You mean this thread can't possibly only be here out of the goodness of the heart?
No.
I'm shocked. Almost too shocked for words.
Proposal Discussion: Things that could help player groups identify a possible bot attack
- Thread starter CMDR Jal'Bur
- Start date
You mean this thread can't possibly only be here out of the goodness of the heart?
No.
I'm shocked. Almost too shocked for words.
For the record: I'm against botting, cheating and anything that is against the rules of the game.
But after Carcosa I don't trust the people who brought this up further than I can throw my carrier at earth gravity.
The witch hunts are already starting, a peer pressure is constantly built up by raising (questionable) numbers of "law-abiding citizens".
And again, referring to Carcosa, I know what a disinformation campaign and witch hunt can turn into.
It nearly broke my will to play this game any more.
This gameplay is definitely more interesting than simply spreadsheeding who has the most CMDR-drones (not bots) running stuff.
But I still hate BGS work. Biggest mistake of FD imho was raising the stakes so high by allowing CMDRs so much connection to the minor factions.
- first of all, i'm not one of those anarchists in the game, being one of the filthy independent cooperatives. my main gripe with the OP comes from asking for commander identifiable information in 3 of 4 cases - having experiences with stalking and doxxing in other games.
- and i'm also aware, whether a request for "more data" matches proportionality of the cause. I read in this thread until now no good reason, why the data requested would help identifying bots (i don't know enough about bots to know which kind of data would allow identifying them more easily - so I'm following the discussion).
-as somebody spending a lot of time once in a while to understand certain aspects of the BGS, i love more data available. Ian D.'s massive data collection on colonia for exampel has created previously unknown knowledge about how markets work.
- as somebody helping a group managing their >50 system imperium (which has signed the AB-A asap), i'm happy about data being available. but i can't say more data would be helpfull for that.
- the only time more data would help is when things do not go as planned - for forensics. and i can clearly understand concerns, that more data would lift the fog of war. so after following this thread for a while now, i think concerns of balancing if opposition numbers of commanders and tactics would be easily to spot are valid. there are games like chess or go, which work without the fog of war. there are other games which work with fog of war. imho elite is the latter.
- a compromise could be, to delay more non-cmdr-specific data - for exampel, if you got a more detailed traffic and trade report with the weekly server maintenance. if the data has any relevance in spotting and reporting bots, it doesn't matter, whether you know it today or next week.
Last edited:
In my experience the vast number of these accusations in games are false. In another game I played it was so common for players to just say whoever beat them was a cheater without evidence, it became a meme: "HACKusations" the community dubbed it. If you said someone was cheating you were laughed at and dismissed, and we were all better for it to be honest.
If this is so common and "obvious", where is the hard evidence? I see none being presented here.
Last edited:
'Hard' evidence goes to Frontier confidentially, even the reporter wont know whats happened.
Hard evidence posted here would get Moderated so fast under so many breaches of TOS.
We are not the judge nor jury. Witch hunts and online abuse are real.
The topic is basically 'how' do we / players get that 'hard' evidence in the first place to raise a ticket, Im all for barring people making too many false accusations from making any more. The system relies on people reporting to draw Frontiers attention to a potential offender so they do need some info beyond outcomes alone.
Other people want more info for their own satisfaction, some maybe for BGS, you can play Checkers or Chess on the same board but maybe not at the same time seems to be the issue.
So as usual its how much security / privacy do you give up for more security or transparency? What info is 'acceptable' and what isn't?
Hard evidence posted here would get Moderated so fast under so many breaches of TOS.
We are not the judge nor jury. Witch hunts and online abuse are real.
The topic is basically 'how' do we / players get that 'hard' evidence in the first place to raise a ticket, Im all for barring people making too many false accusations from making any more. The system relies on people reporting to draw Frontiers attention to a potential offender so they do need some info beyond outcomes alone.
Other people want more info for their own satisfaction, some maybe for BGS, you can play Checkers or Chess on the same board but maybe not at the same time seems to be the issue.
So as usual its how much security / privacy do you give up for more security or transparency? What info is 'acceptable' and what isn't?
... which needs to be argued in the first place. i still remember the peculiar case, working in switzerland, when the local conservatives (!) voted against (!) more security cameras in public places, not because privacy, but because they deemed it ineffective in terms of reducing criminality. it's not like more data = more security necessarily. or the discussion around encryption for raising security, or breaking encryption to do so.
Frontier runs the game. I don't buy this narrative that they need player input and reporting exclusively to do something about this problem or to even be aware of it. Trust me, they know more than you about what's going on in their own game.
