What behavior? Winning? I have been in games where the team that was full of clanners lost. Clans dont necessarily mean win. It does help them skill up for clan matches or tournaments and stuff. Clans dont stack? suuure they dont....
No. Stacking all clanners vs all non-clanners. Clans don't get their own servers and convince players to play against their stack. They win like 95% of the games if they stack. Playing against stack can be fun once or twice, but if it happens every game, it's annoying (for both sides). Skilling up for clan matches and tournaments (and stuff) can be disallowed by the server owner. Clans can rent their own servers for that purpose if they want to. Note that stacking is a violation of KM server rules: You haven't addressed the rest of my post. Why?
Nice to know theres a server where I cannot work as a group with others....which takes away from a team based game.. also the rest of your post was either additional data to prove a point and not needed to be quoted, or it was talking about something else.
Reef seems to be having a hard time saying what he wants to say. So, lets make a better example which actually happened. In 95% of games where I see 5 or more BSID members playing, they are on the same team. They treat the match like a clan match, only vs pubbers, and it makes it NO FUN to be on either team. If you are with them, it becomes a boring rapefest. So what, you won. Big deal. Did you have FUN getting there? Was there any sort of challenge? No. Against them, your team gets raped, because 5 skilled people vs 1 skilled person just doesn't work. Constant clan stax just ruins the game for everyone else. The idea behind this plugin is to prevent a small group of people from taking the fun out of the game.
I'd rather be on the clanstack team than on the other one, because working together with cohesion is nice in itself, but better than both would be two evenly matched teams.
Everyone should be on the clanstacked team! As for the other team, well... We can spawn random turrets so the clanstacked team will have something to shoot at.
There have been many games I've played where Simon and Opie float between the losing teams. Good on them for that. The absolute best matches are when you switch to the losing team and help to turn them around and pull off a win at the very end.
The problem with the Anti-stack plugin is that it forces you to play,with people that you dont like or worse cant play and dont want to learn to be a good teamplayer. The sort of people you want to hit with a hard and heavy item. The endless no armor apc buyers, the only scout playing base snipers...
If those players exist, SOMEONE has to play with them. Oh, to make the game resonable (not skillstacked) We must put some really good players with those those retards to compensate. If We won't do that, it will be like "team with retards" VS. "team with no retards" == skillstack. Jcw87, Demented, knighttemplar: join KMBalancerTesters steamgroup if You are not already in it. As for the todo list: Refactoring code to match updated database schematics Properly logging weapon properties (headshots, roadkills, melee etc) GetPlayerSkill() - this is the last thing needed to enable actual team balancing 1. Is done, not tested yet 2. We don't need it, I deleted the table and stuff. This is not a statistical plugin, it stores some stats but only usefull for balancing. I see no good in storing information about headshots, roadkills and weapon types. Killed, died, gained skill - I only need to gather those. 3. GetPlayerSkill(). This is the next thing I am going to code when I will be able to. I am so tired after work that I don't know my name. If You guys have any ideas how we could improove the plugin or how we can detect end game quitting etc - write it in this topic! It will help!
Next testing session will be at 14:00 CET this Sunday (13 hours 30 minutes from time I post this message). Event is scheduled in the KMBalancerTesters steamgroup so You can check Your timezone there if You need. My test server lags like pregnant slug if more than 9 players join, so I need exacly 9 people. There are two known bugs that will be (hopefully) fixed before the testing session will begin. 1. The plugin has a hard time processing non-ascii characters in player names and it kicks out players. I've talked to sourcemod devs and they told me how to fix it so We should be good. 2. It miscalculates stuff when player is trying to toggle from team to team (without leaving to spectator). This needs a bit of work to refactor things but we should be good. The testing session is just about playing like You always do, only diffrence is that the server is small (9 slots). It would be good to have some good players and some emp_recriuts to check if plugin actually rates them how it should.
