We have temporarily disabled The Houndmaster (Bone Chill Event queue) and Baermar Uraz's Ugly Sweater Cosmetic (all queues) due to issues affecting gameplay.

Visit the Kill Switch Master List for more information on these and other current known issues: https://forums.bhvr.com/dead-by-daylight/kb/articles/299-kill-switch-master-list
The Dead by Daylight team would like your feedback in a Player Satisfaction survey.

We encourage you to be as honest as possible in letting us know how you feel about the game. The information and answers provided are anonymous, not shared with any third-party, and will not be used for purposes other than survey analysis.

Access the survey HERE!

Machine Learning was growing on me - until it was nerfed again

Lets get straight to the point: I played quite some time with Machine Learning after the recent buff and it was a lot of fun, because it brought something new to the playing area. Lets face it: this buff was not a buff, but an essential necessity! Before the perk was so unbelievable cumbersome to use, with the two kicks and you really needed to double plan which gen you actually wanted to kick.

So while its now was more easy to plant and plan, it still needed the target gen to get completed, and with 2-3 gens popping targeted and pressured at the same time during the start, its pretty much impossible to get the perk to fire more then 2 times per match.

When it takes effect, the boost of speed and the undetectable are very fun and a good compensation for losing a gen, but the last round of tuning in nerfs really killed the perk for me. You can't plan around the time when the perk activates, and you might be in a really bad position or have nothing around. The 60s gave enough leeway and breathing room to get your bearing and formulate a strategy: maybe you saw a survivor over there, maybe you wanted to stealthily approach a gen from an angle, maybe you were in chase and wanted to quickly end it and to then capitalize on the remaining juice of the perk.

60s gave me all the time needed to play around with my new found speed and undetectable and maybe make one little mistake, but 45s doesn't offer any free space, you either can fully capatilize the moment the perk procs (well, its not programmed randomness, but you get what I mean), or its wasted. And for a perk with such a steep activation condition its really feels like it gives too little for that.

Its a real shame, because I really liked the perk and had planned to integrate it into my new general playstyle, but -15s really killed it and my time of experimentation with Machine Learning is basically over. I really don't get nerfs like this: the idea behind this is to buff perks that basically no one uses and make them good enough that you might consider them, instead of any other more meta perk. Machine Learnign was good and fun, but it was nowhere good enough to become meta, yet still it got clobbered down.

If BHVR really wants us killers to use any of this rarely used perks, don't make them attractive to use and don't kill them the moment they look viable, otherwise its back to Franklins and Weave Attunement again (until thats killed off next month).

Comments

  • bjorksnas
    bjorksnas Member Posts: 5,700

    Id rather see them reduce the duration for each time it procs rather than give it an overall duration decrease (oh they also either don't have it displayed right or actually nerfed it down to 40 seconds ingame

    2-3 procs per game should be impactful but to keep it more in line even if you manage to get a 4th / 5th time you should start having greatly diminished returns (especially because at minimum you always can get 1 per match since the last gen being done also counts as the other 2 gens being completed and thus always activates it if you kicked one of the last 3 gens)

    It should just get a new effect that says (after each time this perk activates gain a token decrease the duration by 10 seconds per token)

  • Akumakaji
    Akumakaji Member Posts: 5,487

    Hmm … but then you would have even less incentive to use the perk, if each activation made it worse? I don't really understand this reasoning.

    But your post gives me an idea, if we go by the kinda lore behind the perk, it should actually get better and better the more often it activates and the more the algorythm learns. Either that or maybe give it automatically the whole endgame as a duration? How cool would that be :D

  • bjorksnas
    bjorksnas Member Posts: 5,700

    Well ok heres an example

    current version activates for 40s per use

    with the decreasing returns the first activation might last 60s, then the next would last 50s ect, so even if you only manage 3 procs per match you would get 150s of total time from the perk compared to 120s with the current version

    It could work the same way in reverse though

    For each generator completed gain a token, for each token extend the duration this perk is active for by ~5 seconds, so early game ones are less impactful but late game ones can last a very long time

  • Akumakaji
    Akumakaji Member Posts: 5,487

    Ah okey! Well yeah, I guess that could work. I actually only managed to get 3 activations a couple of times, 2 seems to be the norm. And for such a hard activation, I would like to see some more bang for my bucks.

