I think this is a valid concern regarding Challenge 3 of my roadmap.
It is absolutely important that these bot-tests occur for non-botters as infrequently as possible, and possible botters as frequently as possible.
Im however still having difficulty how you propose to distinguish between the two.
Im my proposal, Ive only got so far as to be pretty certain bot-tests could potentially shave 33-66% up-time off all bots that otherwise would be run 23/7 and/or when the player is afk (such as asleep, at work etc)
Thats the best I have managed to offer so far.
(Imo, a very significant drop in bot efficiency)
Your proposal is an interesting addition to that, so as to reduce the annoyance of these tests to non-botters, but Im going to need more explanation from you on that.
Like I’ve said a few times now, It’s first important to remember what we are doing, we are not identifying bots, we are identifying active humans. This isn’t just some game with words, it is a fundamental difference in thought and approach.
Like you have said, the EvE client has a lot more functionality than a website does, and this extra power can be used. If I understood you correctly your main thought here was that this could be leveraged mostly for making cooler captchas, but it’s a lot more powerful than that.
Obviously, and as I already mentioned, there are many aspects of the game other than the task you are doing that can be looked at, but I will example just with chat:
Can a bot “fake” chat? Sure, but mostly in some from of SPAM, unless it’s Turing complete, it can’t fake conversation.
So “typing in the chat box” would not pass you off as human, but reasonable replies could. When I say “reasonable” I don’t mean the content of the messages, as that would take a Turing complete algorithm to “understand,” but I mean reply patterns that follow a normal human. The best part here is that you aren’t just looking at the potential bot’s side of the conversation, CCP can see both sides, and the bot can’t control the other side. Although the content of the conversion can’t be read, it can be scanned for keywords.
Go to Jita and watch chat for 10 min, and see how the scamming players and the scamming bots interact differently, notice how other players respond to bots vs scammers, look at the back and forth “you’re a scammer”, “I’m not a scammer” conversations that go on between the real players.
When I say to watch chat, I’m also not restricting it to local, which isn’t used much outside of trade hubs, but do you chat in corp? Do you chat in help? These are things humans do that bots can’t really do.
And once again, not doing these things, does not mean you are a bot. But doing these things does mean you are an active human.
Chat is just one example of one thing that can be watched. And the more things that CCP watches, and the more data collected, the easier it is to use the data in the future.
Your captchas themselves can be used to decrease the tests for non-botters an increase the tests for bots.
if a bot can fool CCP into thinking it’s an active player regarding the things I’m talking about, then it can beat your captchas. So, active players never get captchas while being active.
This means no worries about a captcha in the middle of fleet ops or other things that would make everyone cry. (The crying would be justified here.)
Everyone else can be subject to captcha tests.
I’m not saying all the time, but they should expect them.
NO-body expects the Spanish Inquisition!
But as you pass the captchas, proving to CCP that you are an active human, the frequency of tests should decrease.
This should never drop to zero as to prevent “training” an account to be free for botting.
For this scale of how often you see a captcha, a failure should be much stronger than a pass.
A failure doesn’t make you a bot, sometimes ■■■■ happens, but it does give CCP the right to ask you more often if you are a bot.
The frequency of being tested should also have some basis on ratios of time in game verses time doing things people use bots for and time as an active player.
At some point, based on failed captchas all of the data is turned over to the humans on the security team, and only they can rule that anyone is a bot and take action against them.
This can be gamed by botters, such that they pass bot-tests manually while botting whilst at their PC, thus reducing their ratio of future bot-tests whilst they bot whilst sleep/work afk.
Creates complications between statistical analysis of account vs character, as to how many bot-tests they have passed/failed each, and in what period,
This I can agree with. More data is good.
Albeit, I hope my bot-test will reduce bot effective activity time so significantly, that CCP can instead focus on other metrics such as inhuman online activity times and repeated predictable/identicle behavior by characters/accounts (including multi-box botters), to identify and deal with the remainder.
Furthermore, the loss of 33-66% of effective bot time (out of afk time), may have sufficient negative impact on many botters, that it will no longer be fiscally feasible for them to continue botting, thus even further reducing the effort/drain on CCP resources to identify and deal with the remainder.
Imo a better implementation of these bot-tests, is so they “fit” into EVE as seamlessly and unobtrusively as possible.
I propose the following as preliminary/draft guidelines:
A) Bot-test should never occur when the character has been targeted by either a player or NPC in the last 1 minute.
B) Bot-tests should never occur when <5km from a gate, Citadel or NPC station.
C) Bot-tests can, however, occur whilst in-station.
D) Bot-tests should have a random chance to occur, roughly equivalent to 1 per hour.
(Yes, roughly 1/hr. Deal with and HTFU.)
E) A bot-test popup disallows ALL ingame action (except for the “Esc” options menu), until completed.
F) The account cannot log-out safely whilst the bot-test is incomplete.
G) The account has 3 attempts to solve the bot-test, before being immediately ejected from the game as unsafely logged-out.
