Okay, so flagging is on the table, just outside of your immediate suggestion?
Yes, by all means.
I just havent gotten that far to explore/consider that since Im still dealing with other objections/posts further down the chain.
My most difficult challenge, is getting people to realize the paradigm shift Im proposing. That instead of running around after bots after the fact, to instead focus on preventing them/their efficacy in the first place.
If this mechanism can be used to flag botters in addition to this, Im all for exploring that. Just havent gotten that far yet due to the volume of other posts Ive had to deal with.
Badly backlogged with guys claiming my use of the term “epidemic” was wrong and another guy claiming that I was ridiculous by suggesting any and all means should be used, to mean that shutting down the servers is the logical conclusion of that. Guys that dont understand that I mean a bot-test that uses the versatility of the EVE UI/client, rather than the “captcha” of text/number/image clickibg they are used to in web browers. Guys that think preventing bots just makes them stronger (somehow) Guys that immediately refuse discussion on it simply cos they dont want a bot-test inconveninencing them etc etc.
Getting everything including the kitchen sink thrown at me.
Okay.
I was mostly addressing your third point:
Bot-like behaviour needs to be tested/challenged, there is no way around this in your system. This does mean “bothering” players who are AFK or semi-AFK while doing things bots do (I’m not attacking all AFK here) will be a necessary side effect of the system.
As I already sated, looking at other in game activity should be the first thing to prevent “bothering” active players, they should never see your captcha.
If there is no other in game activity, then the player is playing in a more bot-like manner, and they should see your captcha, but it doesn’t have to be in an obtrusive manner.
I will example with highsec courier mission botting (because it’s an easy example, but the ideas can be extended to other bot activities also) This is not about flying using AP, but actually issuing jump commands, which is therefore either semi-AFK, or botting.
- You are only doing the task at hand, nothing else in the game, so you are flagged for a captcha challenge
- A gate along your route is randomly chosen by the system.
- When you land though the gate (which conveniently provides a sound that a semi-AFK player would be listening for, if not watching) you have a 60 sec cloak.
- The captcha comes up at the start of the cloak, until you solve the captcha you cannot warp out and continue your route.
- You then have those 60 secs to solve the captcha and start your next jump (I’d be okay with a longer cloak even if you have a captcha).
- if you solve the captcha, great, you’re back on your way. If you continue to play like a bot, you might get another one in an hour or so, but this is not an every gate thing.
- if you fail the captcha, you now uncloak, and are sitting stopped at the gate. I’d even be okay if you then went suspect.
You can’t move, or lose your suspect flag (if implemented) only after you solve a captcha.
Other botting / bot-like situations would need their own set of rules and consequences but the ideas could transfer to them.
It isn’t really. People use that phrase way too much for things that aren’t even close to a real “paradigm shift.” But, this is an aside and doesn’t need more attention here, just a pet peeve.
This is also a word that gets misused, especially regarding discussions on this forum (any forum?). 
If you are doing things with the kitchen sink, then you are AFK or semi-AFK and you should be bot tested.
Understood. Ive been bogged down in challenges 1-2 so far. My reaction to your earlier post asking about that was unclear. I should have laid out my proposed roadmap earlier.
This is an interesting notion, but as I pointed out before, the nature of bots as scripted, is they can emulate that behavior, such as by periodically opening the ingame map, saying Hi in Local etc.
Again, as I said above and earlier, the nature of bots is such that they periodically send commands to CCP servers, as if the the player was present.
The inverse result of your proposal, if I read it correctly, is that only genuine afk players (non-botters) would receive the bot-test (because bots are not considered afk by the EVE servers, as they send commands, whereas a truly afk player does not).
For example, for a drone ship (or other autotargetting ship, especially missiles), its very difficult to ascertain whether the player is present or afk. Even more difficult to ascertain if that drone ship is being run by a bot.
This leads into the meat of the 3rd challenge, which is what activities/behavior can the bot-test be applied to.
That was one of the inspirations behind my impetus in this direction, especially inlight of the recent thread with reports of a dozen accounts running them 23/7.
I would propose that this is one of the instances where a bot-test is applied each time someone tries to activate the courier mission. If you cant pass it, you dont get the mission.
(Edit: Doesnt necessarily need to be everytime, Sufficiently frequent as random should enough to screw the 23/7 cycle, at least somewhat till the botter returns and manually completes the outstanding bot-test.)
In that instance, players managed to fk the botter by opening a trade window.
But the botter/scripter can program the bot to detect that pop-up in a predictable place on the screen, and to click past it, so the rest of the script can run.