This
seems to me like evidence anybody could sent to FDev to have a look into. No need for more data.Out of curiosity, how do you know that the bots are failing missions, pax or otherwise? The op let me a while back in on the "secrets" of tracking the trade stuff and using the leaderboards to analyze inputs, so I got a good idea how that is done. That conversation is also one of the reasons I'm against more data, be it extra or existing more granulated. I'm genuine interested in the answer as I can't see where failed missions would show up. BM trade is another one as it doesn't show up anywhere either.
I'd expect FD do the analysis on this stuff - not players.
Interesting point. I've been playing Mechwarrior Online recently and shot another mech in the back doing enough damage to kill it - the sheer amount of accusations of hacking immediately afterwards were shocking
Many people can't tell the difference between standard gameplay and 'hacking' - or botting in this case
There are some people who believe that if someone is able to better them it can only be by cheating...
This forum is no different in attitude to that.
(I have to do that post on BGS forensic at sone point...)
- negative actions lead to a different structure of influence gains and losses, than positive actions for all other factions:
- as losses are distributed as gains relative to influence in system, low influence factions gain less from losses, than factions of higher influence - and even that in a predetermined way.
- while gains distributed as losses by positive actions for all other factions lead to low influence faction gaining more than high influence factions, and can't really mimic the other.
here an exampel calculation:
System with 4 Factions A,B,C,D.
Influence:
A: 70
B: 20
C: 9
D: 1
exampel I: applied negative influence of -9 to faction A
leads to:
A: 64,5
B: 23,7
C: 10,5
D: 1,2
exampel II: positive influence applied to B, C, D +3 each at original value of 70,20, 9, 1:
leads to:
A: 59,8
B: 23,9
C: 12,3
D: 3,8
... if you watch out for D, you can easily tell negative and positive influence actions apart
System with 4 Factions A,B,C,D.
Influence:
A: 70
B: 20
C: 9
D: 1
exampel I: applied negative influence of -9 to faction A
leads to:
A: 64,5
B: 23,7
C: 10,5
D: 1,2
exampel II: positive influence applied to B, C, D +3 each at original value of 70,20, 9, 1:
leads to:
A: 59,8
B: 23,9
C: 12,3
D: 3,8
... if you watch out for D, you can easily tell negative and positive influence actions apart
- during lockdown many other options for negative influence actions are disabled, so failing passenger missions is one of the few reasonable ways to go about (- if you decide to go about; i personally wouldn't as that feels gamey to me: you have to fail a mission, so outside of having it run out of time, bumping into a station for hulldamage with passengers on board, which will leave the ship after some is a common technique. it will cost your rep, and you'll lose access to more missions, so can't be repeated often - if you don't have a lot of commanders or accounts doing so.)
... so that's how people conclude, "what we are seeing despite lockdown is negative influence effects. this negative influence effect can more or less only come from failing missions. failing missions means probably failing passenger missions." - and probably, with some other numbers, like 200 condas on the traffic report, come to the point concluding that what they are seeing can't be for legit gameplay.
Last edited:
Thanks for the detailed answer. So it's more or less "educated" guesses than hard data that would lead to somebody concluding that the other party is failing missions on mass. Of course the guesses are only as good as the understanding of the mechanism themselves and more important their interactions with each other. But I would also bet on failed missions being a big part of it.
I still don't see how a more detailed system report, short of listing what was done for each faction pos/neg, would help detect bots. And then we are in the territory that the ones with more sophisticated out-of-game bots will have not only an advantage to direct bgs actions more efficiently as they do already but more important the data to find out what the opposition is really doing. That would be the day I would quit playing the bgs as the fun of finding new approaches will be removed and it will just become a who can grind the most game.
From a bot operator perspective it doesn't matter if they send 200 condas once a day, one 200 times, ten 20 times, or 25 eight times the work they do is the same, just the traffic reports would look different if they were in the overall number/individual ships or CMDRs to now.
The "equal %" neg inf distribution is what I saw during my testing of some bgs mechanics not that long ago, alongside some more interesting stuff which needs more testing. I also more or less confirmed the 2k/t profit diminishing return for myself and it was interesting to see at work. Science is the most fun part of the bgs. The failing missions is a stable in what we do now, just like for everybody else. We do use an approach a little different to the pax missions stuff though. Seems to be easier, less costly, and mostly more efficient.
Example II could also be neg against A and some pos for faction D, the result would look exactly the same.
Ok so thank you all for your constructive feedback...