The non-ascii characters bug has been fixed and tested. Player toggling from team to team seems to be fixed. Plugin behaves weird, but I must say I expected it to happen. It has done something like putting 2 Ghosts in BE against 4 other players in NF. The fun thing was that both teams yelled that the other team is stacked. I was in NF, I must say the game was challanging, kinda hard because of skilled enemies. For BE it was hard too, they had to leave their cv to engage us into fight, but they needed to look on the radar all the time to check if the CV is safe and they were outnumbered... Players themselves admitted that the plugin has ruined the stack. We tried joining teams in a way it would stack, but the plugin didn't allow us to do so. When someone disconnects, teams are not even anymore, but on the other hand, noone wants mid-game balancing. I figured the plugin can analyze the situation after player disconnect and suggest player to switch teams. If player will not want to switch, teams will be still stacked... but all players would be notified who's fault it is ;-) The remaining CRITICAL things that we need before plugin will be installed on [mush] or KM are: AutoSkillWeight. Currently destroying vehicle is worth 1 point and repairing 100hp is worth 1 point too. Defusing 2 mines is worth... 1 point. This cannot be. The sql queries are simplified. That causes plugin to determine player skill independently per map and per team. If player never played on a given map and given team, his skill is defaulted to 1. This cannot be. This is hard, I need some SQL-masta to help me improove the query. Map win and map loose events are not recorded I also have a bug, normally skill is from 0.3 (idler) to 9, but Ghosts have skill like 120-130. Mathematical formulas will take time to tune The rest of the TODO: fix on demand personal stats display (command "statsme" in console) Reconnecting to database and handling broken database connection Mid-game SUGGESTING players to switch if the teams are suddenly stacked (f.e. because someone left). Commanders would be protected from this. Plugin would try to pick someone that is playing short (like 2min) rather than someone that has build the whole base etc. Detecting players trying to ragequit Detecting players with the same IP address and always putting all of them to the same team. Allowing every player to mark a buddy. When I mark Blight as a buddy and Blight makes me his buddy, We will be always assigned to the same team and we will always play together, we will not be mid-game switched (or we will be both switched). Notice You can only have one buddy and he has to choose You as his buddy too to make this work. Making algorythm not greedy, it should be aware that players join in large groups in the begining of the map instead of handling them one by one Current features: Balancing teams by player skills instead of number of players. Eliminates skillstack. Detecting players that try to mid-game join a weaker team (that may happen f.e. when a good player leaves NF and someone from BE wants to help poor NF) (it allows that) Detecting players that try to leave their team just after they were assigned to a team. This is blocked. When both teams are empty, allowing the first player to join a team of his choice Allowing the player to join the team of his choice if it wouldn't cause major stack.
Thats not a bug. What if i play 10 rounds absolutley retarded so i get less points so i can stack with people that have many points???
Good point, but at the current plugin implementation, this kind of behavior will not increase Your ability to play with the stacked team even if You would lame 1000 rounds in a row I must do something about the ghost's skill. They get many points in short time, they hardly ever die, they kill everyone.
Thats not true... KM is mostly populated with players that not play Empires long enough to be a challenge for people that know how to use the weapons. So its easy to get many kills without much effort. And you know im the king of dying
Ye.. What was it in that round in Escort? 30 deaths? :p It was a fun round thought.. Back to topic: If you play several rounds bad just to be able to stack you are: 1. One hell of a booring guy. 2. You need to get a life. 3. People will get angry, maybe voting for a 'lil bann..
Even if it worked it would be a huge amount of effort just to stack 1 game in 10, which is already an improvement over the randomiser stacking of 1 game in 3. Thanks for doing this Reef, I've skipped on Empires a lot lately because I struggled to find a fun game that wasn't stacked. I play games with a group of 4-5 friends all the time, and we do try to play together. But we also appreciate that sometimes we have to play on different teams to maintain the balance. If we really desperately wanted to play together we would set up our own server or organise scrims, but I don't think it's reasonable to expect to join a public server as a group of 5 and play on the same time all the time. Another thing that I have noticed as a result of clan stacking is that if a group of clanners all join one team, typically other good players on the server see that stack and decide to join the same team rather than be on the side they see as guaranteed to lose. If you took all those unclanned, skilled players and put them on the other team there would probably be a reasonable battle, but those unclanned players do not trust other unclanned players to do the same. Even removing clan tags would not really solve this, as your typical skilled player probably knows the names of a lot of clanned players anyway, it would just hide the fact from the remaining players. You can also get a lot of "churning", where say BE is being repeatedly stacked while NF players play a round, lose, play another round, lose, and quit. New players join to take their place and get caught in the same cycle, so not only are the skilled players stacked but those on that team have also been playing together for 10+ games while others have not. I apologise for the length of this post, it's just a few things I've observed when playing mainly Empires though also other games, and it's the kind of thing that makes me fed up of one and start playing another.
I still say this whole plugin idea where it can track players is a BAD IDEA. But thankfully there are other servers to play on that hopefully will not use this system.
