You misread the whole thing. Continue with the suggestions instead.
It’s also a forum and not a questionnaire 
1 Like
I do know it, my team is going through it, I see the toll and the cost every day. That is why I said “sometimes” (we’re a bunch with bad luck, part of that “sometimes” lol). There are other solutions too, which actually fit both sides more. Like, ending an old contrat of way of working, for a better one with a lesser cost. You already have the budget, but are now saving thanks to the new solution, and not on the employee’s back. It does exist too.
I will read the rest of your post and comment later, been up since 4 am for work and can barely think straight in French. See you
2 Likes
Thanks for the detailed reply, @E.y . 
I don’t have the time to finish reading everything now, but later I will check the topic.
Just have the servers run AI powered GM bots that can analyze building piece count, land claim and player behavior. Train the AI on real world server map save examples so the AI can learn what’s acceptable and what’s not. Then once it’s working right have it start warning players and then recommend bans to human GMs for repeat infractions. No need to hire more employees just spin up needed server resources as needed. Problem solved.
Consider the current amount of Official servers and the labour involved in a proper moderation system for this type of game.
Currently there are 780 Official Servers.
By last count based on my own tickets and what I’ve seen, there’s a general reviewer and then 1 for each God save Derketo and Crom. I’ve seen Yog, Mitra, Ymir and Zath. That means there’s 5 in total handling Zendesk tickets. Give or take.
5 reviewers handling 780 servers and we know nothing about their internal policies in handling Crom coin issues or how much of that labour is devoted to which types of tickets.
If anyone wants to correct me on additional reviewers please do so.
1 Like
See above post. Use AI
Interesting suggestion for sure but by their own admission and ToC violations have no correlation to building piece count.
It would also require a budget.
1 Like
Of course but I think the budget would be a fraction of the cost of hiring abunch of employees
1 Like
This + human error are just some reasons why manual moderation is an awful idea form a user experience perspective, as well as financial perspective. This is why a good game design is needed with “moderation” via in-game mechanics only (in other words, built-in restrictions which are an integral part of any game).
EDIT:
We’re running a 10k members DayZ server. Moderation is such an absolute nightmare. It’s simply impossible to handle everything and make everyone happy. Even when we simplified rules to a bare minimum, nobody was reading it and I don’t blame them for it. Video games should pose restrictions by design to make sure the gameplay is fair.
We’ve developed our own in-game mechanics that prevent people from abusing the systems that normally cause a dispute and make people quit. Reports dropped to almost none and everyone are happy, despite us creating limitations via in-game mechanics. It’s a good trade-off, just takes some getting used to.
And again - these limitations via in-game mechanics should only be required on official servers.
2 Likes
100%
It could remove the possible need for the subset of pay to play Officials, free up the current staff to focus on Crom coin tickets and only focus on the cases that do need attention for sure.
I like the idea of warnings ^^
1 Like
For sure, it’s always been a poorly done system. I doubt even casual bystanders like it.
It doesn’t mean we have to have block limits and for the reasons listed and more it wouldn’t work like you think it would.
Yea, it doesn’t have to be a hardcap block limit. There’s more of ways to solve it. Just needs some brainstorming by the design team.
Just that with this system they have to make sure to encompass all cases. Block limit will not stop people from spamming foundations around someone else’s base. There are plenty of interesting solutions that could be implemented when it comes to land claim. Many of them existing already in other games.
Because Funcom already employs a bunch of people who are AI experts and know exactly which DNN model to use and how to transform the game database into inputs that can be used by that model, and because the computational resources needed to train that model’s weights are really cheap? 
3 Likes
Less a correction, more so a history. I remember some time ago Umborls said they had hired extra help to conduct moderation, no idea what that would put the team size at. I also remember that Exiles moderation is not the only job function (maybe not even primary) of Umborls, and perhaps the entire team under them. I’ll try to find evidence of my memory.
All that to say the numbers are likely even more bleak than you suggest.
Edit:
Oct 27,2022
Mar 2022
3 Likes
I and others have suggested an upkeep system instead, solving serial refreshing and land claim problems together. Possibly.
The problem with anything like this is that it creates a need to fundamentally change the code and how the game is played.
Three questions come to mind:
Does that align with Funcom’s vision? We don’t even know what their intentions are.
Do they have the budget for it?
Are players going to be ok with this?
1 Like
You say this as if AI isn’t the future of gaming and software dev though lol. Big names are already working on AI tools/APIs etc to simplify the adoption of it by game devs . It’s definitely something to consider.
Even just having monitoring/grafana dashboards with alerting based on live server DB queries isn’t difficult to implement from a ops point of view (though I’m sure database changes on game servers would be needed to facilitate queries). Maybe a custom game server log messages indicating certain player behaviors that when a certain threshold is reached a GM gets an alert to proactively check a server or a specific region. And when a player report AND an alert correlate then more weight to a ticket is given
1 Like
Yeah, but we’re really far away from AI being “a fraction of the cost” of having better human moderation practices
Sure, from the ops point of view, that works. I’m not sure how hard it would be from the dev point of view. Their game database is SQLite, so any process doing queries on it would have to live on the same machine as the game server.
But you don’t really need live queries. The servers reboot every day. They could just copy the database during the reboot and have a tool that inspects that copy and flags potential infractions for human inspection.
EDIT: I want to make it clear that “just” having a tool that inspects the database and flags potential infractions is still a considerable lift. It really isn’t easy to code logic like that, at all.
3 Likes
I understand your request (or that of the community) and I am almost nearly accepting it but I remain septic on Funcom’s ability to provide us with an Adequat service.
I paid a subscription every month to play Final Fantasy 14 Realm Reborn, I see or go my money! In addition to a clientelle service that holds the road, from an efficiency server, Square Enix puts content regularly and this without destroying or calling into question the existing content.
If threw on that giving money to Funcom would actually serve the players of the players, then yes: but I have a big doubt about it; The money irrelated probably in particular in the development of dune
1 Like
True, human moderation action can be improved, AI could help take the workload off human Gms though. Funcom could hire an AI dev to chase down AI use case scenarios etc but tbf it’s hard to speculate what headcount costs would be without knowing general salaries or the country they would be working in. But 30 GMs I imagine even if entry level salaries is a bit much. Servers/software are of course another cost too but easier to stomach than say laying 30 ppl off down the road.
AI is ultimately an investment that can play into other things. Like next level thralls/npcs, a hidden karma system that triggers in game events, AI powered npc clans that play like players etc but I get a board not wanting to approve large project expenses unless it ties into other things. I guess it just comes down to cost and scope.
On the ops/devops side if changing dbs or sending data to a more capable db isn’t an option then logging could def be used and from there monitoring/alerting etc. no matter how the cat is skinned though I think a more proactive monitoring posture of on server player actions could give current GMs better insights on where to swing the ban hammers. Anything to help GMs act effectively is a win for everyone
1 Like