The carpet ban put a stop to most of it. BSID hardly has any active players, so clan stacking for them is kind of impossible. Fixed. I'm going to be working with spawn soon to get a larger sample data set so we can evaluate the skill system. If the system is accurate, we will move to create some kind of auto balance system. I think that any Empires auto balancer needs allow veteran players the most amount of freedom (playing with friends, picking NF or BE, etc) possible while still preserving fair play.
It won't be such a big issue when we have balanced teams...or rather, it'll be the same where you can kinda live with it. To be honest, the real problem is educating players probably. The real reason people stack (although we do have a lot of trolls in Empires) is because they actually want to play with a competent team.
No, I'm saying deal with it. Stacking will always happen. The only thing which needs fixing is blatant number and clan stacks.
Number stacks don't currently exists due to mp_autoteambalance. As for clan stacks... By what logic? Clan players are somehow inheriently "worth more" than clanless players of equal (or perhaps even greater) skill? Maybe it's the part where the clanners may be using a third party voice com? Clan stacks are like regular stacks except it's easier to say "oh yea that's stacking because look at all the tags". So if I quit BSID then I'm no longer as good a player? Suddenly my ability to play well in a squad is destroyed? Suddenly without being in TS I feel too small to use the ingame chat to talk to my team? If I'm not in the clan but I join a clan TS, does that make me worth a clanner, or does the tag in my name actually do it? I propose two types of stack, skill and numbers. Skill stack can be countered by Skill stack (aka the "counter stack") or by Numbers stack. Surely we could agree that 10 clanners vs 10 average pubbers is uneven... but what about 10 clanners vs 30 average pubbers? 50? Sooner or later we'd have to both agree that the team with the greater number has an advantage and that would be a number stack. Finding a balance where "X skill is equal to Y numbers" is a possible solution. As for the how, the rating, I suggest a very simple yet (apparently in my limited testing) effective system. If someone dies, they lose rating. The larger the gap between them and their killer the more points they stand to lose. Their killer gets whatever points they lose. If your team wins, you get rating based on your score. If you lose, you lose rating based on a percent of your current rating. Points are finite, meaning the winning team only gains as many points as the losing team gave up. Dying to someone with an equal or greater score costs two rating points, Dying to someone with significantly less rating than you costs more than 2 depending on just how much less they had than you. This means in a back and forth (1v1 District where we both go back and forth for example) our ratings would average.
The problem with a BSID stack isn't that BSId is godlike because of their tags, it's because they KNOW they're tilting the game in their favour, and they're doing it on purpose.
I'm going to need a more compelling argument than that to cease developing what I feel is a solution to one of the bigger problems in day to day play. How about move past the "yay or nay" concept and consider that there already is an auto-balancer, and all it does is makes the teams even +/- 1. Do you think that's enough to make balanced games? The debate isn't auto-balancer or no auto-balancer, it's what more needs to be done to balance a game so it's not a steamroll whenever a group of vets decide to stack one side and/or how to prevent a stack? Or we as a community can decide that stacks aren't actually problems, they are just challenges for teams that may have no chance of winning... and we move on, but if we do that, I expect BSID and every other clan can stack freely because we're all accepting it as "mp_autoteambalance does enough to balance teams". You can't know their intent any more than I. I, however, can know my own intentions, and can say you are wrong when talking about me, and that you can't know when talking about any of them. Regardless, the balancer I've made targets fixing this by letting players join a team provided there is "space" for them, and possible variations include balanced ratings and balanced numbers as well as balanced ratings and 2v1 type numbers ratio (or more, there's no limit provided a large enough difference in ratings). The individual player ratings combine to make a team rating, and if the teams are "close" you can choose either team you want regardless of your rating. Once enough stack one side, players would have to join the other side to balance it out. In my testing this does not stop players from joining the team they want unless all the good players want to stack a side, and since that's what it's designed to stop, it stops them. I will try to throw together a demo video explaining how it comes together because I get the feeling people think this somehow forces you to not play with friends... in fact it does the opposite if anything, it allows you to play with friends, it stops you from playing with too many skilled players when it would lead to a steamroll of a game (or suppliments the lesser skilled team by allowing more players onto it). It creates games that are in my opinion more enjoyable because you can be outnumbered by less skilled players which leads to more close calls, narrow escapes, and epic moments. It also makes the less skilled players feel useful because they can gang up on a vet and stand a chance.
