mindset that the 3 gen mechanic was coming to take away their ability to play the game. It's hard to get rid of a mindset if you've already embraced it, even if objectively the change with the increase kick damage / no gen tapping is a buff.

    4: People take BHVR statements way too literally. BHVR said they wanted this to only impact Killers who 3 genned from the start. Many seem to have treated this as a vow and not a design goal.

    I will give credit to @RpTheHotrod - he at least provided a video as requested by the OP and noted it was unusual. I was beginning to think we weren't going to get any instances of it happening.

  • Chiky
    Chiky Member Posts: 785

    Using a sample of pro players is not the same as using a sample of general players. It's just a really bad sample, especially using just 2 players. It's like League of Legends balancing the champions based on the pros....

    With my oreo build I kick gens all the times. I havent counted, but definitely more than 20 kicks per game. Maybe way more than that. And even with the rest of killers, the number of kicks is high too. Not sure if a kick has been kicked 8 times tho (not in PTE, and not when doing a kick a gen X times with the help of a survivor), so I can't really tell if that's acatually a thing.

    If the change was just kicking 8 times turns the kicking regression useless, I'd say it'd be completely busted to the survivor side, but since it has a nice regression and also renders the gen tapping useless, I think it's all good keeping it on 8. Of course I would need to test it, but it sounds fair.

  • RpTheHotrod
    RpTheHotrod Member Posts: 1,934
    edited January 15

    Not necessarily a win. They could have been kicking my ass by also hitting other gens. They had an incredibly lucky killer shack right next to 3 center gens on top of having the center building available for looping. In this example, sure, the survivors totally screwed up, but only because they were an swf obsessed with wanting to do altruistic swarms. If they actually tried, the match could have gone very different, and I would have been completely locked out from one of the 3 center gens. That's absolutely being punished all in the name of "stopping the killer from just sitting at the 3 gens and never doing chases from the beginning of the match" as BHVR claims is ONLY when this would kick in and "normal non camp matches will never reach 8 regression events". That is completely incorrect.

  • Coffeecrashing
    Coffeecrashing Member Posts: 3,794

    The concerns dropped because the popular concern was never really valid. People aren’t often doing 8+ regression events on the same generator, in normal gameplay.

    The real concern should have involved what happens if survivors were literally trying to force 8 regression events, by repairing in short bursts and excessively hiding.

    We have a major double standard, where it’s absolutely believable that killers would go through extreme lengths to win games, but most people think it’s super unbelievable that survivors would be willing to go through extreme lengths to win games.

  • Firellius
    Firellius Member Posts: 4,429

     If they actually tried, the match could have gone very different, and I would have been completely locked out from one of the 3 center gens

    Except you also said that if they actually tried, you wouldn't have hit 8 regression events on the same gen, because the gen would've been done in the meantime. Not to mention that this took quite a long time to reach. And keep in mind that there's regression buffs it's being paired with, so they would have less progress overall, meaning they've wasted more time and are more likely to lose.

  • CodeDB
    CodeDB Member Posts: 285

    My sample games were simply what I could provide and aren't meant to be definitive proof for balancing. I, like you, had a sense of what my personal games have gone like but I have found that simply stating what I, the average or more likely below average player, experiences isn't really notable for this forum. Anyone can say their personal experience so I wanted to find some actual numbers to help demonstrate why I felt a particular way and help facilitate an actual discussion.

    I chose the two players I did because

    1. I knew about them previously and trusted them to know how to play the game
    2. I generally enjoyed their content anyways
    3. They were streaming live content so I knew it was not edited

    All that being said, I think you missed the actual question I posted with this. I specifically asked for others to add their input and give me other game footage to check out to get a more holistic view on the issue. If you have someone who plays in a more similar fashion to you that you think would be beneficial to bring to the discussion, I would love to know about them.

  • CodeDB
    CodeDB Member Posts: 285

    First of all, thank you! Surge/Jolt was exactly what I wanted to see in use and I agree the game was unusual. That center gen must have been up to over 90% with how much the survivors were trying to push it and just couldn't commit.

    That being said, I counted 6 gen kicks and 4 Surge/Jolt procs which were mostly split between the gen near Shack and the gen near the Basecamp (though this map makes it hard to tell with its lack of...well much of anything lol) so I don't even think this game would have used the anti 3-gen mechanic (and nor should it). Gen events could have been a bit closer as you mentioned with the early grab or the hook being a bit closer to the gens but the fact that they weren't I think is a good sign for the mechanic.