    The problem I have is that you have basically no control over when the perk activates: even a 99% gen could be abandoned by survivors if they feel its unsave and they think you might camp it, thus they split up and pressure others. And while the bonus speed and undetectable is really sweet, the unpredictability really hampers its positive impact on the game. I feel that the 60s was a really good spot and that the 45s was an overcorrection that the perk didn't needed. I am on the verge of dropping the perk alltogether, when I just started to like it and I think thats not a good positon for an off-meta perk to be.

    Of course, I am just one voice of many, but this perk is rarely incountered in the wild, and the nerf probably didn't help its popularity. While Peanits gave it a good explaination in one of his last posts of CM, ie that with its new easier activation process they wanted to reign in its power a bit, I feel this was a not a direction this perk needed to go.

  • bjorksnas
    bjorksnas Member Posts: 5,700

    I feel like it could use a bit more now. Its a bit boring and a bit too niche to really care about any more

  • Devil_hit11
    Devil_hit11 Member Posts: 9,090

    i think perk is worse than what it pre-buff because only one killer excels at using the perk. that killer was end-game legion. if you equip BFF add-on for 7% haste. Battery included for 5% haste, Noed for 4% haste, Machine learning for 10% haste and No way out to block exit gates. when you kick a gen and final gen is completed, machine learning activates and all these haste bonuses stack to make legion be around 140-150% m/s for 60 seconds. they saw this as too strong so now the perk is worse than what it was when it got buffed.

    i think they should make the perk automatically trigger when any gen get completed so killer gets use out of their perk and it is not just a joke 1 time end game perk. At the same time, It is BVHR. BVHR simply isn't serious with improving any of killer perk for reliable daily use in killer gameplay. I want a perk that i can seriously equip in a build and not feel bad about not using meta. Not a perk a buff to laugh for 1 game and never use again.

  • fussy
    fussy Member Posts: 1,720

    I agree, this is also my story with Zanshin. Just found a new perk in my builds once in decade and they nerfed it (thanks Nurse for it again, I guess) to the point I don't want to use it anymore.

  • bjorksnas
    bjorksnas Member Posts: 5,700
  • HolyDarky
    HolyDarky Member Posts: 844

    I love Machine Learning because it is a funny perk with a strong reward that requires good game sense and braincells to use it right. Even though they made the apply easier, it is still a difficult to active perk that can also be countered by survivors in many different ways. Therefore, I don't understand why the developers nerfed the perk from 60sec to 45 seconds even though it was still difficult to activate and not much used due to the fact that you need to think ahead. Now, even less players will use this perk and this is so sad. The perk got over the year many small buffs to become more viable and the last buff made it finally useable only to get nerfed next hotfix without any options to talk about it.

  • yukiday
    yukiday Member Posts: 48

    I honestly don't understand why this perk got nerfed. It's just as stupid as the nerf of Batteries Included.

  • Rokku_Rorru
    Rokku_Rorru Member Posts: 1,326

    Because haste for that long and that potent isnt healthy for the game. the amount of disparity between survivor and killer movement caused by all these haste perks create situations where people don't even have any input especially with how awful the servers are, I use it on trickster to let me aim and move faster for fun.

    40 seconds is still a long time, you could pair it with tinkerer to get more undetectable value from it and to notify you when its coming up or to kick the gen and pair it with claustrophobia (i forgot its current name) too and coupe de grace/dissoloution/blood price.

    it still has synergy, and it's still useful lol, I'm also not saying you are all clown mains coz of your portait, but it's funny having the thought of seeing a killer who would benefit from it massively all saying its not useful now lol