Barotix wrote:more bots != more money.
Ban the bots. All of them. Condense the server.
Imagine if you dropped a piece of chocolate behind your desk. It sat for a few days before ants found a way in. One day you look down and see some ants so you kill them. The next day you see some more ants, so you kill them. Now, you can keep killing the ants every day, but they will always come back. But, if you were to find how the ants were getting in and why, then you could stop them for good.Banning bots is analogous to killing the ants you see. It doesn't solve or address the real problem on hand. Sure it gives you the satisfaction of a short term resolve, but it is one of those things where you have to be in it for the long term. Here are some problems and solutions that Joymax could do to take care of this game better. I'm a programmer myself, so I understand the feasibility of everything I am presenting.
On client based bots:There is not much Joymax can do in this area to prevent people from using client based bots. Any client side protections that they use, gameguard, hackshield, etc.. are just deterrents for the average person trying to mess with the client. The bot companies have skilled labor that specialize in reversing such technologies so their products work with them. If JM put a protection back on the client, it would disrupt client bots for a few weeks, and then they would come back in full swing. However, while the small % of client bots that are not being used are gone, clientless gold bots will fill those slots.
The best defense JM has against client based bots are server sided database heuristics. The most obvious one is the "both path". If JM were to add a few key trigger locations that if walked in a specific order, the account would be flagged to be banned, then JM could quickly wipe out a lot of player bots in one swipe. The obvious benefit of this is that once player bots start getting banned, people are less willing to bot. Plain and simple. It doesn't matter that the client bots would still be around, once faith in the bots is shattered like that, people will stop botting (and most of them will also leave the game).
The other subtle things JM could do is add some coordinate checking on movement packets. If the movement coordinates sent to the server don't match the map files, increment some number in the account. At the end of the week, accounts with a high ratio of invalid movement flags gets banned. The reason you have to set a cut off is because you can send invalid coordinates to the server. If you happen to click on a dock for example, the height will be of the docks and not the terrain. But, for most other areas that are simple flat lands, this technique would work,
The next thing ISRO could do is track "grinding time". vSRO limits players to 5 hours. I don't think something like that would fly in ISRO, but just by tracking how much time a day is spent actually attacking monsters, you can start to see accounts that are either being played 24/7 by people or bots. Of course that doesn't have a rule against it, so they would need to add some "acceptable play time" clauses. The client already has logging in it for play time, it's just disabled.
Joymax could add any number of client security features, such as process/module list handling, CRC checks, etc.., but it'd only work once. When the client remains unpacked, as it is now, those things are easily found and defeated; given you are looking for them. That pretty much sums up the client specific stuff. You can slow down botting, catch a few unlucky people, but overall, server sided analysis is the best technique.
On clientless based bots:There is a lot for Joymax to do in this regards in banning clientless gold bots and other clientless based programs. Most of the traffic Silkroad has is from clientless gold bots. They far overpower the number of users that use client bots.
The easiest way to ban clientless bots and keep them from becoming so predominant would be weekly client updates. If one patch Joymax changed their security algorithms and accepted any packets, they could silently flag and ban accounts coming from gold farmers. Further more, they can do this over a few weeks and then perform IP bans to prevent those same networks from coming back. This would work once very well, after that the bot makers would know what to look for each patch, but be delayed more so than they are already on updates. Since weekly updates is never desirable, JM would have to invest in a more realistic update system. One that would patch the existing client rather than send a new client on trivial changes.
Packet opcode changing is another way to easily ban clientless bots. If a packet format is silently changed, they can identify, flag, and remove those programs. Couple that with the server sided heuristics of checking paths, they could force bots to update to make them less detectable. There are quite a few clientless bots that support image code detection, so Joymax could address that by changing their captcha to the captcha that is used on their site, if not a different style. If they change captcha styles, then there is no way clientless bots could auto login anymore.
The final grand scheme to ban clientless bots would be hidden packets in the client. Let's say one update JM adds a packet handler to the client that will send a response to the server when a packet is received. One day, they broadcast the secret packet to all connections. All accounts that are connected and not using a client would not send the opcode. This would be a dead giveaway in identifying a clientless bot.
One misc issue that I'll throw in here is how they can stop gold spammers in town. Each account could be given a "mute flag", in which their chat packets are only sent to themselves so they see their own chat, but no one else does. I'm pretty sure SRO has this already due to some error messages in the PK2 files involving GM chat blocking. JM could hop on every week, mute spammers and never have to worry about those accounts again. Of course, remember that gold companies pay game companies to advertise in their games and spare their characters, so I doubt this would ever happen. (See latest news on WarHammer's response to gold farmers)
ConclusionSo, Joymax could make some game architectural changes to save the game if they felt it was critical to take care of the current bot problem. That would clean up the game for a while. However, bots would eventually return. In that case, it is just a matter of consistent game maintenance to ban bots when found. No, they won't be able to keep them all out, but they sure could keep more out than they are doing now,
if they truly wanted it.
Remember that Silkroad is a "free to play" game, and having studied their game architecture, it's not a cheap game to run. They have 32 servers now, so from a perspective of "success", they are doing quite well. All of the money they are making from people investing in Silk and buying gold for the game to attract players is getting the best of them. However, it is always important to understand that Joymax is a business, and they will run their business how they feel is in the best interest for themselves.
What I have written about is not the only way to make the game "better" though. I've only addressed making Silkroad more bot free, more or less the "how" in my ant analogy above. The "why" is something that would change the direction this game is going in. Party mobs, fortress war, and CTF are all things that help, but JM needs to do more in making the game "playable" and not just "grinding" (referring back to vSRO which only allows 5 hours of grinding, forcing people to job, cape, or guild war, or log off) We could speculate and offer ideas all day though, but there is little chance they would consider it.
Of course, we can all sit around and throw out ideas and fixing what we perceive are their problems, but until we are actually the ones in their position, we won't ever understand what they are going through or the reasoning for their actions. But hey, at least it makes for an interesting discussion.
