Murder Karma System Improvements 3

[aiden]

So, this went cold again so I have to revive it again

The main problem with my previous suggestion was that new players who don't know the rules or unlucky people will lose more Karma.

Karma right now works quite well, but the problem is that there still are quite a few RDMers. This might actually be a really useful feature.

Karma should take into consideration other factors in determining how much Karma is to be subtracted:

1. The frequency of RDMs in the past month and the past week for a player
2. The average frequency of RDMs per game for a player
3. The average ratio of not RDMing anyone in a game with a gun to not RDMing
4. If a player who was killed by another bystander is an RDMer
5. How many games the player has finished/completed fully
6. How many random kills in the game (@Kra3tys)
7. If the person shot had a gun (@Kra3tys)
8. If they RDMed in the previous game (@Kra3tys)

Here's why said factors were chosen respectively:

1. This is to determine if they are an RDMer or not.
2. Again, determining if they are an RDMer.
3. This can determine if they RDM just because they can.
4. This determines if another bystander with a gun was trying to stop an RDMer.
5. Determines how new the person is, reducing the toll on newbies who don't know the rules
6. To determine how much to multiply the base Karma penalty by
7. Less effect on targeters
8. To determine more accurately if someone is a RDMer

Formula: (@Kra3tys)
Base = 25
If meeting criteria to conclude that a person RDMs, then base = 50
Multiplier for people shot = 2.5
like normal lol

With this in mind, the system should then determine how much Karma to subtract within a certain range. Each variable should have a different impact on the final result, and also should not rely on previous results. (e.g. Subtracting 5-75 karma depending on the previously listed factors)

There should be a one-month notice before the new system takes effect so that some people can clean up their act. This is so that once the new system is in effect, they'll lose less Karma.

This results in less of a toll on people who don't purposely RDM, and a bigger one on those who do randomly kill.

As well as this, rules must be exposed far more heavily, and made sure that everyone knows the rules.

This is a huge improvement over version 2. If anyone has ideas to make it better, then please help.

 

[Timelord]

So, this went cold again so I have to revive it again

The main problem with my previous suggestion was that new players who don't know the rules or unlucky people will lose more Karma.

Karma right now works quite well, but the problem is that there still are quite a few RDMers. This might actually be a really useful feature.

Karma should take into consideration other factors in determining how much Karma is to be subtracted:

1. The frequency of RDMs in the past month and the past week for a player
2. The average frequency of RDMs per game for a player
3. The ratio of killing the murderer to RDMing for a player
4. If a player who was killed by another bystander is an RDMer
5. How many games the player has played
6. If they have read the rule book; how many pages
7. The amount of time between each RDM kill within the last week
8. Amount of gun pickups after someone has shot an innocent
9. Amount of random kills over the span of all Murder games played

Here's why I chose said factors respectively:

1. This is to determine if they are an RDMer or not.
2. Again, determining if they are an RDMer.
3. This can determine if they RDM just because they can, or if it was an accident, and they're trying to kill the murderer.
4. This determines if another bystander with a gun was trying to stop an RDMer.
5. Determines how new the person is, reducing the toll on newbies who don't know the rules
6. Determines if they know and understand the rules
7. Attempting to determine if someone is killing on accident
8. Determines if they still want to continue being a Detective running the risk that it may happen again
9. Further determination for if it's accidental or genuine

With this in mind, the system should then determine how much Karma to subtract within a certain range. Each variable should have a different impact on the final result, and also should not rely on previous results. (e.g. Subtracting 5-75 karma depending on the previously listed factors)

There should be a one-month notice before the new system takes effect so that some people can clean up their act. This is so that once the new system is in effect, they'll lose less Karma.

This results in less of a toll on people who don't purposely RDM, and a bigger one on those who do randomly kill.

As well as this, rules must be exposed far more heavily, and made sure that everyone knows the rules.

This is a huge improvement over version 2. If anyone has ideas to make it better, then please help.

Um I like the idea but how will my 1,800 innocents shot with 32k games played do?
Is there a well put cap to the amount of karma you will drop or have?
Also for (1. Frequency per week or Mounth) will this take into account games played? Because if I were to play 100 games per week and get 10 in shot then 1/10 ratio. But a player who does 10 games and shots 1 player will get the same 1/10 ratio. Yet per week this player will have 1 while I would have 10. This kinds puts playing a lot a bad thing.
...
I'm going to say no or -1 for now because of the complexity and unknown ways this could end up.

 

[Kraetys]

Um I like the idea but how will my 1,800 innocents shot with 32k games played do?
Is there a well put cap to the amount of karma you will drop or have?
Also for (1. Frequency per week or Mounth) will this take into account games played? Because if I were to play 100 games per week and get 10 in shot then 1/10 ratio. But a player who does 10 games and shots 1 player will get the same 1/10 ratio. Yet per week this player will have 1 while I would have 10. This kinds puts playing a lot a bad thing.
...
I'm going to say no or -1 for now because of the complexity and unknown ways this could end up.

