Not just complaining about sbmm

So just like the title says this is not about complaining about the system. There are plenty of those threads already. Instead let's discuss what the system could track to be better than it currently is.


Looking for something productive here so im going to start of with the sbmm should track how effective you are at escaping the killer. You have lots of long chases before going down your skill level goes up lots of short and it goes down (protection hits needs to be factored in too) thoughts?


Also please be aware it's against the rules to discuss the current mechanics behind sbmm so if you are aware of them do not mention it as you will get this thread deleted.


Any other ideas for what's should be tracked?

Comments

  • Basing MMR on each of the four emblemic gradings for survivor team mates (i.e. an altruistic survivor may get paired up with similar players, or objective low scorer put with other low objective scorers, but it's about getting all 4 emblems as closer as possible.

    For killer, it may have to be a bit more basoc as cpmparing killer emblems with survivor emblems could be tricky. Although, if they did, I would suggest:

    • Lightbringer paired with Gatekeeper

    • Evader paired with Chaser

    • Unbroken paired with Devout

    • Benevolent paired with Malicious

    So, effectively, have an MMR scoring for each emblemic section and match with opponents closest to each MMR. That way, you'll work with not just similar skilled people, but also people who have a similar playstyle, so it encourages people to participate in all sections. Someone who solely does gens will get a low Benevolent score, and will be teamed with other gen people, so they'd need to branch out the skills.

  • The problem with escaping chase and all that is that you can escape by dropping a God pallet, or by the killer just dropping chase. Especially common with obsession related perks. Or just by running towards house of pain, most killers will instantly drop unless they are doing something weird like hunting for BBQ stacks. Heck, just using an exhaustion optimally will result in an escape.


    And that is VERY likely the reason they do not consider it. It would be the easiest thing in the world to inflate. Same with the opposite, you just use every resource and purposely not break chase.

  • @ThiccBudhha said:

    The problem with escaping chase and all that is that you can escape by dropping a God pallet, or by the killer just dropping chase. Especially common with obsession related perks. Or just by running towards house of pain, most killers will instantly drop unless they are doing something weird like hunting for BBQ stacks. Heck, just using an exhaustion optimally will result in an escape.

    And that is VERY likely the reason they do not consider it. It would be the easiest thing in the world to inflate. Same with the opposite, you just use every resource and purposely not break chase.

    True but If you are consistantly able to escape quickly or run long chases I'd say that makes you a higher skilled player no?

  • @wxnickxw said:

    https://forum.deadbydaylight.com/en/discussion/comment/2499423#Comment_2499423

    True but If you are consistantly able to escape quickly or run long chases I'd say that makes you a higher skilled player no?

    Of course, I just do not believe it would be a healthy system if it were ever discovered. And given where we are at now, like.. I have never seen an mmr system datamined to be honest.

  • @ThiccBudhha said:

    https://forum.deadbydaylight.com/en/discussion/comment/2499434#Comment_2499434

    Of course, I just do not believe it would be a healthy system if it were ever discovered. And given where we are at now, like.. I have never seen an mmr system datamined to be honest.

    I don't follow why would it be unhealthy? Becuse it would promote survivors running long chases?

  • @Marc_go_solo said:

    Basing MMR on each of the four emblemic gradings for survivor team mates (i.e. an altruistic survivor may get paired up with similar players, or objective low scorer put with other low objective scorers, but it's about getting all 4 emblems as closer as possible.

    For killer, it may have to be a bit more basoc as cpmparing killer emblems with survivor emblems could be tricky. Although, if they did, I would suggest:

    • Lightbringer paired with Gatekeeper

    • Evader paired with Chaser

    • Unbroken paired with Devout

    • Benevolent paired with Malicious

    So, effectively, have an MMR scoring for each emblemic section and match with opponents closest to each MMR. That way, you'll work with not just similar skilled people, but also people who have a similar playstyle, so it encourages people to participate in all sections. Someone who solely does gens will get a low Benevolent score, and will be teamed with other gen people, so they'd need to branch out the skills.

    Oh you want them to curate the teams as opposed to just a relative skill. I'm not gonna say that what your proposing wouldn't be better but I'm not sure they are capable of that. The devs I mean not humanity in general.


    Start with something a little smaller of an improvement than there current mechanics

  • @wxnickxw said:

    https://forum.deadbydaylight.com/en/discussion/comment/2499413#Comment_2499413

    Oh you want them to curate the teams as opposed to just a relative skill. I'm not gonna say that what your proposing wouldn't be better but I'm not sure they are capable of that. The devs I mean not humanity in general.