We talked about that a month ago. It's all relative. Clan stacking *can* be different than regular vet stacking, but most of the time is *not*. Perhaps, but the system needs to be flexible to determine that on the fly, and account for the skill of all the players involved. We'll certainly see once the system starts collecting data from the major servers. *yawn* /me autobalances complete /me lols That's why SourceBans has Steam group banning... Under no circumstances will the major server networks consider steamrolling by vets (purposeful stacking) acceptable... I'm going to stick up for spawn here... He's relatively new to Empires, and even newer to BSID. I don't think associating him with the worst of BSID simply because of his tag is acceptable (yes, I did in the past, and at the time, he was directly associated with the aforementioned players). Most of us have been here since 2.12, and we know most of the player base, and certainly the history of Empires and its clans quite a bit better than you do. Interesting. Very interesting.
How do I get a hon account? (shit... forum admins are going to kick me this time for being off topic... hmm lets hide this message with some EMP talk: I really dont like the points system in empires. If I go and destroy as eng or ninja a whole enemy base with rax, vf, radar then i get... 3 points... Pfff I might as well just camp around my squad getting free points. So I believe the base of this classification system based on points for balancing is rotten. It will also tend to drive people away: "hell i ended up with the stinky NF (or BE whatever i am not being racist here) team... I am out of this server". And jezzz we only have about 1 or 2 populated servers at a time. We will drive people to play more Hon (please an account???)
Hmm no I just want to play with BE these days because I believe the research is unbalanced towards it. (Mark IIs with ER). (like 2 versions ago -was it 2.22??- it was NF with the super lights). Oh common dont nerf the ER cannons now!!!
This is an absolutely horrible system. It really is. Making people lose points for dying will do nothing more than encourage camping. Which it DID around the time g4tc stats was on and you lost points for dying. And causing people to lose points based on their team losing? People will leave the losing team far earlier, and join the winning team far more often. Clan stacks will actually be worse. I know for a fact I'd encourage EPIC to stack as a result of the stats system, because even if it is 10v30, we'll still be getting way more points than the other team. More people to kill, easier kills, we'll know the system better. A stats system makes things worse, ESPECIALLY when hooked into balancing. At the very most, a system in which someone who truely sucks is worth 0.5 players, and a top end vet is worth 1.5 players. Anything either side of that, no matter how much better or worse they are than someone who is truely rated at 0.5/1.5 and not affected by the cap, will completely skew things. Reef made an autobalancer system, and I'd still like to try it out. It pretty much uses points alone as opposed to WHO you kill and which team you were on. It still won't compensate for consecutive point whoring, which may mean it won't be suitable. And saying "ok we can go around it by implementing this idea" is just a hackjob. It's an inherent problem with the entire system. I used to be all for autobalance stats type things, but after seeing it in practice, I totally see why Dubee used to hate it. Balancing people's stats based on other people's stats (How many points you get for killing a nub, etc) will NEVER work. There's far too many variables to take into account. The best thing you can do is balance how many points you get for things in empires in general, then make the odd exception, such as consecutive healing points on the same person (point whoring underwater etc). That is the ONLY way it will prevent point whoring and stop people adapting their style of play to the stats. Punishing people for deaths, for noobs killing them when they're on 1hp, etc will not work. Simple as. Hell, giving points for killing isn't even a great idea, damage dealt might be better, for stats at least.
