Posted on 25 July 2016 - 06:38 PM
@ArchybotWasn't sure who to tag, and I apologize if this is unimportant, and I understand that you are working on a lot of things.
Previously, I successfully made a bot that played on the Japanese Riichi Mahjong server called Tenhou. It was definitely an above average player, but I never contacted the website owner because the bot wasn't THAT good and nobody really cheats or gets banned honestly, and uhh, well I don't speak Japanese.
I would like, if allowed, to use an alternate account I have (currently called "Daisangen") exclusively from here on out to try and design a bot that can play Iron Soup.
Now lemme be clear, I don't intend for the bot to have an unfair advantage at all. I am thinking it will have something like this at first:
CPS: maybe around 8?
Aim: Struggles a little with good jukes by players, struggles with keeping up with circle strafing.
Blockhitting: Solid, but has some small % chance of messing it up.
Souping / inventory manipulation: Above average, but at a reasonable speed.
Ghostkilling: Ideally I'd like it to be respectful and not do this, but idk.
My idea isn't to edit the code of Minecraft in any way, but rather to make a Python script that sits /outside/ the game and interprets the image data on the screen, moves the mouse around, hits keys, etc, again not to get an unfair advantage, but simply for the purpose of doing it. Making a bot that can play Iron Soup at all, even decently. The arenas are small, flat, pretty much ideal rooms to make some sort of bot like this.
Monitoring heart count, souping, knowing when to refill soups, doing that, etc, this is all basically easy and just a series of tedious programming. The difficulty will be in detecting the other player, like, differentiating them from the walls even lol, and then how close are they? what direction are they moving? This is gonna be really confusing x_x.
I may not even do this if the difficulty becomes too great, but I wanted to know if this sort of thing would be acceptable, and if so, I obviously wanted to be very upfront about it. If it is okay, and I did start experimenting with this, I'd get some basic stuff sorted out in a single player Minecraft world against mobs and maybe against myself on another computer, and then after that, I would pretty much isolate Daisangen's play to Unranked Iron Soup on this server. Honestly with the difficulty with detecting the player I'd be surprised if it could even hold its own in there at first.
I am asking before I even get started because if it isn't allowed, there isn't much point in putting in the time.
If I do get going with this and it kinda works???… maybe there can be different difficulty levels… who knows… all reasonable of course.
But yeah it'd be really cool for me. I could post videos about the progress of it on Youtube, I'd learn more about Python and some programs that help with Python, etc.
Posted on 25 July 2016 - 07:15 PM
thats swesome, how long did it take you to make it??Posted on 25 July 2016 - 07:25 PM
This would be great, but I doubt it's allowed since it's basically a shared account. You and your computer share the account. Your computer does your soup ladder, and you do the rest. Badlion is all about effort and practice, and you want to get better. But as for the actual coding experience part, that's great!Posted on 25 July 2016 - 07:28 PM
I think it would be pretty cheap to have a bot that can actually kill players and take their rating. Doubt this would be allowed.Last edited on 25 July 2016 - 07:46 PM by greywalk
Reddy23 wrote
The Riichi Mahjong bot was like 1700-rated on a server where the starting player is like 1500 and the Danteh-equivalent on there is like 2360 http://tenhou.net/sc/?rate4. My bot wasn't crazy or anything lol, but pretty good. It took like 2 months x_x…
Idk how long the Iron Soup bot would take, because I don't yet know a great way to detect players, how far away they are, etc o_o.
edit:
the above posts didn't really read the op. I said it would be restricted to Unranked Iron Soup X)
Posted on 25 July 2016 - 07:48 PM
greywalk wrote
Reddy23 wrote...
The Riichi Mahjong bot was like 1700-rated on a server where the starting player is like 1500 and the Danteh-equivalent on there is like 2360 http://tenhou.net/sc/?rate4. My bot wasn't crazy or anything lol, but pretty good. It took like 2 months x_x…
Idk how long the Iron Soup bot would take, because I don't yet know a great way to detect players, how far away they are, etc o_o.
edit:
the above posts didn't really read the op. I said it would be restricted to Unranked Iron Soup X)
If it's restricted to Unranked, which is pretty much where you practice, then there's no point. Unranked is there for practice.
Posted on 25 July 2016 - 08:29 PM
BlueJewell wrote
Tbh, this whole idea is super weird, so like when you try to apply normal ideas to it, it doesn't really make any sense. What is the point of making a bot that plays a game? What is the point of playing Minecraft? Are spoons just for eating food or can you use them to play the drums on pots and pans?
Let's just wait for an answer or for the thread to fade away.
Posted on 25 July 2016 - 08:30 PM
greywalk wrote
BlueJewell wrote...
Tbh, this whole idea is super weird, so like when you try to apply normal ideas to it, it doesn't really make any sense. What is the point of making a bot that plays a game? What is the point of playing Minecraft? Are spoons just for eating food or can you use them to play the drums on pots and pans?
Let's just wait for an answer or for the thread to fade away.
Alright, sounds about right.
Posted on 25 July 2016 - 10:47 PM
You should make a bot that plays a simpler ladder like Diamond, then move on to more complex stuff like thisLast edited on 25 July 2016 - 11:07 PM by Archybot
I would have no problem with this but @MasterGberry would also have to approve.My stipulations would be -
1. Players cannot lose points from the bot.
2. Players will not lose ranked game count from the bot the bot wins.
3. Players could only be matched up with it a few times a day
4. I would have to make it immune to GCheat
Ofc 1-3 wouldn't apply if it was only ran in unranked.
This is always something I've wanted to do but never had the time to do it.
Posted on 25 July 2016 - 11:52 PM
Archybot wrote
My stipulations would be -
1. Players cannot lose points from the bot.
2. Players will not lose ranked game count from the bot the bot wins.
3. Players could only be matched up with it a few times a day
4. I would have to make it immune to GCheat
Ofc 1-3 wouldn't apply if it was only ran in unranked.
This is always something I've wanted to do but never had the time to do it.
Yay / ty ^_^
Alright well I'll let ya know (supposing he's also cool with it) when it is maybe ready for Unranked. Might be a little while. I'm trying to get it to kill a zombie rn lol.
Last edited on 26 July 2016 - 12:09 AM by RyanOyster
greywalk wrote
But yeah it'd be really cool for me. I could post videos about the progress of it on Youtube, I'd learn more about Python and some programs that help with Python, etc.
Do you think once all is finished it would be cool to have a live feed of stats? like a website that states wins, accuracy and other things?
Also are you going to create a GUI or just have the script? it would be pretty cool to be able to adjust difficulty mid game through a GUI
BTW honestly this is an awesome idea and an awesome project gz
Posted on 26 July 2016 - 01:32 AM
:O sounds AWESOME!Wanna include me please? I know python really noice and i can do stuffs :D
plox i always wanted to such a thing
especially since Archy says he likes the idea
Posted on 26 July 2016 - 11:25 AM
MadeInSlovenia wrote
Posted on 26 July 2016 - 11:32 AM
Sounds great, best of luck if you're allowed to do this :DD