On the discord for the ABA the following idea was presented and discussed by a number of users. It is a different take on an idea for how frontier could deal with botters in BGS and PP. I shall copy it here for further discussion.
Anti-Botting, Telemetry & Verification approach
Suggestion for FDev
Scope : Permanent Telemetry is possible for them AFAIK - but it costs Performance and Server resources, for all I know it can also degrade Gameplay experience (lags). [ taken from memory from an FDev Forum post a >very< long time ago ] As such, it is assumed that permanent Telemetry on all Clients is neither desirable nor feasible.
- Phase 1 :
Selectively, a reasonable amount of Clients are assigned or moved to a Telemetry Server for a limited and fixed duration. Sole goal is to perform detailed spot-checks on these Clients while minimizing additional Server loads. Trigger for candidates of these spot-checks shall be a sufficiently significant amount of dedicated BGS- or PowerPlay activities (rationale : those are prime candidates for botting negatively affecting legitimate Players).
- Phase 2 :
Candidates that show no anomalies are dropped and cycled out. Candidates that show unusual and irregular (suspicious) behavior are moved into an investigation bucket. Some basic details from Phase 3 used to determine suspicious Clients.
- Phase 3 :
The investigation bucket is specifically subjected to a longer check of Client behavior in order to perform a valid verification of the suspected abnormal Client behavior.
Telemetry targets are a combination of :
- Phase 4 :
FDev (humans) analyze the presented remaining bucket of "high chance of being a bot" Clients and associated bot scores - and takes appropriate action. Lessons learned from the attained, detailed Telemetry Data flows back into optimizing the detection parameters for Phase 2 & 3, this way FDev can also react to any changes to existing or future Automation Software.
One way to catch Bots would be to implement a very basic "Trap check" on Clients that are subject to Phase 2 or Phase 3 testing :
As AI is officially banned in ELITE lore, one could eventually take this a careful step further and implement occasional but various Inbox-related checks (the most friendly in-game version of a dreaded Captcha). A tiny variation of "Please complete Pre-Flight Checks before being permitted to Launch" so to speak that comes in at random and only for suspected Clients. Having a variety and keeping things changing over time could be the easiest to implement "bot stopper" that throws a wrench into its machinery, repeatedly rendering it useless (on top of highlighting the abnormal Client behavior - repeated failure to pass this simple test of variation can also be used to throw Clients into the "highly suspect" bin right away).
Absolutely minimal effort - big effect.
On the discord for the ABA the following idea was presented and discussed by a number of users. It is a different take on an idea for how frontier could deal with botters in BGS and PP. I shall copy it here for further discussion.
Anti-Botting, Telemetry & Verification approach
Suggestion for FDev
Scope : Permanent Telemetry is possible for them AFAIK - but it costs Performance and Server resources, for all I know it can also degrade Gameplay experience (lags). [ taken from memory from an FDev Forum post a >very< long time ago ] As such, it is assumed that permanent Telemetry on all Clients is neither desirable nor feasible.
- Phase 1 :
Selectively, a reasonable amount of Clients are assigned or moved to a Telemetry Server for a limited and fixed duration. Sole goal is to perform detailed spot-checks on these Clients while minimizing additional Server loads. Trigger for candidates of these spot-checks shall be a sufficiently significant amount of dedicated BGS- or PowerPlay activities (rationale : those are prime candidates for botting negatively affecting legitimate Players).
- Phase 2 :
Candidates that show no anomalies are dropped and cycled out. Candidates that show unusual and irregular (suspicious) behavior are moved into an investigation bucket. Some basic details from Phase 3 used to determine suspicious Clients.
- Phase 3 :
The investigation bucket is specifically subjected to a longer check of Client behavior in order to perform a valid verification of the suspected abnormal Client behavior.
Telemetry targets are a combination of :
- UI timing patterns, i.e. fixed and long-term stable delays between standard UI interactions and/or using basic Ship functions (showing no human/normal variation in terms of interfacing with the Game), indicating the Client is likely operated by a Bot instead of a human
- presence of maximized use of in-game automation (Adv. Docking Computer and SCA fitted AND activated in a narrow timing pattern, in line with UI interaction anomalies)
- presence of fixed loop cycles (i.e. during a Session, only a single major BGS or PowerPlay task is performed very often by the Client in a stable pattern and showing no expected variation)
- abnormal and fixed pattern to typical gameplay-interfering in-game events (i.e. Client behavior after an NPC Interdiction) - fixed Flight Control Inputs : Client is controlling the Ship in a digital/robotic pattern, showing no normal (human) variation to basic flight inputs and navigation
- Optional/Supportive indicator : unusual times of Operation (i.e. during the entire night of the respective time zone - Bots need no sleep)
- Optional/Supportive indicator : i.e. Client exclusively in Solo (Bots will normally never be operated in Open Play)
- Phase 4 :
FDev (humans) analyze the presented remaining bucket of "high chance of being a bot" Clients and associated bot scores - and takes appropriate action. Lessons learned from the attained, detailed Telemetry Data flows back into optimizing the detection parameters for Phase 2 & 3, this way FDev can also react to any changes to existing or future Automation Software.