    Start with something a little smaller of an improvement than there current mechanics

    Yeah, I admit it's quite complex, but the results ought to be very good for matchmaking, whilst also encouraging players to get out of their comfort zones.

    This is the thing with SBMM: it needs to truely represent the skills of the player, and that can't be done on a simple escape/kill tally. It needs to take in various factors.

    Ironically, the emblem system for those 4 categories is the perfect foundation for it! It's just that they've not capitalised on the idea.

    Currently, I like the new SBMM system in terms of my experiences. However, if the datamined information is correct, I don't think "Skill-Based Match Making" is an accurate phrase to use. It's more of a simple league table, like football.

  • @Marc_go_solo said:

    https://forum.deadbydaylight.com/en/discussion/comment/2499519#Comment_2499519

    Yeah, I admit it's quite complex, but the results ought to be very good for matchmaking, whilst also encouraging players to get out of their comfort zones.

    This is the thing with SBMM: it needs to truely represent the skills of the player, and that can't be done on a simple escape/kill tally. It needs to take in various factors.

    Ironically, the emblem system for those 4 categories is the perfect foundation for it! It's just that they've not capitalised on the idea.

    Currently, I like the new SBMM system in terms of my experiences. However, if the datamined information is correct, I don't think "Skill-Based Match Making" is an accurate phrase to use. It's more of a simple league table, like football.

    Also something to severely lower players who haven't quite got the fundamentals yet. For example if you fail skill checks regularly your gonna be kept in the low tiers till you get better at it


    Edit this was in reference to their points in the category dealing with doing objectives. Obviously this has to be separate from the chase and other categories or people will abuse it

  • The problem with measuring chases is that a longer chase doesnt need to be better. There are many factors, which need to be considered:

    The place: looping around an almost completed gen or an hooked person is not good

    Ressources: if you use every good pallet at the start to get a long chase timer, everyone after that has shorter chases, even when they are more skilled

    End of the chase: doesnt matter if the killer drops it or gets you, if you lead the killer to an objective and he interrupts that, it is a bad move. You could equip bond and run to other survivor and easily escape the chase.

  • @TicTac said:

    The problem with measuring chases is that a longer chase doesnt need to be better. There are many factors, which need to be considered:

    The place: looping around an almost completed gen or an hooked person is not good

    Ressources: if you use every good pallet at the start to get a long chase timer, everyone after that has shorter chases, even when they are more skilled

    End of the chase: doesnt matter if the killer drops it or gets you, if you lead the killer to an objective and he interrupts that, it is a bad move. You could equip bond and run to other survivor and easily escape the chase.

    But going to those good places is kind of the mark if a good survivor. True on good chase doesn't mean much but if your consistently running long chases then it doesn't really matter why your running 4he killer for a long time what matters is that you do

  • @wxnickxw said:

    https://forum.deadbydaylight.com/en/discussion/comment/2501088#Comment_2501088

    But going to those good places is kind of the mark if a good survivor. True on good chase doesn't mean much but if your consistently running long chases then it doesn't really matter why your running 4he killer for a long time what matters is that you do

    Running shack or just dropping the pallet immediately, is a big difference. If we talk about a 5-gen chase, then it really doesnt matter. In a comp scenario it would be the right play to drop the pallets immediately, so we even could say thats skill.

    But if every survivor gets measured for their chases, i would just run atraight to the killer at the beginning. Then i would have all pallets available and would get the best chase.

  • @wxnickxw said:

    https://forum.deadbydaylight.com/en/discussion/comment/2500022#Comment_2500022

    Also something to severely lower players who haven't quite got the fundamentals yet. For example if you fail skill checks regularly your gonna be kept in the low tiers till you get better at it

    Edit this was in reference to their points in the category dealing with doing objectives. Obviously this has to be separate from the chase and other categories or people will abuse it

    Yeah, true. Granted, it would take some time to implement it, but if this new system takes even a little more into account then they can build on it. I'm worried they won't, but moving forward they ought to have even just a small team that focuses on MMR improvements throughout. I'd be all for that.

  • @Marc_go_solo said:

    https://forum.deadbydaylight.com/en/discussion/comment/2500971#Comment_2500971

    Yeah, true. Granted, it would take some time to implement it, but if this new system takes even a little more into account then they can build on it. I'm worried they won't, but moving forward they ought to have even just a small team that focuses on MMR improvements throughout. I'd be all for that.

    Unfortunately they don't look to be interested in developing it as they have been actively suppressing conversations about what the system in place is