Fixing MFK in jb (Somewhat)

[Kaarma]

I think the way to fix most (intentional) mfks are make a zone in maincells where most mfks happen and if somebody kills 3-5 people in a short time period 1-2 seconds in that zone they would be banned and the people who got freekilled would be respawn. This is not a perfect solution but it is somewhat would reduce the ammount of mfks that generally happens.

 

[Kaarma]

This would help with better player experience if they got killed during an mfk and they don't have to sit and wait for the round to be over.

 

[kaZZ]

-1 this would be hard to set up and there will be a lot of false bans, this will give admins/managers more work with no benefits

 

[4g0tt3nSou1]

Fact of the matter is this. We can have the most robust auto admin system, it still won't "fix" MFK/MassRDM. There isn't a solution.

However this could be one tying in the tool belt.

My biggest concern is there could be MANY reasons why A CT would kill 5 people in quick succession in early round.

Fact is free kill is entirely situational. What is free kill in one scenario isn't always free kill in others, vise versa. So how would this plugin/modification know? It physically can't.

Look, I appreciate the thinking outside the box. But I don't see a clearcut way where this does as intended without adverse reactions.

Imagine 5 guys in a gang go through and gang up on one ct. They get em banned for killing them even though that CT was in the right

This just doesn't feel like it'll work. -1

 

[Kaarma]

Fact of the matter is this. We can have the most robust auto admin system, it still won't "fix" MFK/MassRDM. There isn't a solution.

However this could be one tying in the tool belt.

My biggest concern is there could be MANY reasons why A CT would kill 5 people in quick succession in early round.

Fact is free kill is entirely situational. What is free kill in one scenario isn't always free kill in others, vise versa. So how would this plugin/modification know? It physically can't.

Look, I appreciate the thinking outside the box. But I don't see a clearcut way where this does as intended without adverse reactions.

Imagine 5 guys in a gang go through and gang up on one ct. They get em banned for killing them even though that CT was in the right

This just doesn't feel like it'll work. -1

You could have it as places where t's normally get placed like cell buttons and what not. It would be a lot of work but it could come out as good

 

[kaZZ]

You could have it as places where t's normally get placed like cell buttons and what not. It would be a lot of work but it could come out as good

"Stand on the cell button and AFK freeze"
"Get to the red marker by the cell button and do not be on the cell button"

These commands will kill multiple Ts if they don't follow the command, and will "trigger" the "autoban"

There's no way this will ever be useful, and you have to set the "zones" on every single map. There are so many situations that will cause false bans

 

[jekgog]

Or my 1up plugin suggestion gets added

 

[Kaarma]

Or my 1up plugin suggestion gets added

/shrug

 

[Wahtur]

The issue is that a machine will never be able to tell what is a freekill or not.

 

[gliptal]

bad idea
Good idea -->Slay day
slay days are pretty common on css servers and its basically just banning the guy who mfkd and slaying all cts so it goes to the next round.
There are a few problems with this so there would probably need to be some proper implementation.

problem #1 -> +1 death to all cts for no reason (only matters on gameme i think)
Pretty much no one cares about gameme but this easily fixed by changing the round time to 1 second like how LR lowers to 2:30(i think, if lowering isnt possible then idk).
problem #2 -> warden queue reset for warden of mfk round
Many ways to fix but why I think this would need proper implementation. Basic fix is giving the warden of the mfk round 100% priority the next round.

this obviously wouldnt be done every mfk, but is clearly better than the current solution of doing nothing and possibly having 30+ people wait up to 5+ minutes in spectator.

 

[WaSe]

Here me out,

If we ban everyone, no one can mfk

 

[Kaarma]

Here me out,

If we ban everyone, no one can mfk

nerd

 

[nitro]

i personally autobahns are stupid in general

however, the idea wouldn't be too bad if it was, say, an automatic spectate mover?

±1

 

[Ritzee]

i personally autobahns are stupid in general

however, the idea wouldn't be too bad if it was, say, an automatic spectate mover?

±1

maybe not spectate but slay just in case it is false they dont have to wait in queue again

 

[4g0tt3nSou1]

maybe not spectate but slay just in case it is false they dont have to wait in queue again

This is the At-Best for this suggestion.
This is the best approach at a non-aggressive system as we can get without making it worthless and without any fangs.

However, the bigger underlying point still stands;
Zoning every map is tedious and ultimately wasted effort, as the system would be unable to identify MFK by guess in a majority of cases.

The only proper way to do it, in my opinion, is basically impossible*. That'd be to use a Machine-Learning algorithm to train a model to self-identify these MFKs.
This is not plausible or probable with the tools in-front of us just couldn't interface with such a algorithm. We can really only use SM for modifying the server, and there isn't easily a way to export clips of the MFK/MassRDM/etc, and pass those through such an advanced algorithm. Not just that, but all due respect to tech. I doubt they could implement this anyways; EVEN IF IT WAS PLAUSIBLE to interface with it through sourcemod.
This is also entirely OVERKILL. The amount of work to make this work is more than valve puts into their anticheat using the same tech; which also as you know sucks.
As Nerd-Boner inducing as having a ML-Based anticheat/rule enforcement system would be for our community; it just won't happen anytime soon, even possibly ever.
* Now, I don't believe that things are impossible, to me impossible is theoretical; but this definitly isn't in the realm of what eGO could accomplish. Especially in our current position.

TTT Has been toying with ideas like this for a min, (Link is the open thread about it) but it is getting more on people's nerves rather than solving anything.
Further, Let's say things go similar to TTT's implementation... People QUICKLY found out how to exploit it and effectively punish players for doing the right thing. The stories of that implementation speak for themselves; and while I'm not as active on TTT as I used to be, I haven't heard of it ever catching someone who MassRDMs.
If anyone has success stories for that system; I'd be interested in hearing them. But I think much like TTT, an Anti-MFK system in JB would have the same, if not worse, failure.

Look, The idea you are ;putting forward has good intentions, and your heart is in the right place. But tried and tested, the only way to really go about administrating these is to have a human do it.

 

[Spencer]

bad idea
Good idea -->Slay day
slay days are pretty common on css servers and its basically just banning the guy who mfkd and slaying all cts so it goes to the next round.
There are a few problems with this so there would probably need to be some proper implementation.

problem #1 -> +1 death to all cts for no reason (only matters on gameme i think)
Pretty much no one cares about gameme but this easily fixed by changing the round time to 1 second like how LR lowers to 2:30(i think, if lowering isnt possible then idk).
problem #2 -> warden queue reset for warden of mfk round
Many ways to fix but why I think this would need proper implementation. Basic fix is giving the warden of the mfk round 100% priority the next round.

this obviously wouldnt be done every mfk, but is clearly better than the current solution of doing nothing and possibly having 30+ people wait up to 5+ minutes in spectator.

Slay days sound cool, just make sure the guy who got Warden gets it next round.