My bot-test would severly impair that.
On the issue of AFK, by whatever form, activating a bot-test, Im wary.
Players are allowed to afk, nor does it mean they are botting.
The paradox there, is that a bot doesnt “afk”, only players do.
Ill think on this more and perhaps you can elaborate on how you propose to detect the difference between a player that is afking whilst doing nothing, and a bot that is never afk whilst the player behind the screen, is.
My bot-test should help that distinction, but Im unclear on your suggestion of “other ingame action” as a means of determining where/how/when the bot-test should/could be applied.
1 Like
Humans don’t normally do things “periodically”, this is where you are missing the point, it is very hard to script something that looks like real active human behaviour.
No, because they aren’t engaged in a bot-like activity.
They would need their own set of “rules of engagement” for the anti-botting algorithm. I haven’t suggested one yet, but it would have similarities to the one example I gave.
Okay, they don’t get the mission, but nothing else happens. I could still take the mission, and then start the bot. Death in space is much better, and more EvE-like. Also it wouldn’t effect bots running trade routes.
Right, it doesn’t mean they are botting, but just because they are allowed to doesn’t mean CCP can’t ask them if they are botting, nor does it mean that their can’t be added risk to that allowed activity.
Unless you are talking about an actual robot, I’m sure that there is nothing actually at the keys issuing commands.
Active = a human at the computer interacting with EvE like a human would
Semi-AFK = a human at or near the computer doing non-EvE things, but auditorial or visually aware of what is happening in EvE.
AFK = no human at or near the computer or paying attention to what is happening in EvE.
My suggested system would test those who are not “active” but interacting with the game (giving commands, mining, ratting, etc.), semi-AFK players would easily pass, bots and AFK “players” would fail the test.
So yes, if you are a real person and you leave your auto-targeting ship ratting while you take a shower. Then yep, you might get tested as a bot and fail that test, you might lose your ship, and it might flag your account for review by a human GM to see if you are a bot or not.
That was true up until about 2 years ago or so. In another game I played at that time, which was action oriented combat BTW so much faster and harder to do, and with each toon a vastly bigger array of abilities then ships here have. It got plagued by those types of PvP bots. Those bots had to use an overlays for inputs into the actual game. So what the devs did there they disabled the game whenever something would attempt to put an overlay on it. Not entirely, but as soon as any sort of overlay was used, even teamspeak or steam, game was minimized automatically and the client would pause. If overlay was not shut down the client would stop working entirely.
The AI is actually tiny, remember its only for a game and only for specific parts of it, for example combat, or map/teleportation/movement, or buying/selling, etc. They give you the system resources it uses on their web pages.
And things have progressed since that time quite a lot. I would give links but its against the forum rules, but not exactly difficult to google, just use some of the advanced filter options. Anyone can look at the stuff.
Agreed, and theoretically this “should” be the flag that indicates a bot (even allowing for advanced bots that can even detect/warp to anomalies and engage, as scripted).
Afaik, this is data CCP looks at when investigating potential bots. I dont know what CCPs data/computational capacity is to automate that process out of all activity in EVE, so as to flag suspect characters for staff to investigate.
Theoretically, CCP could collate more data, more efficiently, with more computational power to crunch that data, to flag suspect characters, but they still need to be investigated individually by staff with limited time.
This is again approaching the issue from a remedial perspective, rather than a preventative one, which is my onus.
Right.
So a client that is sending no commands to CCP, can rationally be assumed to not be a bot.
Agreed.
Correct, you wont get the mission until you manually pass the bot-test.
Also correct, that after that, you could initiate the bot, per normal.
But you would have to be at your PC, more or less 23/7, manually passing those bot-tests, and initiating the bots therafter.
I agree, but we have to crawl on this issue before we can walk.
Its not unreasonable for the bot-test to pop-up occasionally right after initiating warp to a gate. Such that if you dont pass the bot-test during the warp, you will not pass the gate. There is ample time for a human player to complete the bot-test during the warp to gate. Autopiloting players/bots would not be affected by a bot-test, but also are forced to slow boat the remainder of the distance as vulnerable.
I think CCP can program the algorithm so that the bot-test wont affect fleets orbiting gates awaiting a jump command from FC, so that some random fleet members arent hampered by time taken to complete a bot-test, and thus not jumping in a timely fashion with the rest of the fleet.
If the client is sending no commands to servers, it can be assumed its not running a bot script, at least for that period of inactivity.