    While I thought 8 events would be on the high side, I do think examples like these show why it can't be much lower (if at all) than 8. I'll be really curious the number of Surge/Jolt hits on a similar game on a map like The Game or Hawkins.

  • alpha5
    alpha5 Member Posts: 364

    I had a game recently where the first gen got hit by Pain Res + Pop three times, so six events. With Surge or Eruption on top it would have hit eight easily but that game was kind of a shitshow anyway.

  • GroßusSchmiedus
    GroßusSchmiedus Member Posts: 555

    The system only affects Killers who 3 gen but not the majority of the player base. Sounds like it works like BHVR intended, in fact its more of a buff for most normal Killers who don't 3 gen since gen tapping was removed and instant regression was increased. Now we only need some map/spawn rng changes that would make 3 gens more difficult because there are truly some unholy gen spawns out there.

  • Seraphor
    Seraphor Member Posts: 9,424

    A survivor playing like this is throwing the game or trolling their team mates. This isn't the argument you think it is.

  • Pulsar
    Pulsar Member Posts: 20,792

    Because a battle of attrition usually favors the Killer.

    If Survivors are playing in that way, they'd need to be astronomically better than the Killer for it to be viable. That does happen, of course, and it certainly should be called out as an ######### move if you see it, but I don't see it much differently than I see a flashlight squad.

    Annoying? Yep. Mean-spirited? Sure. Looking to exploit worse players? Yes. A huge issue in the grand scheme of things? No.

  • Seraphor
    Seraphor Member Posts: 9,424

    This is the thing, that many hooks in a short window without any gens completed almost certainly puts the killer in a winning position, or near enough.

    The question isn't "can a killer reach 8 regression events?" It should be "does a killer need more than 8 regression events?" And the answer is almost certainly no.

  • Akumakaji
    Akumakaji Member Posts: 5,468

    With the old gen tapping a survivor could have very easily baited out 3 or so events by just running around a tight loop and 8 might have run out pretty quickly, but with the new gen tapping mechanic its not that easily abused and quite fair game.

    I guess many players are just seeing it from the old perspective, without taking both mechanics into account. With both mechanics in place 8 regression events just puts a hard counter on the whole thing, but only a true, classic 3 gen situation will ever get close to this numbers.

    Just like with the anti camp, it's more the looming threat of running out of events that will make killers commit to chases, then killers actually running out of regressions, and I think that's a good thing.

  • oQuva
    oQuva Member Posts: 8

    With this update you can still play 3-gen. The update with 8 kick events was directed only, for example, to the hard 3-gen skullmerchant (before the rework). In any other case, the killer can still benefit from 3-gening. But if he didnt do anything in 8 kicking events, then he was destined to lose. Before the introduction of this update, you can see either long games that are won by the survivors in the end, or equally long games that are won by the killers. (its 30% surv win - 70% killer win i think? taking into account hard tunnel 1 surv. And these numbers will not change, except maybe a little 10%+ - in favor of the survivors) This mechanic simply speeds up the decision-making process for the killer side, which doesn't fundamentally change the 3-gen meta, it just speeds up the outcome of using this meta. It also helps that 1 kick is 5% instant regression + perks. This will allow faster regress gens (without hook perks like scourge and pop) than before, giving you the time you need to catch a survivor faster. And also survivors cant one tap gen to stop regression. For all 7 years in dbd, Im sure that I used more than 8 kicking events for 1 gen like 100 times. I think this is small compared to the time I spent in the game and such cases are incredibly rare. So 8 is a good number. And most likely it was made for such perks as surge(jolt) or smth, so that you have a reserve for regular kicks in case you have such perks as pop, overcharge etc.

  • RpTheHotrod
    RpTheHotrod Member Posts: 1,934

    Doesn't matter if it puts them into a winning position. If the survivors are making mistakes, they shouldn't be getting freebie catch up mechanics to give them a chance to win. That's would be like making NOED basekit for killer. (Frankly, I don't like NOED at all to begin with).