Targeters for life man.

1. The frequency of RDMs in the past month and the past week for a player
2. The average frequency of RDMs per game for a player
3. The ratio of killing the murderer to RDMing for a player
4. If a player who was killed by another bystander is an RDMer
5. How many games the player has played
6. If they have read the rule book; how many pages
7. The amount of time between each RDM kill within the last week
8. Amount of gun pickups after someone has shot an innocent
9. Amount of random kills over the span of all Murder games played

1 - Acceptable, would take a new system, but relatively simple compared to the stat system already there.
However, sometimes you do get a series of unfortunate targets accidents and end up killing a lot of people in a single week.
It depends on the curve algorithm.

2 - Acceptable, and could probably be tracked through karma loss.

3 - Often times when you miss the murderer and shoot someone else, there is no second chance.
This seems like it'd be pretty pointless.
Unless you used it in a reverse psychology kinda way.

4 - Very good! Now I can target without my rank being affected so much.

5 - No lifers would have too much of an advantage, this may not be the best idea.
Not to mention the skew of games can be easily abused.
As long as you leave after you die, the game is always counted.
It's easy to commit suicide on murder.

6 - I don't think this is traceable and the rulebook was removed.

7 - Idk about this one, If you're any good at being a targeter you're going to have those intervals set really short.

8 - Not sure what this would do.
If it's bad, you can accidentally pick up a gun anywhere, it will skew the results.
If it's good, same.

9 - Already traced, but targeters have the higher number of random kills ironically.
You wouldn't be affecting the target demographic. (Hehehe).
We have more because we're good at it, and can keep our karma up so we can continue to play.
Real randomers have very little karma, and no real skill, 'ergo the randoming.
This makes it harder for them to random, again, the irony.

 

[aiden]

Um I like the idea but how will my 1,800 innocents shot with 32k games played do?
Is there a well put cap to the amount of karma you will drop or have?
Also for (1. Frequency per week or Mounth) will this take into account games played? Because if I were to play 100 games per week and get 10 in shot then 1/10 ratio. But a player who does 10 games and shots 1 player will get the same 1/10 ratio. Yet per week this player will have 1 while I would have 10. This kinds puts playing a lot a bad thing.
...
I'm going to say no or -1 for now because of the complexity and unknown ways this could end up.

I'd say cap it at 75.
What I mean is games over the past week, so 7/10 (bys shot) times 100 (games played) = a score of 70. The system would find the mean of all variables then divide by 2. Something like this formula might work, but I need some more time to see if it'll work for every variable.

Targeters for life man.


1 - Acceptable, would take a new system, but relatively simple compared to the stat system already there.
However, sometimes you do get a series of unfortunate targets accidents and end up killing a lot of people in a single week.
It depends on the curve algorithm.

2 - Acceptable, and could probably be tracked through karma loss.

3 - Often times when you miss the murderer and shoot someone else, there is no second chance.
This seems like it'd be pretty pointless.
Unless you used it in a reverse psychology kinda way.

4 - Very good! Now I can target without my rank being affected so much.

5 - No lifers would have too much of an advantage, this may not be the best idea.
Not to mention the skew of games can be easily abused.
As long as you leave after you die, the game is always counted.
It's easy to commit suicide on murder.

6 - I don't think this is traceable and the rulebook was removed.

7 - Idk about this one, If you're any good at being a targeter you're going to have those intervals set really short.

8 - Not sure what this would do.
If it's bad, you can accidentally pick up a gun anywhere, it will skew the results.
If it's good, same.

9 - Already traced, but targeters have the higher number of random kills ironically.
You wouldn't be affecting the target demographic. (Hehehe).
We have more because we're good at it, and can keep our karma up so we can continue to play.
Real randomers have very little karma, and no real skill, 'ergo the randoming.
This makes it harder for them to random, again, the irony.

Targeters for life man.


1 - Acceptable, would take a new system, but relatively simple compared to the stat system already there.
However, sometimes you do get a series of unfortunate targets accidents and end up killing a lot of people in a single week.
It depends on the curve algorithm.

2 - Acceptable, and could probably be tracked through karma loss.

3 - Often times when you miss the murderer and shoot someone else, there is no second chance.
This seems like it'd be pretty pointless.
Unless you used it in a reverse psychology kinda way.

4 - Very good! Now I can target without my rank being affected so much.

5 - No lifers would have too much of an advantage, this may not be the best idea.
Not to mention the skew of games can be easily abused.
As long as you leave after you die, the game is always counted.
It's easy to commit suicide on murder.

6 - I don't think this is traceable and the rulebook was removed.

7 - Idk about this one, If you're any good at being a targeter you're going to have those intervals set really short.

8 - Not sure what this would do.
If it's bad, you can accidentally pick up a gun anywhere, it will skew the results.
If it's good, same.