I am beginning to understand your proposal, but I see it as an “advanced” capacity that can be applied later so as to streamline the bot-test mechanic. CCP can fine tune the algorithms of when/why/how the bot-tests occur later, as well as if/how those would apply to specific accounts/characters, their behavior, and what kind of data CCP collates off that to inform/flag them to investigate bot suspects.
Actually a gate, post-jump is the safest and easiest place to put a bot test (as I already detailed above). Or it could even come mid-jump, in the session change.
I think that’s too high. It should depend on what you are doing, but I think 3/day (8hours) would be enough, and the more you pass it drops, but never to zero.
I just don’t think this is the EvE way, making you flashy at a gate in a disabled ships, or siting at a belt in a disabled ship is much more fun, makes a greater loss to the bot, and gives more data to CCP regarding the botting. Why would you want to logoff the botting carrier, when it’s dead in the water waiting for the next wave of rats (also increased rats could be a response) to come pick it clean?
If a player fills bot-tests for 8hrs per day, for 5 days, whilst botting but present at PC, and then bots 23/7 over the weekend, they will have skewed their ratio far enough to enable/extend the weekend constant botting,
See what I mean?
No offense, but this is not a cogent argument.
Noted. I included this guideline to prevent undocking ships and fleets at gate from incurring bot-tests, which could be disastrous.
Note: In my guidelines, there is nothing preventing a bot-test occurring mid-warp.
Caveat added for bot-test being possible post-jump, before gate cloak drops,
Fine. I dont.
I have no problem with solving a simple bot-test 1/hr.
Thats 5-20seconds per hour. I can handle that.
I spend more time picking my nose… (I’m not ashamed!)
If 5-20secs per hour means we can cut 33-66% off bot up-time, I will do it gladly.
As I said in my guidelines, as long as the bot-test is up, you/bots are incapable of other ingame action, ergo sitting ducks in space.
Depending on the exact scheme of the bot-tests, they may require an additional click to “send” the result, or not. Ideally, Id prefer they dont involve an additional click to send, for a number of reasons.
I’m open to changing the failure to complete the bot-test via three attempts, or within a reasonable time period, to lead to flashy etc as you propose. If thats what people will be more amicable towards and what it takes to sweeten this idea, I have no problem with that.
CCP would have to change the log-out system though, for that, so as to keep the ship present (and inactive) in space despite the bot/player disconnecting their client, but hey, why not.
Edit: The more I think of that, the more I like it.
Random flashy bots stranded in space with no way to act even if the player returns, for the rest of us to blow up and laugh.
There is one simple fix to the bot epidemic, but one that will raise howls if rage from the most organized, and loudest segment of the game: the RMT cartels.
Why do people bot?
Profit. Whether the cash is strictly for ingame use, or “other use”, profit is defined net income after whatever out of game resources needed to bot PLUS the PLEX needed to maintain the account PLUS the ingame costs of ship losses, drones, fuel, etc.
Where do the vast vast majority of bots operate?
Null sec, possibly wh’s (though I have not lived in a wh in years, so not sure about that one). The ISK/ hour is at least an order of magnitude better in null sec than low and high sec, and far safer.
So if CCP radically reduced the ISK / hour potential in null sec, right across the board, a lot of bot’s would no longer be profitable, or at least profitable compared to the risk. I am talking about a 75% reduction in all bounties in null, and a 75% reduction in ore yield.
“But , but, but that is not fair to the honest RMT cartel members!!! You can’t penalize them for the tiny fraction of bad apples inside the various null sec alliances. Think of the damage to the economy of the game, and the amount of players driven from the game.”, would be the immediate reply from the cartel propaganda machine.
And CCP’s reply should then be. “Tell you what RMT cartels. You demonstrate that you are self-policing bots in your alliances. And by self-policing, we want demonstrable results, one that we can measure with our super-secret tools. If we see that bot usage has dropped by X%, we will reinstate your ridiculous income potential. And, if we see that bot usage is creeping up again, we bring the nerf hammer out again until you guys fix the problem”
It is clear that CCP does not have the resources and/or competence to deal with this crisis, and more than one person has suggested that self-policing is the only thing that will work. CCP CAN control the ISK faucet. That carrot/ stick factor should be more than enough incentive for the cartels to clean this up.
Keep in mind though, the cartels collect the same tax rate off of bots as they do non-bots. It would be enlightening to know how much income flows to the cartels, as a percentage of their total income, from bots.
I wonder if the push for alts instead of more interesting game play led us to this point. So many things in eve require nothing more than watching modules cycle. Mining… If the game required more interaction could it be harder to bot and less profitable? I only mine to remember the pain… and it is pain. There is nothing for my brain to do when mining… As nothing effects the cycles of my lasers. I cant aim the laser better to get more ore… Align and mine.
Ive been a strong proponent of a mining minigame for years.
This would constitute an EVE intrinsic bot-test, as is my impetus with all of my above posts.