I personally would be ok with that, but I expect most will not (especially NS ratters, whom may or may not be botting). They will argue they are allowed to leave their ship ratting afk using ingame means (sufficient active/passive tank, ingame automated targetting/dps application, orbiting at speed etc) without botting.
But!
Then we had the recent Nyx bot fleet incident, which was farming like mad for who knows how long, before PLAYERS intervened.
In-light of that, its a pretty strong case that a bot-test should somehow find its way into PvE as well.
This is a fascinating benchmark case on bot behavior.
The bot knew to flee upon arrival of an unknown in its system, and to warp back, later, right back to what it was doing, but it didnt account for a bubble there. It also had no programming to engage players attacking it.
The botter in this case made fundamental mistakes, but again, this extent of repeated, perpetual botting would have been prevented, or at the least significantly reduced in efficiency, by my proposed bot-test, by forcing the botter to manually pass bot-tests, over and over, and hence be at his PC doing so. (Not sleeping, at work or whatever).
Even if he is infact present at his PC, whilst botting, for all his waking hours, that would still mean he has to repeatedlt check through the EVE client to see if there is another bot-test, after which he has to re-initialize the bot program, per account.
But everyone also needs to sleep. That is impossible to avoid. So for at least 6-8hrs per 24hr period, he would not be able to operate these bot fleets, due to not being at his PC manually completing the bot-tests. (Discounting account-sharing, which is itself a CCP EULA/TOS violation).
PS: Its a dead-giveaway of a bot or account-sharing, if the account actively sends commands to the CCP servers, day after day. Its humanly impossible for anyone to play EVE actively for more than 3x23hr periods without sleeping.
So even at worst, my bot-test proposal would shave 1/3 off actual/efficient bot time, from ~23hrs, to ~15hrs.
A 33% drop in bot output, is worth it.
That may be enough to make many individual bots or bot fleets unprofitable vs cost of PLEXing them, and thus cease or curtail botting operations.
No matter how many botters there are, all of their output will be reduced by ~33% at least, at no added effort to CCP Security team.
As this is an automated prevention model, it does not involve CCP Security staff time/resources/involvement to enact.
This frees CCP resources to investigate the remainder, as well as potential flags indicated by accounts failing the bot-test.
I suppose in some ways my bot-test proposal boils down to a humanity-test, as a factor of no human player being able to sustain activity 23hrs per day for longer than 3 days at a time.
Additional to that, many, if not most, will have real 8hr jobs and IRL requirements (daily chores, transit time, family, hobbies etc) whilst afk, thus reducing effective botting duration ,if impaired by my bot-tests, by another +8hrs, so 2/3 reduction compared to current bot uptime/efficiency.
Thus we could infact be looking at a >66% reduction in botting output, per bot.
For most people, in aggregate, its 8hrs sleep, 8hrs work and the remaining 8hrsnof that day is divided among everything else that is required by their IRL, with EVE time typically being what is left-over,
So out of a 24hr period, most people are not at their PC to answer bot-tests for ~16hrs/day + time spent on all other IRL concerns.
Weekends/days off/vacations are an exception, per individual, but those are a fraction of the week/year, and most people then prefer to do something other than babysit bots in EVE.
TLDR: My proposed bot-tests not only wreck bots, but even more importantly, they wreck the capacity of most players operating them, to keep them running.
This won’t completely prevent NEETs that stay at their computer for 16hrs per day, but it will wreck them while they sleep for 8hrs (-33% efficiency). More importantly, it will wreck any botter with a “normal” life of 8hrs work, 8hrs sleep and the remaining 8hrs divided among IRL concerns and EVE (>-66% efficiency).
2 Likes
This is great, if your only running one client, monitor, computer, etc. I do things on my other clients while in warp tunnels. Ive even done the race to the bathroom for emergency bio break in warp while on CTAs. This to you is me being afk. As is getting another beer from the fridge, tossing something into the microwave and then returning when its done.
I think you need to rethink how and why Eve can be multiboxed while other games cannot be to understand why so many people will false positive on “bot-like” behavior in Eve.
On a regular day I have about 8 or 9 clients going across 7 screens. Most are doing mundane stuff. I used to multibox ice miners years ago and have multiboxed a lot of things as well. Most are doing disparate things these days but again mundane without a lot of need for staring at a screen all day watching paint dry. In fact it was because of those long breaks of nothingness within Eve I started multiboxing just to have something to do in the meantime and to “fill” up the gaps with more Eve.
I bet you on any day you would flag most of my accounts as being bots simply because they would appear this way when I am still right here. I might be eating a meal and watching a movie on another screen infront of me at my computer but I am still actually right here.
