Eve Bots - are they a bad thing, is CCP complicit in their use?

Communications Center General Discussion
551

On a scale of 1-“I know how to program basic and advanced windows applications”, I know how windows programming works.

“Botting” is moot. What you do with the application is moot. The simple fact of the matter is that getting the window contents is literally one line of code. Sending mouse or keyboard input to the application is also literally one line of code.

Text recognition can be imported from libraries, so any query requiring user thought would be easily parsed. Image recognition has been around for decades.

At that point, it’s just basic logic, and the more thought-out the bot is, the more robust it will be.

3 Likes
552

Show me the code/process to break the captcha image I posted.

553

In psudocode, it is the following process:

  • During initialization of the botting application you get the handle for the Eve Client
  • During your processing loop, you get the windows content using GetClientRect
    If there’s a captcha, you parse it using whatever logic you decide to use. Basic image processing stuff at this point. If there’s a phrase, throw it through an OCR library, translate it into logic, and then handle that logic appropriately. This gets more complex but it is still only about a day’s worth of programming if you’re good.
  • Otherwise, you handle the game (overview, active modules, etc).
  • When you need to make some sort of input to the game, you send a message to the application via it’s window handle, with SendMessage

This is all very rudimentary, but it’s also very basic WINAPI stuff.

4 Likes
554

How many botters in EVE do you think are currently capable of that?
Your response involves human interaction to access various databases.
This defaults the issue, as a player present is not technically botting.

555

Anyone who can read this:
http://www.learncpp.com/

Plus, Eve has a significant number of people who do IT-related professions working in it (about the only higher profession I see in this game is military personnel). We learned how to program a LONG time ago.

No. A library is a collection of code someone else already wrote. To use it, you import the library (commonly a DLL) and then you can use their code in neatly wrapped function calls. No databases required, no human interaction required (beyond that needed to program it in the first place).

You are still botting if at any point your ship is doing things that you didn’t expressly do with your own input. Even if your supposed “default” was the correct conclusion.

3 Likes
556

How would you deal with the colors of the two dots to be connected being different by 1 in hex, which whilst imperceptible to the human eye, would be considered concrete, and different, each by reference?

557

CCP actually bans all automation. If it is just warping to zero by a script of autopilot kind or full fledged bot farm.

558

Computers are remarkably good at checking to see if one number exists between a range of numbers. Another single line of code to hit literally every shade of red that can ever exist on a computer.

Edit:

3 Likes
559

Then the botting problem is worse than I thought.

What do you recommend as a solution?

560

Computers can even describe emotions on images with a script and using this:


CCP should just focus on automation and characteristics of a bot account.

2 Likes
561

Not Empty Quoting.

1 Like
562

Captcha was then, still, at least one of the solutions.

563

Absolutely. I agree that a captcha on pilot logon will help, always did. A captcha that requires “common knowledge” that a bot programmer simply cannot impart (such as what a street sign is, in my previous post) WILL beat a bot.

The net result is that a botter would have to individually answer each captcha. For a small number of bots, that’s not a problem. For a large number of bots, that would be cancerous to say the least.

It has the downside of being onerous on legitimate multiboxers as well, however. Logging in 10 accounts is not against the TOS, assuming you aren’t botting them.

Where I disagreed was in saying that a captcha during gameplay would be unnecessarily disruptive to the player, and thus be to severe detriment of game enjoyment. A sufficiently complex captcha will absolutely stop botters. But it would destroy the game at the same time if it’s done during gameplay.

2 Likes
564

All in all bots are relatively easy to catch. All that is needed is better bot reporting system for players to report, for example market bot, combat (ratting) bot, FW bot, courier bot, cloaked AFK bot, etc. etc.

On top of that algorithms can also be used to flag characters / accounts. For example, characters remaining online 23+ hours out of every day, characters never leaving a station or system, many other types of flags can be done.

On the technical end, flags can be set to check for input type, for example is the input joystick, controller, mouse, keayboard, or is it a macro or script ? Is a character often or continuously bradcasting from the client to an external app or ingame channel ? etc. These again, are just flags, not actual bans.

Now, these are just flags, not bans. To actually ban you need to verify bot status vs actual player status. This is where the majority of difficulty comes in. For example is an AFK cloaker legit and at keyboard and typing info into an intel channel ? or is the broadcast automated ? is the combat automated or is it a legitimate player ? Are the market transactions actually done by someone that dedicated themselves to that style of play, or just blindly .01 isking until a threashold is met and done by same character 24/7 except reset times ?

