Glorious Alpha Two Testers!
Alpha Two Phase II testing is currently taking place 5+ days each week. More information about testing schedule can be found here
If you have Alpha Two, you can download the game launcher here, and we encourage you to join us on our Official Discord Server for the most up to date testing news.
Alpha Two Phase II testing is currently taking place 5+ days each week. More information about testing schedule can be found here
If you have Alpha Two, you can download the game launcher here, and we encourage you to join us on our Official Discord Server for the most up to date testing news.
[Suggestion] Behavioral pattern analysis to identify griefers/trainers in dungeons.
TheRealScar
Member, Alpha Two
People aren't going to run dungeons if they just die all the time. They will specifically avoid dungeon activities resulting in death when its a solo player training mobs into every group in the dungeon, regardless of motivation. If GMs are not going to actively monitor solo player activity in the dungeons, look at their damage/healing output when they gain aggro from mobs, or look at pathing patterns of said solo player... Maybe look at automating it with alerts.
Conceptually, a system could be coded to collect the necessary data to identify players training in dungeons that result in other players deaths, and alert GMs to this activity. A lot of this info has been pieced together from MMO coding questions with multiple AI platforms to understand some player movement/logging/database activities in MMOs and the resources required to make that happen (I didn't want to post something if it was such a heavy programming lift it made it unrealistic), it kind of morphed into how to automate player tracking for potential griefers/trainers (categories and some of the behavioral definitions are fully credited to ChatGPT query responses)
1. As a group enters the dungeon area, they get assigned some identifier, solo players get their own identifier, associate the following data points with identifiers.
Patterns of player movement in dungeons:
Paths taken by players in real-time, logging player positions and if the player is moving or not.
Patterns of mob movement in dungeon:
Identify mobs that are aggro'd, mob pathing, and who mobs have targeted.
Interaction between players and mobs:
Log the damage/healing/abilities/buffs/debuffs
Proximity between players/groups:
Log the distance between groups in the dungeons. Are the players on the same level/elevation? What is the distance between groups? How often are players reducing that distance to zero?
2. Behavioral Rules (ChatGPT, 2025)
Define behaviors that indicate potential griefing. For example:
Aggro Hopping:
The player consistently pulls multiple mobs and redirects them toward another group without engaging in significant combat.
Target Switching:
Mobs frequently switch from the player to another group in suspicious patterns.
Mob Density:
The number of mobs in a group’s vicinity spikes abnormally, and the pattern repeats.
3. Pattern Recognition Algorithms (ChatGPT, 2025)
Use algorithms to detect griefing patterns:
Path Analysis:
Compare the player's path to the mobs’ aggro paths. If the player repeatedly leads mobs directly toward other groups, flag the behavior.
Heatmaps:
Generate real-time heatmaps of mob density around groups and correlate spikes with nearby solo players or groups.
Repetition Analysis:
If a player engages in this behavior multiple times within a specific time frame (e.g., 3 instances in 10 minutes), increase suspicion level.
4. Behavior Scoring System (ChatGPT, 2025)
Introduce a scoring system to assess the likelihood of griefing:
Weighted Metrics:
Assign weights to actions such as:
Pulling mobs without engaging them (+2 points).
Leading mobs into another group’s proximity (+3 points).
Lack of meaningful combat engagement (e.g., <10% of the mob’s HP dealt in damage, +2 points).
If the score exceeds a threshold (e.g., 10 points), trigger further action.
5. Automated Responses
Implement in-game responses based on detection:
6. Log and Review (ChatGPT, 2025)
Event Logs:
Automatically save logs of flagged behavior, including timestamps, player paths, and mob interactions.
Admin Alerts:
Send notifications to moderators or admins for further review if the behavior persists or is particularly disruptive.
I'm sure lots of people have talked about this stuff, I would hope Steven and his team are discussing this stuff as well, I just figured if not, a solution starts somewhere, and complaining about it isn't going to fix it.
