Should survivor mmr be based on the number of survivors that escape?
Currently, when you play survivor, you gain mmr if you escape and lose mmr if you die, this is the only metric that counts.
Should it instead be based on the number of escapes (+10 points when a survivor escapes, -10 points when a survivor dies)? If a survivor disconnects it counts as a death to prevent exploits. On the short term you can lose or gain mmr because of teammates, however if you do the maths the chances of your mmr being impacted long term are so low that they are negligible.
Should survivor mmr be based on the number of survivors that escape? 33 votes
26 votes
7 votes
Comments
-
Yes, make it based on the number of escapes instead of just you
Survivors should absolutley be scored as a team, not individuals, since they are a team, not individuals.
5 -
Yes, make it based on the number of escapes instead of just you
Yes 100%.
1K 3E
The one Survivor that died lost MMR points for being the only death, even tho the rest of the team escaped. It's a team game, Survivors should be ranked based on how good their team, as a whole, was
3 -
Yes, make it based on the number of escapes instead of just you
I don't think mmr should only be calculated from team performance, but I agree that makes more sense than what we currently have.
4 -
Yes, make it based on the number of escapes instead of just you
Ula in my opinion, team performance is enough, on the long term.
Let me explain: yes, by bad luck you can get bad teammates. However, in the long run, your luck will be average (binomial distribution, if you do many games the chances of you being negatively impacted by teammates are very low).
0 -
No, it is better as it is
Survivors should only be scored as individuals - not a team.
If this were the case, the individual scores of each surtvivor's performance can easily and horrendously become skewed by raw teammate input.
You can be the greatest survivor player ever, but get stuck in elo hell because the SBMM places you with rank 20 noobs that don't know anything but how to crouch to the map's edge when inside the killer terror radius.
You hardly escape, but when you do - it stunts your impeccable score because of poor teammate performance.
Then you have the opposite, where a noob gets into a few lucky matches by sheer chance, or plays with friends a single time. They go down in seconds, never touch a generator or help their team - and are completely unfamiliar with core concepts of gameplay. However, they are sparring with red ranks because, although they do virtually nothing and die very quickly - they are scored based on their team.
In addition, the better your team does, the better you do as a survivor. They all feed into one another.
How many times do you get to post-game screen and you see 3 survivors with <10k bp and 1 survivor sitting on the top with a kingly sum of 30k?
How many times do you see survivor scores often clumping together into the same ballpark, especially at higher levels?
Just judge a survivor based on their own performance - and create higher scoring events for when players do what they must to ensure a 4e.
If the system has a few kinks in it, it's best to fix what's causing them than to reach for the gaff tape.
1 -
Yes, make it based on the number of escapes instead of just you
@Valik but in the long run, like in 100 games or so, being screwed or carried is nearly impossible. When you have 300 random teammates, then with binomial distribution you should know that it will even out (if you know what this thing is). The noob that got lucky will eventually face binomial distribution and get back to the rank he deserves, also, if he gets too high his incompetence will drag others down, causing defeat. Do you want me to explain what binomial distribution is? I can do it if you want.
Also, according to your logic (that escapes = skill), survivors that hide in lockers and just move to avoid crows all game and then use left behind or a key to escape have done a good job, since they escaped. Very easy to exploit. Good job on the survivor that always hides when exit gates are powered, lets others open them and then escape while leaving behind the teammates who did all the work.
0 -
Yes, make it based on the number of escapes instead of just you
Yes 100%. Survivors are not individuals, they are a team. If I could have escaped, but I died in order to ensure my teammate's escape when they otherwise would have died, I don't consider that a loss on my end. This change would also promote more altruistic gameplay instead of leaving teammates behind in endgame who can realistically be saved.
Now it might make some sort of sense to have it based on both your performance and your team as a whole, but it should not be based off just whether you escaped and nothing more.
1 -
People that die extremely quickly early on shouldn't be getting higher mmr if everyone else escapes
0 -
Yes, make it based on the number of escapes instead of just you
@TeabaggingGhostface but this is very unlikely to happen in the first place in a 3v1, just look at binomial distribution.
0 -
@slunder the amount of people i see that suicide on first hook is too high
0 -
No, it is better as it is
Not at all.
How many games within 100 will you preform accurately to your own skill level while you're alive?
How many games within 100 will you have bad teammates, DCs, or get boosted?
Even if less than half of the games have you mismatched with teammates, or sport crippling drawbacks - it's still infinitley worse than simply scoring players based on their performance in the first place.
Making a problem even worse doesn't check out merely because it can be diluted with volume.
SBMM should be a light switch that turns on or off. Not a system that - over the course of 100 or more uses will be more correct than not.
And I'm not the one who believes that escapes = skill. That is simply the current system. I 100% agree with you that escape scoring is a horrendous choice and should be changed.
I personally think the system should consider escaping to be the cherry on top of the gameplay sundae instead of the only factor in determining player skill - but we have what we have.
I respect your opinions, but I hope that you can see my point when I say that diluting a problematic system to create a 'generally correct' result through sheer volume is not a concise or ideal solution.