So please rethink things from the ground up on this one.
What is the process of contacting someone when you are wrongfully accused of botting ,and what will be the result of ignoring requests like mine to be reviewed?
Would you be willing to endure the inconvenience of occasional bot-tests per account, so as to potentially wreck botters competing with you by up to +66% of their output?
If I may answer on his behalf:
Being accused of botting means nothing, unless they also file a report to CCP.
If you are reported for botting, and subsequently found guilty by CCP with resulting consequences, you can appeal your case by contacting CCP Customer Support.
which I immediately did, 4 DAYS LATER ,still no reply. I’ve paid too much money into this game for this kind of customer service.
Just to be clear, my statement was about the notion of ‘survival of the fittest’, not evolution itself.
Language evolves, but is “dog, perro, or inu” more fit to mean Canis (lupus) familiaris?
There might be a reason that docga won over hund in English, but it’s not because one word was more fit than the other. In fact, they changed places. Dog is now the general term, while hound usually refers to a working animal. It was reversed in old English.
–Gadget likes doggos
Edit - Sorry, didn’t mean to get all pedantic. Was hungry - no Snickers.
1 Like
Because CCP can’t tell when you are multi-boxing, and that can’t be factored into the algorithm? So, maybe I already considered it, but considered it too trivial to address as the system would work the same with 1 or 8 computers.
The example you gave is not a “false positive” for “bot-like behaviour” it is a true positive for bot-like behaviour.
False, I think you missed the point of… well most everything I said.
-
First “bot-like behaviour” does not mean you are a bot, I never said this, I never implied this, I never meant anything like this. It means you are behaving in a way that looks like a bot. No one is being flagged as a bot because of bot-like behaviour.
-
Bot like behaviour only would tell the server to challenge you if you are a human paying attention to the computer. So, even if you got the challenge, it wouldn’t flag you as a bot, you would just have one of Salvos’s captchas to solve. Which in your given case you’d have no problem passing.
-
If you then failed the captcha or failed to respond to it within a reasonable time, then there could be in-game consequences and you could then be flagged as a “potential bot.”
-
After being flagged as a potential bot, all of your game play activity would be reviewed by living breathing members of the security team to determine if your account activity matches that of a bot. Then, and only then, would you be “flagged as a bot.”
When I say bots are bad for EvE, I don’t mean that it’s bad for players to use bots, but good for CCP to hand over the policing of bots to other bots; I mean bots are bad.
“Bot-like behaviour” is a way to know who to test without testing everyone all the time. (This is for @Salvos_Rhoska too.) No one can test for a bot, that isn’t what the test is doing. The test is testing for an interactive human, and only an interactive human would be able to pass it. So, until bots become much more advanced, there would be no “false positives” (a bot that passed as an interactive human). And, if the tests are something like Salvos is suggesting there would be no “false negatives” either, because anyone who can figure out how to undock should be able to solve them.
No, if i understand you correctly you are looking at too small of a time frame. For example, AFK VNI ratting is something done by players and bots, which involves a few commands given and a lot of time with no commands being given. Intel bots, which can be ran from docked inside a station never give any commands, but they are still bots that are reading from the environment. The only thing that can rationally be assumed to not be a bot is an active player doing things that player do, but bots don’t do.
So you want a captcha for each mission? No, that’s crazy. I think, no matter what they are doing, if a normal player were to see three captchas in one day it would be too much. (Exception, failing bot tests should increase your chance of seeing them, and passing them should decrease your chance maybe to once a day - but still only when playing with bot-like behaviour.)
I think death in space is more of the “crawling” (not that it really fits the crawl/walk analogy). It more fits with the ideas of EvE and EvE’s “justice” system. And it fits better in a lore sense. “Do a puzzle and I’ll give you something to do” just doesn’t fit, but “while travelling through space some ■■■■ happened, fix it quick” makes a lot of sense. It increases the risk of those botting, and cedes to those who want to kill bots.
This was my first thought, but warp times vary a lot, so it could be bad, or predictable. This is why I suggested after the gate, it’s always 60 sec. That immunity time could even be increased for gate that you raise a captcha. So when going through a gate you could have one of two things a 60 sec timer like normal, or a captcha and a 90 or 120 sec (for example) timer. This way completing the captcha does not put anyone at any risk, only failing it does.
Yep, and that is why this would introduce a new in game thing they would need to protect against: random failure. If the AFK “player” is anywhere near their computer such that they could respond to a warning that they are being attacked or that a non-blue has entered their system, then they would be able to respond to the “random failure” captcha.