9 - Already traced, but targeters have the higher number of random kills ironically.
You wouldn't be affecting the target demographic. (Hehehe).
We have more because we're good at it, and can keep our karma up so we can continue to play.
Real randomers have very little karma, and no real skill, 'ergo the randoming.
This makes it harder for them to random, again, the irony.

1 - 4 fixes this.

3 - I tried fixing it. Look now.

5 - Would games completed work?

6 - Removed, thanks.

8 - Removed.

9 - Removed.

I need more variables now, it's been narrowed down to 5. Any ideas?

 

[Kraetys]

I'd say cap it at 75.
What I mean is games over the past week, so 7/10 (bys shot) times 100 (games played) = a score of 70. The system would find the mean of all variables then divide by 2. Something like this formula might work, but I need some more time to see if it'll work for every variable.



1 - 4 fixes this.

3 - I tried fixing it. Look now.

5 - Would games completed work?

6 - Removed, thanks.

8 - Removed.

9 - Removed.

I need more variables now, it's been narrowed down to 5. Any ideas?

Spoiler: My reasoning.

Well the thing is, the current karma system affects the target demographic perfectly.
If you're a real randomer, the results are catastrophic.
I've noticed a few variables it takes into account.
1 - How many per game, it's multiplication now, not addition. (Whatever the base loss is x5). (at least for two innocents).
2 - Whether or not the person you shot had a gun. (x2, does not stack).
3 - Whether or not you randomed the last game. (x2, does not stack).

The base is 25 unless the victim had a gun or you randomed the last game, or both, in which case the base is 50.
Then it's x5 for each new person shot. (Guesswork here, I've not shot 3 people in a game and I never plan to).
so it's 25(50), 125(250), (Guesstimation here), 625(1250), etc etc etc.
So if you're actually mass randoming, the results are utterly catastrophic to your karma.
This doesn't affect targeters single shot randomers so much now, and it does prevent multiple game mass randoms somewhat.

Spoiler: Solution.

What they need to do is fix their emerald system.
The amount of emeralds needed is supposed to increase based on how low your karma is. (Don't ask me for the algorithm here) (Extremely complicated amounts compliments of Gigo).

But, the way they're doing it now will just turn randomers into hiders.
(An emerald cost of 50+ for real randomers, there is no chance of getting a gun).

They need to make a softcap on the emerald cost increase system of 15 so there's a "chance" of getting a gun. (There isn't trust me).
Then make it so guns picked up = emeralds, not gun.
Then have the emerald cost increase apply to the gun, so it lowers the value of the gun with your karma, all the way to 1 if needed.
Then actually get it all working.

 

[aiden]

Spoiler: My reasoning.

Well the thing is, the current karma system affects the target demographic perfectly.
If you're a real randomer, the results are catastrophic.
I've noticed a few variables it takes into account.
1 - How many per game, it's multiplication now, not addition. (Whatever the base loss is x5). (at least for two innocents).
2 - Whether or not the person you shot had a gun. (x2, does not stack).
3 - Whether or not you randomed the last game. (x2, does not stack).

The base is 25 unless the victim had a gun or you randomed the last game, or both, in which case the base is 50.
Then it's x5 for each new person shot. (Guesswork here, I've not shot 3 people in a game and I never plan to).
so it's 25(50), 125(250), (Guesstimation here), 625(1250), etc etc etc.
So if you're actually mass randoming, the results are utterly catastrophic to your karma.
This doesn't affect targeters single shot randomers so much now, and it does prevent multiple game mass randoms somewhat.

Spoiler: Solution.

What they need to do is fix their emerald system.
The amount of emeralds needed is supposed to increase based on how low your karma is. (Don't ask me for the algorithm here) (Extremely complicated amounts compliments of Gigo).

But, the way they're doing it now will just turn randomers into hiders.
(An emerald cost of 50+ for real randomers, there is no chance of getting a gun).

They need to make a softcap on the emerald cost increase system of 15 so there's a "chance" of getting a gun. (There isn't trust me).
Then make it so guns picked up = emeralds, not gun.
Then have the emerald cost increase apply to the gun, so it lowers the value of the gun with your karma, all the way to 1 if needed.
Then actually get it all working.

This is really good.
The only thing is that I think a factor of 5 is too excessive. If you shoot 3 people and meet the criteria required for the base to be 50 then you'll already be in the negatives with just one game. Too much of a drop. I think a multiplier of 2.5 would work best.

 

[Kraetys]

This is really good.
The only thing is that I think a factor of 5 is too excessive. If you shoot 3 people and meet the criteria required for the base to be 50 then you'll already be in the negatives with just one game. Too much of a drop. I think a multiplier of 2.5 would work best.

This is something already ingame. If you shoot 3 people you need to get something straight.

 

[aiden]

This is something already ingame. If you shoot 3 people you need to get something straight.

I've seen it a lot.