(Also, no explanation is needed, thanks for the offer though! It's always good to have people excited to share knowledge)
0 -
Yes, make it based on the number of escapes instead of just you
@Valik yes, but you can also be mismatched with teammates better than you. And yes, I am aware on the short term it is not perfect. If there are better solutions I am open for them, but the devs never really managed to create good ranking systems that work, neither on the short or long term.
0 -
No, it is better as it is
What would you think about a BPPM model?
Tweak scoring events and average out the amount of BP gained in a match for active time (Sitting on hook or 'find help' while on ground doesn't count)
Do you think that Score Per Minute systems using DbD's current BP scoring would work out well?
Why or why not?
0 -
Yes, make it based on the number of escapes instead of just you
@Valik I think there would be problems.
If you have 2 survivors, one uses 5 pallets and gets a stun every time and goes down in 1 minute, the other uses 0 pallets and goes down in 1 minute, then the first survivor gets blood points for the stuns but the second one does not.
Then, the survivor who unhooks all 3 teammates but then dies instead of leaving gets less blood points that the one who just leaves.
I could give many examples, but blood points are not really based on skill.
Also, a survivor could literally just sabo hooks when the killer is far away and gain more blood points than if he was doing generators.
I could literally give examples like this all day, just tell me how many you want.
0 -
Yes, make it based on the number of escapes instead of just you
I think that idea is generally more suitd for actual groups of survivors (swf). Solo wouldnt care about others, right?
Whereas swf players care for each other, thus such rules are better suited for swf. Can be extended to bloodpoint scores too. ( bonus bp for healing a survivor not in your group). This would need the game to track swf in the trial tho.
0 -
No, it is better as it is
Uhhh...
Yeah, because stunning the killer is a good thing.
But, if this was such a grand issue - it would be extremely simple to fix, by creating a scoring event that triggers once every 20 seconds of chase that a pallet is not used.
If a survivor stuns the killer with each and every pallet in the map without getting hit, good on them. This is not, by virtue, a bad thing at any level of play. But if a survivor manages to avoid getting hit or throwing a pallet, giving them a hefty reward would be an awesome Survival scoring event that could help round out the otherwise anemic BP category.
Also... you are fortunately (unfortunately?) mistaken. This is the beauty of SPM/BPPM models. A survivor that makes two heroic unhooks, takes protection hits and dies has a much more compact spread of BP than those who flee from the hook and make an escape.
Survive for 5,000 bp for the escape and ~250 for opening the exit gates and waiting there with a single heal each (750) for everyone to get out after 45 seconds. (8,000 BPPM)
Or getting two safe late unhooks with two protection hits, the chase conclusion score, a little recovery and wiggle put together ( 2,000 + 2,000 + 200 + 200 + 600 + ~1,000) in 30 seconds. (12,000 BPPM)
In the situation as you described is not too uncommon, but it also illustrates how - without any tweaking - the numbers can work as-is in many cases like this.
If this scoring event based matchmaking system was live, the late unhook hero would be considered twice as skillful as the two others that opened the 99'd exits and healed up before taunting the killer to leave.
Survivors on generators also earn tons of blood points. Skill checks can give 300bp each, Repairing a generator can grant over 1,000bp with +25 for each second preformed along side another person. if in the killer's terror radius, they also gain some boldness.
If someone sits on a generator for 30 seconds with a teammate that gets chased off halfway through - they can gain over 2,000 bp (1,000 BPPM)
Someone who is able to sabotage 4 hooks in the same time will gain 500bp each 2,000 bp (1,000 BPPM).
Yeah, you could run around finding hooks to sabotage, but you're better off just repairing and hitting great skill checks.
If BPPM was implemented, it would be exceedingly simply to change the BP score of Sabotaging to 200bp and have it increased to 800bp if done while the killer is carrying a survivor.
That being said, the system would only need minor changes to scoring events to even this out.
For instance, only rewarding points for throwing a pallet if it stuns the killer or is broken quickly, while also scoring for survivors that conserve pallets and make do without.
I understand if you still disagree, but it sounds like you have the wrong mental picture. I hope this helped clear some of that up!
0 -
Yes, make it based on the number of escapes instead of just you
@Valik I think that a rework in which the devs transform the bloodpoint system into a system that is accurate for determinating skill would be a very long and complex task. They are too busy releasing new dlcs of trash maps and killers whose power us just a remix of older killer powers.
0 -
No, it is better as it is
A got camped by Bubba, with Kindred.
Either B or C or D has Prove thyself and able to rush the remaining Gen. 3 escape.
A's MMR increased just by hitting Space in struggle phase
1 -
No, it is better as it is
Fair is fair. That may be where we disagree at the core. I personally think that changing numbers like +500 to +200 is probably the easiest kind of patch that can be implemented in any game, at any time.
I appreciate where you're coming from though - I think we can all agree that the current MMR system needs some serious work before it's truly in an acceptable state for the community.
0 -
Meanwhile its ok for Bubba's MMR to increase while he stands infront of the hook to get his kill?
0 -
No, it is better as it is
Its killer choice to go for cheap kill.
0 -
Doesn't change the fact the killers mmr is gonna rise while the camped survivor ones drops, even if he looped the Bubba for 5 gens and then ends up getting face camped.
0 -
Wouldn't this make higher ranked survivors go against babies more often?
If you escape regularly but it doesn't really increase mmr unless more of ur teammates escape i feel like you would go against more baby killers
0