If someone is completely AFK and not docked or safe spot cloaked, they should expect to come back to their pod (or more likely clone bay). What is their argument, “well I had good tank and all, so I was safe”? No, this is EvE, you are never safe, unless you activity make yourself safe.
@Fluffy_Moe I was thinking about your overlay bot example and I actually think that would be a great kind of bot to have. (I’m serious)
The overlay bot can do a lot of things, it is very powerful, and this is what made it so successful in the other game you cited. What it can’t do, is do those thing like a human. So, even if such a bot could be used to solve the captcha, at the same time it would have to tip off CCP that a bot was used (this is in the nature of how overlays work), which would defeat Salvos’s prevention plan, but it would be even better for a punitive plan.
Too many of these post and replies are far too short. Step it up and double their length. 
The bots will at least remain busy reading all this splooge.
1 Like
This exact phrase is the issue with Eve online period regarding this matter. The fact that normal gameplay looks like botting behavior means that bots are harder to detect and the game play is something very repetitive and simply so that it becomes boring for a human to do at length. This isnt good game design.
As long as they are integrated within the gameplay the annoyance is kept as a minimum. Your ideas are great Salvos. Just make sure they are within the game as lore or gameplay rather than the hated captchas and other such junk.
I hate bots and botting because it ruins my characters, isk and other value in the game I enjoy playing. Though I could say I am one of those no lifers that can effectvely play 16+ hours/day. My max is 22 hours btw, and then I needed like 2 or 3 days of sleep and Eve avoidance.
So yes it is something I want very much for this game.
Yet the idea that most PvE or simple things within the game are coded and created to mimic bot like behavior is a huge issue. To minimize this the game itself needs to be changed, not hugely, but to integrate features in fun or unique ways. Aka things break going through old Empire jumpgates that the empires are sick of paying for repairing and need to be fixed before you warp off. Random capacitor overcharging needing Scotty the docking manager to take a look at in the nearest station to prevent a catastrophic meltdown or shutdown of all systems leaving you stranded in space. Realigning warp core coils. Fixing offlined and damaged weapons or modules due to flying through a magnetic storm while in warp.
Give it something lore based and integrate it and youll have your solutions imo. 
This is almost ALL PvE in Eve. In fact I can gank in such a way as to incur bot like behavior. I can gate camp in such a way as well.
You have previously stated that you dont care about those leaving that game in a bot like behavior yet you must realize that this is all PvE activities in Eve. And as your example stated that unless a player is utilizing 100% of the available activities that a single client/character can offer; such as PI, market, manu and R&D, then almost all activities other than pure PvP will get flagged under your system as bot like behavior. Then to say that its okay they just leave the game is absurd and ruinous to Eve as a whole.
I did point out that the bot-tests can be “Evefied”.
The EVE UI/client offers ample possibilities to make the bot-tests seem like part of the game, rather than something stuck to it with duct-tape and caulk. The EVE UI/client also has capabilities that can be used to stump bots.
People have a knee-jerk reaction to the term “captcha”, but what I mean is a test that requires human interaction to solve. Basically, just another series of mini-game (which at least most bots will be unable to solve).
My 3 graphic examples where very crude, and just to somewhat indicate what Im getting at in terms of bot-tests, and how the EVE UI/client is more dynamic than a web browser based captcha.
2 Likes
No. I’m sorry, but you are still looking at this backwards.
The default for a captcha system is “test everyone”. That doesn’t mean they are “flagged as a bot” it means you don’t know until you test them. My suggestion doesn’t pick people to test, it picks some people who don’t need to be tested. As I said before:
Everyone can play almost any aspect of the game in a “bot-like” manner, this means that an algorithm can’t say you look like an active human, and so you get a captcha. This is EXACTLY THE SAME as what would happen in any system without an idea of “bot-like behaviour.”
The idea of an algorithm that checks everything a player (not restricted to a single toon) is doing to see if it can say, “yep, that looks like an active human” is not flagging everyone else as a bot, it’s flagging the active human players as active human players, and thus omitting them from the captchas and making the captchas less annoying and intrusive in general.
Now I think you are intentionally trolling or making a straw man. First, I specifically said that it was not limited to a single client/character, yet you present that here. Second, I never said nor implied any requirement close to 100% utilization. I said these are things that humans use in a certain way, that bots can’t mimic and by looking at those things some players could be pre-screened and therefore wouldn’t need to solve a captcha.
I absolutely did not state this.
Agreed, things should be changed to make the game more interesting. But, that doesn’t mean all bots should get a pass until they do.