One way to catch Bots would be to implement a very basic "Trap check" on Clients that are subject to Phase 2 or Phase 3 testing :
- i.e. after a certain amount of times docking during a Session (bots doing what bots do best), a simple "Acknowledge this to continue" Message is sent to the Inbox of the Docked Client
- the Message must be opened and acknowledged before being permitted to launch the Ship again, with a short but reasonable expiration timer (i.e. <60 Minutes); this event must not enter the Journals
- the acknowledgement can consist of a basic Checkmark Option (valid choice) amongst several "X" Options (invalid choices) in random order (= no fixed sequence of UI Inputs)
As AI is officially banned in ELITE lore, one could eventually take this a careful step further and implement occasional but various Inbox-related checks (the most friendly in-game version of a dreaded Captcha). A tiny variation of "Please complete Pre-Flight Checks before being permitted to Launch" so to speak that comes in at random and only for suspected Clients. Having a variety and keeping things changing over time could be the easiest to implement "bot stopper" that throws a wrench into its machinery, repeatedly rendering it useless (on top of highlighting the abnormal Client behavior - repeated failure to pass this simple test of variation can also be used to throw Clients into the "highly suspect" bin right away).
Absolutely minimal effort - big effect.
Last edited:
Excellent suggestion. Anything requiring a little skill! I would love to see some bgs stuff involving getting into and hacking installations without being scanned. That would be hard to bot.
It's overthinking it by orders of magnitude and trying to micromanage the process FDev should take to solve this.
The telemetry stuff isn't as simple or straightforward to implement as it seems because as you describe it you'd need to define "robotic" unambiguously as to look for automatically in a mess of data. I'm willing to bet you'd get almost the same results at looking at accounts who are outliers in influence/powerplay effects achieved without that much extra code.
Assume any trap/CAPTCHA works once and then becomes ineffective so it's only good if you have a large amount of different bots to confirm/purge. Otherwise it's just security theater.
The telemetry stuff isn't as simple or straightforward to implement as it seems because as you describe it you'd need to define "robotic" unambiguously as to look for automatically in a mess of data. I'm willing to bet you'd get almost the same results at looking at accounts who are outliers in influence/powerplay effects achieved without that much extra code.
Assume any trap/CAPTCHA works once and then becomes ineffective so it's only good if you have a large amount of different bots to confirm/purge. Otherwise it's just security theater.
I wonder if perhaps the reason FDev doesn't engage with this is precisely because it's very easy to infiltrate such a discord and put together ideas on how to avoid counters to bots, since this idea is something that someone could mitigate through adding something to the program. By mentioning this you've already given the game away, so to speak.
This to me highlights the issue I, and I'm sure others, have with the wider discussion about bot prevention. One can be a player in a game, or a referee. You can't be both- something about conflicts of interest.
So, conversations about this need to happen privately with Frontier, not on a public forum where bot users can read your ideas. I'd suggest that you've some relationship building to do there as well because that PC Gamer article, as well intentioned as it may have been, brought them a measure of bad press that they could have done without. If it was me I wouldn't talk to you again for fear of things being leaked.
So keep things private, work with Frontier, not creating problems for them and hopefully bot users will have a far harder time here. On that point I genuinely wish you luck.
This to me highlights the issue I, and I'm sure others, have with the wider discussion about bot prevention. One can be a player in a game, or a referee. You can't be both- something about conflicts of interest.
So, conversations about this need to happen privately with Frontier, not on a public forum where bot users can read your ideas. I'd suggest that you've some relationship building to do there as well because that PC Gamer article, as well intentioned as it may have been, brought them a measure of bad press that they could have done without. If it was me I wouldn't talk to you again for fear of things being leaked.
So keep things private, work with Frontier, not creating problems for them and hopefully bot users will have a far harder time here. On that point I genuinely wish you luck.
This is another really stupid idea...