I think if they just put a few people on it, that can take 1 day a week or 2, to verify such flags and rotate out to their primary duties, then ban maybe once a month, instead of 6 or a year, then things would improve a lot over time.

Automating such detection, no matter how good the system seems to be is a mistake, for example : Archage and what happened there.

It will ban a lot of legit players along with bots, then the bots will just find a way to circumvent and come back and continue anyways. hence I think automated systems should be used only for flagging, not verification. Once verified, the actual bans can again be automated, simply due to the volume if nothing else.

It all boils down to manpower, labor hours. Its unfortunate it has come to this as there is of course a lot that should be done with this manpower elsewhere, but I don’t think there is any other choice then that. Otherwise things will just keep progressively getting worse.

The second approach would require a huge amount of resources, and that is to redesign the game as a whole from the ground up to make botting much more difficult. Limit acceptable input types, limit hours someone can be logged in after a couple days of gameplay within a week or month, impliment AFK timers, make combat itself more visual based instead of spreadsheet (overview) based. Limit multi boxing, how many accounts any player can have logged in at any time per machine, how many clients can be actually running. Etc. Each one of these alone is just a small dent, but together, with other additional stuff it can have a large impact.

565

For me in relation to bots its to make them work for nothing while being fixtures in space for players to find and kill. Reducing their means to make isk while not banning them is kind of a win/win imo. CCP collects income from them as they, as long as they remain active, must use the PLEX market and buy, legitimately, from players account time and we as players get to kill them when we find them.

The issue with botting to me isnt the fact they exist but that they make isk illegally. Taking away that ability from them but letting them “run” for nothing is a more viable way of destroying the damage. Just make them targets and useless and then force the botter to biomass or cancel or change them themselves. To me its much more ruthless lol.

Love the “discussion” on what can beat bots between Pervert and Salvos.:heart_eyes:

Having players crowdsource issues imo wouldnt be the way through. It gives way to much ability for abuse, period.

One way of securing client side manipulations to run many clients on a single computer would be to secure the client through blockchain tech. Making each client effectively only be able to be “hacked” or changed through breaking a blockchain that is verified through every other client and the CCP server would be impossible. This would reduce the raw number of botable clients per computer and really raise the bar for “cheap” bot farms.
A side note would be once clients are secure then server load for things like large fleet fights might be able to be passed to secure clients computers to share server load for some calculations reducing lag possibly. As the only reason to not use a clients computer for such calculations would be the insecure nature of clients presently.

Just a thought, but Ill keep checking back on the thread.:smiling_imp:

567

Coevolution suggests that you could end up with smart bots and no “dumb” bots and given the ease with which things can like programs can propagate…

1 Like
568

Make it mandatory for GM’s to put a block time (5hrs, 10hrs, 20hrs) per week or month per GM dedicatd to perusing through BOT reports, AND getting in game and flying their invisible from the playerbase ass to said locations and literally putting eyeballs on the targets…

If you are in the game outside of trading bots, it is actually fairly easy to identify a bot by it behaviors whilst actually watching it.

1 Like
569

Eve Bots - are they a bad thing, is CCP complicit in their use?

In answer to the thread title I have to say yes. Why? Because CCP has and continues to introduce bots of their own into the game.

While the playerbase continues to dwindle down, more and more bots will be added to the game. Eventually a major percentage of active ships ingame will be bots. That’s so the few real live players left flying around the game will have the illusion that the game is active with other real life players.

I base this hypothesis on events that happened in another MMORPG I played over 10 yrs ago called DarkSpace.

It was strictly PvP set in a SyFy space environment which, every once in a while, would spawn NPC’s holding special loot that could be fitted to your ship giving a slight advantage over the other player ships. Over time more NPC’s were spawned with the intention of giving other players a chance to upgrade their ships but unfortunately the players who already had some of those special items were able to quickly kill and collect more of the special loot, thus causing other players to lose interest in playing the game. After a while the game soon started looking like a ghost town.

Since the playerbase had dwindled down to almost nothing, more NPC Bot’s were added to the game so that the few real life players left in the game could have the illusion that there were other players in the game with them.

Anyway, just something to think about.

DMC

1 Like
570

Bot can’t evolve, they don’t have random mutations that then get naturally selected. They are in-fact designed by an intelligence, for a specific purpose. Then they are redesigned to fit their changing environment.

Or if I may quote the mighty Gadget:

571

The most common “bot-detector” idea givers seem to like to say “its easy to spot a bot who is on doing the same thing 20 hrs a day.”. But then wouldnt they just put random timers on their log ins?