Conceptually, a system could be coded to collect the necessary data to identify players training in dungeons that result in other players deaths, and alert GMs to this activity. A lot of this info has been pieced together from MMO coding questions with multiple AI platforms to understand some player movement/logging/database activities in MMOs and the resources required to make that happen (I didn't want to post something if it was such a heavy programming lift it made it unrealistic), it kind of morphed into how to automate player tracking for potential griefers/trainers (categories and some of the behavioral definitions are fully credited to ChatGPT query responses)
1. As a group enters the dungeon area, they get assigned some identifier, solo players get their own identifier, associate the following data points with identifiers.
Patterns of player movement in dungeons:
Paths taken by players in real-time, logging player positions and if the player is moving or not.
Patterns of mob movement in dungeon:
Identify mobs that are aggro'd, mob pathing, and who mobs have targeted.
Interaction between players and mobs:
Log the damage/healing/abilities/buffs/debuffs
Proximity between players/groups:
Log the distance between groups in the dungeons. Are the players on the same level/elevation? What is the distance between groups? How often are players reducing that distance to zero?
2. Behavioral Rules (ChatGPT, 2025)
Define behaviors that indicate potential griefing. For example:
Aggro Hopping:
The player consistently pulls multiple mobs and redirects them toward another group without engaging in significant combat.
Target Switching:
Mobs frequently switch from the player to another group in suspicious patterns.
Mob Density:
The number of mobs in a group’s vicinity spikes abnormally, and the pattern repeats.
3. Pattern Recognition Algorithms (ChatGPT, 2025)
Use algorithms to detect griefing patterns:
Path Analysis:
Compare the player's path to the mobs’ aggro paths. If the player repeatedly leads mobs directly toward other groups, flag the behavior.
Heatmaps:
Generate real-time heatmaps of mob density around groups and correlate spikes with nearby solo players or groups.
Repetition Analysis:
If a player engages in this behavior multiple times within a specific time frame (e.g., 3 instances in 10 minutes), increase suspicion level.
4. Behavior Scoring System (ChatGPT, 2025)
Introduce a scoring system to assess the likelihood of griefing:
Weighted Metrics:
Assign weights to actions such as:
Pulling mobs without engaging them (+2 points).
Leading mobs into another group’s proximity (+3 points).
Lack of meaningful combat engagement (e.g., <10% of the mob’s HP dealt in damage, +2 points).
If the score exceeds a threshold (e.g., 10 points), trigger further action.
5. Automated Responses
Implement in-game responses based on detection:
6. Log and Review (ChatGPT, 2025)
Event Logs:
Automatically save logs of flagged behavior, including timestamps, player paths, and mob interactions.
Admin Alerts:
Send notifications to moderators or admins for further review if the behavior persists or is particularly disruptive.
I'm sure lots of people have talked about this stuff, I would hope Steven and his team are discussing this stuff as well, I just figured if not, a solution starts somewhere, and complaining about it isn't going to fix it.
1
Comments
Talk about double standards.
I know, reading is hard... also...
Ya motha.
Apparently proof reading is also hard.
And if you don't want people calling out your double standards then I would suggest not having double standards.
He wrote “every group.” You gave a hypothetical about a specific group getting retaliated against. You should read more carefully.
Look, I'm sorry you had a bad experience getting rinsed wherever you were farming. The post was written with the perspective of folks going into dungeons to train mobs onto any or every group in there, whether initiating training or responding to previous training, hence, "regardless of motivation" it's supposed to be ambiguous (sorry, "non-specific" for the I-failed-reading-rainbow audience) because training is a garbage tactic that abuses the way the mobs are leashed and aggro.
We all get to have our asses kicked while we play, it's PvP, I don't chug an energy drink, listen to some Fiona Apple, cry into my keyboard then go and try to train a bunch of mobs on people. I'll message some friends to contest, or move on and find something else to farm.
You do you though, boo.
Also... ya motha.
Sure, they want to fix the hate bug that exists, but other than that there is no reason or indication that this is something that isn't intended.
And yet here you are on the forums crying and calling for people to be banned because someone used the environment to beat you rather then mashing buttons on their keyboards, or calling friends.