Firstly demanding that Frontier create a whole slew of code, checks, and measures as well as create special 'telemetry' servers is a total waste of time and effort. Frankly little enough gets done on this game that we would possibly want Frontier to spend any of their development time on this when there are so many other things with higher priority.
Secondly the measures could easily be countered - all it would take is one human monitoring these supposed bots to counteract any of these stupid misguided anti-botting measures.
Thirdly any form of measure that is anti-botting will, unless incredibly well thought out (and the people who are behind the anti-botting alliance have shown nothing well thought out at all), be more of a hindrance to legitimate players than to bots.
Just stop it - there's no proof of bots being available, there's no proof of the imaginary bots being used on even a small scale, but there's plenty of evidence of confirmation bias amongst 'anti-botters' based off of their 'we're losing so it has to be bots' mentality
I looked up that PC Gamer article, didnt know it existed until here. Might be just me but I thought Frontier response was slightly above the normal and less wishy-washy. They also made the point that they want to keep their methods discrete as they dont want to tip-off anybody about their methods. Which seems sensible. And is also the reason they dont announce how many caught and banned at all ever. Which is bloody irritating to the player. Not sure Id do it different myself if I was confident I was secure and catching most cheats.
Been thinking about the 'game of spreadsheets'. Isnt this just the same as any meta-game in anything though, I want to play on 'this level' and you want to play on 'that level' and I hate it but cant actually do anything about it as its an MMO. To me its the same as a meta-build ganker blowing up a non-meta casual player. Both want to play on different levels and the game allows that but clearly gives an advantage to the meta-gamer. It always does in any situation. So to apply that argument here opens up all the other meta-imbalances in game, to me, in my opinion, after thinking about it.
On the other hand, more stats the better for me, if people use them against me well so be it, I could do the same back if I wanted or try and counter it. Thats the meta-game. Or I could just do my own thing and ignore them. If they use them to highlight accounts and notify Frontier then as long as theres checks and balances on 'malicious' reporting its just the same as now, some people just think it will be easier. If they find out its active actual players instead then they have to accept bots not real and you get to laugh and say 'told you so'. Either way we all know the game is 'clean' which surely everyone wants?
Been thinking about the 'game of spreadsheets'. Isnt this just the same as any meta-game in anything though, I want to play on 'this level' and you want to play on 'that level' and I hate it but cant actually do anything about it as its an MMO. To me its the same as a meta-build ganker blowing up a non-meta casual player. Both want to play on different levels and the game allows that but clearly gives an advantage to the meta-gamer. It always does in any situation. So to apply that argument here opens up all the other meta-imbalances in game, to me, in my opinion, after thinking about it.
On the other hand, more stats the better for me, if people use them against me well so be it, I could do the same back if I wanted or try and counter it. Thats the meta-game. Or I could just do my own thing and ignore them. If they use them to highlight accounts and notify Frontier then as long as theres checks and balances on 'malicious' reporting its just the same as now, some people just think it will be easier. If they find out its active actual players instead then they have to accept bots not real and you get to laugh and say 'told you so'. Either way we all know the game is 'clean' which surely everyone wants?
It's about the data, the way the data is pulled and processes, and most important from there the data comes. Some of the data pulled is on the edge if not over to be in violation of the Eula(maybe GDPR too). Also the ones most active in this threads are the ones who use extensive automation (bots) and data harvesting in and out of game to get a competitive edge. Some also use known exploits in game (player bounties, slf's in CZs) and Eula violating shared accounts for bgs stuff (explorer accounts). Clearly some double standard here.
The meta game is to find out how the bgs works and how to leverage it to your advantage. Giving more data removes that as anything you find the opposition can see in the data. Part of the appeal for bgs activities for me personally are the theories, validation tests, and deploying the theory in the wild. Doing the actual bgs stuff in game is the boring part of it.
Especially this proposal (1st post) seems to me more from somebody looking to get there hands on more data to figure out what the opposition is doing, instead of doing the hard part and start thinking outside the box. Looks to me some people rested on their laurels for to long and are now left behind. Instead of trying to catch up the hard way we have a witch hunt and "requests" for more data. Then there are some people unsolicited DMing me and gloating about the amount of they data they "gather" and the way they process it I would rather see less data being available. It would bring some more fun into bgs activities and most likely make it harder for large empires to grow. That looks like a win-win to me.
P.S.(not for you Vetinari): And no, you can't teach me more about how the bgs works as you are one of them who got left behind. With the amount of people and time you needed to flip Carcosa I wouldn't gloat about it. A monkey could have done it faster. It worked out well for SEPP/EN though, didn't it?