  • alpha5
    alpha5 Member Posts: 364

    While I agree in theory this game is full of weird 'freebie catch up' rules and mechanics. Eventually protecting a single gen getting bombed like that would not make a difference in the grand scheme of things. I would have liked to DC after the 2nd hook for the outcome was obvious to me. Instead I had to play a lost game while staring into the void thinking I should have gotten out the moment I saw it was Skull Merchant.

  • Seraphor
    Seraphor Member Posts: 9,424
    edited January 15

    You're right it doesn't matter.

    It doesn't matter if the killer is wiping the floor with survivors 10 times over or 9 times over. If they're obliterating survivors, it doesn't matter if survivors get a tiny leg up that will ultimately make no difference.

    Because that's what 4 Pain res + 4 Pops on a single generator is. 4 hooks and between 110% and 220% gen progress undone.

  • CaptainOkita
    CaptainOkita Member Posts: 15

    Not that I encourage this play style, but if you want to win games, you just tunnel people out one by one and then slug the last two. Slugged people don't work on gens. 3 gen was a somewhat viable strategy and I find it way more engaging fighting a 3 gen than being slugged end on end. I'm not sure how I feel about this change yet.

  • Seraphor
    Seraphor Member Posts: 9,424

    3 gen is still a viable strategy if you use the 3 close gens to capitalise on their proximity and actually chase and score downs. Having 3 gens close together means survivors will be forced to reveal themselves.

    3 gens are a problem when the killer refuses to commit to chases and just holds the gens down, preventing either side from progressing.

  • RpTheHotrod
    RpTheHotrod Member Posts: 1,934

    One thing to also keep in mind, I was only using a single regression perk and nearly capped out what would be my allocated 8 regressions within the first few minutes of the game. If I had been running even a single extra regression perk, the 8 would have been breached probably within 3 minutes of the match starting. Again though, this is primarily at the fault of the survivors for huddling around only that gen, but I don't like the fact that all because the survivors make some poor decisions that I should get penalized for their mistakes.

  • Halloulle
    Halloulle Member Posts: 1,347

    I remember one match on Haddonfield as Nemo where I might have kicked one gen eight times. - It was a group of very stubborn survivors who managed to three gen themselves - or, well, I don't think it was really avoidable on their part; I had two gens in Main and the one next to it, so a really tight three gen - and the first person took chase to main and I was too stubborn to drop it. So no one got around to doing those three gens and I doubt at that point during the match they were aware of the second gen in main. At some point they started to try and do the main gens between chase-laps but I could just do a drive-by-kick. I also did have jolt and eruption, so when I eventually did down a surv I got massive value.

    After that it was just a stalemate that didn't seem to end --- but it's still Nemo. Eventually they won, but I did get six or seven hook stakes out of that extremely dragged out match.

    And tbh. I'm glad that even that kinda stuff isn't going to be possible anymore.

  • Krazzik
    Krazzik Member Posts: 2,475

    I think it's pretty unlikely for most killers to ever reach 8 events on a single gen even with multiple gen perks.

    I will say though that Jolt seems to be by far the most impacted perk by this change. Jolt is built around dealing less gen regression but on multiple gens, so is punished much more harshly with the limit, and I think for killers who run Jolt, especially on smaller maps, they very well could hit the limit much more realistically.

    That said, if they're getting a bunch of downs and hitting multiple gens with Jolt multiple times, they're probably doing pretty good that game :p

    (PS I know it's Surge now)

  • crogers271
    crogers271 Member Posts: 1,842

    BHVR claims is ONLY when this would kick in and "normal non camp matches will never reach 8 regression events". That is completely incorrect.

    If the survivors are making mistakes, they shouldn't be getting freebie catch up mechanics to give them a chance to win.

    So these are both going to fall into the list of four things that come up. This isn't just a killer thing, but the way any player can look at the game.

    The first, the 'ONLY', is an unrealistic and overly exacting demand of a game feature. The idea that a game mechanic will never come up, not even in a 100s of games of play, is unrealistic.

    To take the game you showed, we have lots of 'could have' happened things.

    The second is treating the game design as somehow sacred. If survivors get downed by the same group of gens and you have surge and are kicking the gens, you are still winning. There is a extremely rare scenario where they might not get downed enough to give you the complete victory and you get locked out of a gen, but the survivors have still been punished for their mistakes (lots of downs, massive gen regression). It just isn't quite as punished as they once would have been.