They probably wont change it given that this was one of the charms of playing on older games like EQ where one person could drag 30 mobs to a zone entrance and if they didn't warn people they were doing it well you were shit out of luck.
I posted with no expectations from intrepid. If it helps it helps, if it doesn't, nothing lost. Hell, maybe intrepid considers training mobs part of competing for resources.
"Crying... calling for people to be banned..."
Can you highlight, maybe even provide a quote from my posts where that occurs? The only evidence of someone hurt here is you and the massive baggage you dragged in here with your oddly specific narrative about training. I mean, maybe you're hurt, maybe not, would be presumptuous of me to say one way or the other... either way, you missed the point of the post, and that's sad...
Damn, you're probably just a forum troll, my bad. Do your thing, I'll act super upset and distraught in my next post. If you aren't a troll, message me, I'm here if you need someone to talk to.
And also... ya motha!
They definately consider it a part of the game. How and when it is used is up to players.
The reason I am confident saying they consider it a part of the game is twofold. First, they built the threat system in a way where it is possible, despite knowing how to building in a way where it isn't possible.
Second, when confronted with the excessive use of it in game, their answer wasn't to make it impossible, but rather to state thst they will fix a bug that makes it easier than it should be.
So, imo at least, us players should be working on the assumption that Intrepid intend for this to be a part of the game.
So your going to gather a bunch of data on players and alert GMs so you can what? Ask them nicely to stop? The whole point of gathering data like that would be to ban someone, either temp ban or perma ban.
You don't need a bunch of data to ask someone to stop being a dick. You can just do that.
You could be a Sociologist/Psychologist interested in the data for some research project, but you probably wouldn't use a forum to make such a request, and the form your post takes is that of umbrage with the fact that people are disrupting gameplay in dungeons, not someone interested in the long term sociological effects that running trains (not that kind) in dungeons on a game not yet released to the public has on dungeon participation levels in dungeons in said games.
So, yes. I read between the lines and concluded that although you did not specifically ask for people to be banned you took all the steps one might require to ban someone, and as such want people banned for the perceived slight.
I actually liked this response, thank you for finally being articulate... aside from the part where you continue to draw assumptions about my intentions. Once the GM is alerted to the activity, they can review it, continue to monitor player activity, and do whatever they think is appropriate, kind of the point of being the GM.
Side note, I don't want anyone banned, but I am in no way averted to advocating that GMs have the tools and ability to do so with adequate information to back up their decision. I simply want people to be able to functionally test the dungeons, the mobs, the pathways, and the bosses. Training mobs can prevent this, quite frequently at times. Please correct me if somehow my view of training being disruptive to that (during a playtest) is incorrect.
Still no quotes to back up your previous statement. I don't think there is anything else to say to you at this point...
Other than... YA MOTHA!
I agree, I know my guild simply expects it to happen and we have our methods to attempt to stop them, doesn't always work, but to me I view it as a current aspect of gameplay.
Now whether it is intended to be that way, I don't know, I don't work there, if it ends up never changing, again... nothing really lost here.
Fair points on considerations regarding training being an accepted aspect of gameplay. I also don't ever want to get rid of accidental training or the occasional mad-hatter running through the dungeon causing absolute mayhem with a train of insanity behind them... always makes for a crazy experience, awesome if you somehow come out the other side without dying.
The griefing aspect of it, is what I hope the GMs are handling. I have dark and darker to play if I wanna feel a constant stream of abuse going into a dungeon lol.
Contet exclusion is supposed to be a part of the game. The discussions about instanced vs open world content covered the notion that non-instanced content would absolutely result in people excluding others, and the response from Intrepid was basically "that is the point". It was talked about as a risk vs reward thing. If people wanted to spend time excluding others from content instead of progressing themselves, then they are opting to risk their time for the reward of stifling the progression of rivals. If you want to progress, you need to get stronger than those that are in your way.
You are free to agree or disagree with that, but that is essentially Intrepids stance on blocking content.
To me, using an intended system to return an intended outcome is not and can never be griefing.
So, the question I then have for you is - if we assume that content exclusion is an intended aspect of Ashes, where is the line for griefing?