The minigame can also be designed so that it rewards player presence with greater yield, whereas it blocks bots incapable of solving the minigame out of mining.
Ill try to create and put up one or more examples of what that could be like, graphically, tomorrow when I have more time.
You started on the right path but somehow abdicated to some underlying jurisprudence that can in no way be a solution.
Professional botting practices and RMT go hand in hand for most scenarios.
As you described,leaves CCP open to a negotiation that can’t lead to a means to an end.
Botting and RMT are of nefarious intent. Are you suggesting and possibly conceding that the RMT structure can’t be removed?
The last article I read about CCP finances in relation to amounts given. RMT approximately 2% to 4% compared to the total for CCP annual revenue. Example without using actual figures. CCP $100 mil,RMT $2 to $4 mil
The largest RMT site likely makes $150,000,000 annually. Much of which has no tax paid upon it.
Perhaps tomorrow if time allows me. I will outline the constituents involved,their cause and effect,and a painful truth leading to a solution. Honestly,I ask myself are players and gaming companies ready or indeed interested. I do expect a backlash to the highest degree and question to myself if I am willing to follow through and defend the outline.
Let rip. Fk the haters. You may not be alone in defending it, it there is truth to it.
I certainly will support whatever is valid in it.
But I would ask you do so in a new dedicated thread, so as to spare this one the subsequent apocalypse, and so your thread gets the specific attention it deserves.
Conceptually it would be far from a bombshell more like common sense. I would never hype it as so and prefer others didn’t as well. I am pondering thus value and haven’t reached a conclusion I can accept for said value.
Many components need to be carefully worded and that in itself is daunting,while trying to be concise.
If I get passed that,I will indeed create a new thread.
There is also this. CCP Falcon said he was going to provide some resolve in some fashion. I would know what that is before commenting further. At this point I will refrain until that is public.
Let’s hear what the man has to say. After all he is gatekeeper and key master.
Nope, never to zero their weekend botting will still raise a bot test, and they will fail it, which will end their weekend botting (maybe an hour or two later than without the “training”), failing the bot-test will also undo all the training (remember failing one counts more against you than passing one). And if they try this for a few weeks straight then the pattern makes them more guilty.
It wasn’t an argument, it was just saying “you’re wrong.”
No offence (I love how that’s always used before saying something offensive), but it meant you don’t know what you’re talking about and I didn’t feel like explaining it. I will give a short explanation now.
Legit multi-boxing vs. multi-boxing bots is actually about the easiest form of botting to detect, this is because a human is very limited in how many different inputs they can give at one time. A single bot multi-boxing could try to duplicate that, but a team of bots could not even come close. Registered account owners and IP addresses are easy to combine in a data search (I know 1 IP doesn’t mean 1 person, but there are other things to look at to get a real good idea), and it’s when looking at the big picture that you can clearly see if the guy running 8 ships grabbing the ice fields when they form is a really good multi-boxer or using bots.
I don’t like mid-warp, warp has too many variables to make it reliable timing-wise. The more I think about my post-jump, before the gate cloak drops, I still like it, but I think we can do better: in the jump tunnel during the session change. This can allow whatever time is needed for the test without effecting anything in-game. You sometimes get stuck there anyway for a few min when jumping into a busy system. This could also allow these bot tests to not need to worry about current node load, as they could be handled by a bot-testing node while your ships is being transferred between nodes (if needed).
Of course this only effects travel/mission bots, not ratting bots; they will have their own set of testing/failure result rules.
It’s not fundamental to the idea, and CCP would pick their own frequency anyway regardless of what you and I think.
You could make a poll, and ask the players what would be the highest frequency they would tolerate, then if CCP sees the poll, we can trust that they will ignore it and do whatever the hell they want anyway.
You said in another thread that you wanted to get CODE to support you in your anti-botting campaign, this is how you get CODE to support you. Making things that look act like bots blow up is kind of their MO.
I might concede to a way to get out of it if you come back to your computer, mostly for the pro-AFK group, but it wouldn’t be trivial.
I see you missed the “and” “P and Q” =/= “P”. Especially when “Q” is the more important part of the statement.
I wasn’t saying that there was something fundamentally wrong with a guy walking into a bank wearing a mask, but if he then gets in a huff because people look at him like he might be a bank robber, then there is something wrong with him. And I don’t mind if he wants to leave.
Most drone centric ratting models can be done by whenever you scoop drones one or a few of them break, have bad code from taking too many headshots from the NPCs, etc and need to get “fixed” and are simply put from the drone bay to the cargo hold until they are repaired in a station or through usage of a mobile depot. If no cargo space they are simply auto abandoned in space.
Other ratting models could have Thermodynamic damage build up through over capacitor charge regarding usage. Simply stopping the module, or docking up/mobile depoting to offline/online, or even using nanite paste to repair the damage would all work lore wise. If not then modules would burn out and need a full service repair over time.
has to say. After all he is gatekeeper and key master.