    Like survivors could be complaining that killers are being given a free catch up mechanic with double gen kick power. But there's no sacred rule on what the amount of damage a kick should do, its all about building a healthy game.

  • Coffeecrashing
    Coffeecrashing Member Posts: 3,794

    But it really doesn't mean that survivors will be forced to reveal themselves. It only takes 4.5 seconds of repairing (and even less if it's more than 1 person) to reverse regression, and unless the last 3 generators are literally all within line of sight of each other, survivors could just repair in tiny bursts, and be long gone before the killer can reach that generator.

  • crogers271
    crogers271 Member Posts: 1,842

    The real concern should have involved what happens if survivors were literally trying to force 8 regression events, by repairing in short bursts and excessively hiding.

    We have a major double standard, where it’s absolutely believable that killers would go through extreme lengths to win games, but most people think it’s super unbelievable that survivors would be willing to go through extreme lengths to win games.

    But how would that happen? Not talking ethics or game enjoyment, just purely how do people pull this off while going for a win.

    On the double standard, I absolutely agree with the idea that survivors would go to extreme lengths to win a game, but why would these be a viable approach and how is it different than the game right now?

    In the game right now survivors can go to a gen, work on it, and immediately hide as soon as they hear the terror radius / their SWF tells them the killer is coming. They can then come back out and work on the gen after the killer has left.

    Under the current system, the survivors will finish the gens. Under the new system, survivors will finish the gens. Survivor gen speed far outpaces the killers ability to reset it.

    At least the idea of a mass amount of surges in one area is something I could hypothetically see happening. I just think it would be incredibly rare and, even in that event, the killer has probably won the match anyway. I just don't see from a math standpoint how this can be a valid tactic.

    Survivors might do it to bully the killer, just to prove they could, but they could do that anyway right now.

  • Coffeecrashing
    Coffeecrashing Member Posts: 3,794

    Often in the current game, this ends up being an extended 3-gen, because the survivors are pre-leaving generators early and often, and the killer isn't chasing the survivors that are pre-leaving the generators early.

  • Coffeecrashing
    Coffeecrashing Member Posts: 3,794

    It doesn't take a lot of skill to repair generators in short bursts, and excessively pre-leave generators. You could probably code a bot to do that strategy too.

  • Sava18
    Sava18 Member Posts: 2,439

    Killers literally would rather kick a gen than get a free hit most of the time, they have conditioned themselves to do so. Make it 4 per single gen, something to discourage that less than a bot level of thinking.

  • Coffeecrashing
    Coffeecrashing Member Posts: 3,794
    edited January 16

    Both sides of the game should have some encouragement to interact with the other side, or some punishment if they excessively refuse to interact with the other side. This shouldn't be 100% the burden of the killer, to chase survivors, even if the survivors are literally never anywhere near the generators when the killer patrols by them.

    Yes, there are survivors that pre leave generators at the first sight of a visual terror radius, and runs far away and hides until the terror radius goes away, and if it's a low mobility killer that can't hide their terror radius, then this cycle can just keep going and going, where the killer just never finds anyone. And this is sometimes even worse if it's a SWF, and they are using voice comms to relay the killer's location. Why is that strategy acceptable?

  • crogers271
    crogers271 Member Posts: 1,842

    Why is that strategy acceptable?

    1: How would you realistically detect it? Even if we agreed we wanted to prevent pre-running, how would we determine that? Regression events are a measurable thing that is straightforward (though not as straightforward as if they had just done kicks). I don't see a way it could even be done.

    2: Most games usually have the onus on one side to take an action. No one likes the situation where survivors stand outside a 3 gen and the killer stands inside it ala the worst of the chess merchant days. You could say as the power role that onus falls on the killer, or you could just say realistically at this point in the game's development that's the only feasible path.

    3: The increased gen kick time and lack of gen tapping makes the killer stronger in the 3 gen, it just means its not unlimited. This gives the killer a higher chance to catch the survivors, just not as much time to do it.

    4: The strategy you outline can be done in the game right now. The change mercifully shortens it.

    This is the big one for me. As the game exists, survivors can dart in and out on a 3 gen. If the killer does not chase, the killer will lose. It just can take a long time. The killer still loses here, it just doesn't take as long.