Going to try to condense into key points... "This is an absolutely horrible system." - No it's not, but what you describe after this sentence is. You did not describe what we are all talking about... again leading me to believe I need a presentation or something. "Making people lose points for dying will do nothing more than encourage camping" - Do you realize (honest question) points and rating are not related? Meaning if you kill 20 people and get your second skill but then die a bunch, your score doesn't go down and you lose the second skill... Everything I described was "rating points" not "score". With that in mind, it won't encourage camping because the most important thing you can do is win the game. I'll have to put together a presentation tonight because it's hard to explain in this format... but when you kill someone you get 2-5 points maybe, probably just 2. When your team wins, everyone on the losing team loses a percent of points (on a 20v20 this would be somewhere around 600 points total lost) and distributed to the winning players based on their scores (yes scores as in ever two sabs is a point, every 100 repair is a point etc). So if you camp the whole game and kill 10 people you will probably get 20 points. If your team loses, you'll lose probably 30 points. Moral being, if you aren't helping your team, they are more likely to lose, and if your team loses it's worse on your rating than a few deaths. "And causing people to lose points based on their team losing? People will leave the losing team far earlier, and join the winning team far more often." - No it won't, because it doesn't calculate who owes points based on who's currently on the team, it calculates who owes based on who was on the team previously. If you jump to the winning team (assuming you can, since as they win the ratings are shifting meaning there's a really high chance the winning team will have a higher rating by the time you want to switch), you'll still owe points because you WERE on the losing team. After a certain amount of time it drops this taxing, but if it's late in the game you will still owe even if you teamswitched to the winning side. First sentence you make it sound as if they should get points for their team losing, or that losing has no bearing on your teams skill, I'd argue that results (as in which team won) is the biggest factor in measuring the otherwise unmeasurable (such as scout sabbing, helpful engis, utility roles, etc). "Clan stacks will actually be worse. I know for a fact I'd encourage EPIC to stack as a result of the stats system, because even if it is 10v30, we'll still be getting way more points than the other team." - You mistake a high rating as a good thing. Do you know what you get for having a high rating? "At the very most, a system in which someone who truely sucks is worth 0.5 players, and a top end vet is worth 1.5 players. Anything either side of that, no matter how much better or worse they are than someone who is truely rated at 0.5/1.5 and not affected by the cap, will completely skew things." - Those are pretty extreme values imo, here's the current values. Upon entering the server for the first time, your SteamID is assigned 200 points. You can never have less than 100 (meaning if you hit 100 and someone kills you, they don't actually get any points because you have nothing left to give). There is no max, however the highest rating I've seen thus far was 369. The lowest rating I've seen was about 128 and Duke was trying to see how low he could drive it... took probably 5 matches, and in one one game following (playing serious) he was right back into the 200+ range where he belonged. So with the current numbers, the worst player is 130 (we'll divide by 260) or 0.5. The highest rated player is 370 or 1.42. Now the difference is that your suggestion is far too static and mine meets your current criteria but can shift as time goes on. "The best thing you can do is balance how many points you get for things in empires in general, then make the odd exception, such as consecutive healing points on the same person (point whoring underwater etc)." - This is a separate issue, and I agree. Part of this balancer uses player ingame score (score as in you get skills for 20,30,40 points), so tweaking the amount of points handed out for tasks will be important to make sure the players on the winning team get their fair cut of the rating points dispersed. "That is the ONLY way it will prevent point whoring and stop people adapting their style of play to the stats. Punishing people for deaths, for noobs killing them when they're on 1hp, etc will not work. Simple as." - Again, good players will wish they could reduce their score without dumbing down their play. It's not a punishment to lose points for dying, it's fair, and it's what keeps your rating RELATIVE. You see, this is a finite system with relative stats, so your rating only means something compared to someone elses. "Hell, giving points for killing isn't even a great idea, damage dealt might be better, for stats at least." - No, because causing damage and not killing is worthless if there's an engi/armory around, in fact you may have just given the enemy team squad points or upgraded an enemy player in the case of an engi being around. Good job. Me needs video, because almost every negative comment I've seen regarding this system has been misconceived. It's nowhere near as bad as most people posting think it is. LOL @ my lame attempt to "condense". Step it up Empty and read. Or here's your summary: Ratings based system good on paper, good in limited testing, values may need tweaked. People who don't fully understand system have a lot of concerns. I will make a presentation to explain the system so they can be informed enough to make valid concerns. Will work on video tonight.
Stacking is certainly a problem, but I'm not convinced a better autobalance system is the solution. The other option is to make it easier for the losing team to make a comeback. This can be done by either 1. Hurting the winning team (e.g. upkeep or escalating costs for more buildings and vehicles, etc.) 2. Helping the losing team (e.g. allowing them to salvage dead vehicles, make starting buildings free (or cheaper), etc.)
Maybe you don't realize this, soundspawn, but people like me will obsess over a rating system and do things just to maximize my rating to be number one. If I lose points for being on the losing team, then I'm going to go to spec before my team loses if I can mitigate some of that loss. If I lose points for dying, I'm going to play to never die, ESPECIALLY not against noobs, since I'm going to lose more points to them. It's really as simple as that. Basically if there's a way to alter my game playing style to minmax my rating, I'm going to, and I can also guarantee you that I'm not going to be the only one to do it.
And there is lots of people who are like this! The g4tc stats proved this when 90% of teams were rifleman every game until they were removed from the public. Game play will diminish.
