Subscribe
All of Badlion's current problems(megapost)
Posted on 04 August 2016 - 10:13 PM
Archybot wrote
Anonymous Quote...
Have you ever played a server with not enough devs? They die out faster then servers with good devs that don't listen - look at shotbow/mcsg.
Anonymous Quote...
Currently there is an admin above you on global with less than a 1/3 of the games you've played. We have several admins/devs that are capable of top 50 global, one finished 11th last season. A developer finished top 5 a couple of seasons ago when he tried (S7 or 8, can't remember) on an alt. Ginie is easily as capable of being top 10, as is halfcreeper. I used to be a top player before I developed carpal tunnel and can't even finish a single match of gapple before my hand cramps. Badlion was literally created by me for my faction/team and I to practice on.
And I'm pretty sure you don't have to be a top player to make decisions that are good for competition. You think top players make the design decisions for League of Legends? You think only PRO's make the balancing decisions in CSGO?
So why did u get top players to test KB???
Posted on 04 August 2016 - 10:15 PM
Irelevance wrote
So why did u get top players to test KB???
Two reasons, one of which would happen:
1. If they did it correctly, we would have the best KB ever
2. If they failed, prove my point of just because you're a top player doesn't mean you're good at game design.
Posted on 04 August 2016 - 10:23 PM
Archybot wrote
KarliXon wrote...
Danteh said it's fine, other then the fact that there is an issue with the matchmaker.
In reality, here is the issue:
Our placements spread people too thin. If all the people who got placed incredibly high (E5 or higher) would wait a month, then try to play again, they would easily be able to play matches as people would have spread out by then. In reality we need to shrink the placement range (currently it's like anywhere from D5 to B5) and just make the queue system enlarge a lot faster (like 3x as fast). No one but the top 20 should be waiting more then 15s for a match, max.
Look Archy it may be a better ranking system but all I know for sure is badlion Has not been getting as many players on as they had in s12. Its your server you can do whatever you want with it but I really don't see why u had to change everything so much. Yeah you wanted a better skill curve and more even ratings but there was nothing wrong with elo. Idk why your saying that every1 was complaining about elo I saw no1 complaining. Completely your choice tho
Last edited on 04 August 2016 - 10:43 PM by iStrafe
@Archybot just do the community a favor for once and listen to us. BRING BACK ELOPosted on 04 August 2016 - 10:54 PM
The idea of the matchmaking system is good. Into practice it evidently hasn't worked. There are a lot of small bugs and issues involved in it. People don't like change. It's human nature for you to be sceptacle and pessimistic of something you don't understandBecause of season 13 a lot of players think badlion is going to die like mcsg. Wrong! At least the badlion staff team acknowledge the community and its demands in contrast with failed servers like mcsg. The owner if I am correct has not logged on his own forums in months. Badlion has definitely lost some players but I genuinely do not think this server is going to die now. After all the bugs related with the season should be fixed, I don't see how the server could carry on to lose players.
Posted on 04 August 2016 - 10:57 PM
@Archybot I thought people were complaining about the queuing system more than elo in s12… the only person that may have complained about the elo system was danteh who complained about getting +1 from 2ks… who else?Last edited on 04 August 2016 - 10:58 PM by rkt
Irelevance wrote
badlion Has not been getting as many players on as they had in s12
smhhhhhhhhhh
Last edited on 05 August 2016 - 12:21 AM by Hivlik
MasterGberry wrote
I have 100 ELO more than this person, therefore I am 100x better than him. It just means you have a number that is 100 higher, but it cannot actually be converted to a skill level system.
If Player 1 is 700 elo above Player 2, then Player 1 will gain 1 elo for wins and Player 2 will gain 31 elo for wins. In theory, if 2 players are correctly rated, then out of 32 matches they will end up exactly as they were before the 32 fights. What this tells us is that in 32 matches, if somebody has 700 elo more than somebody else, they will beat them at a ratio of 31:1. If two people have a very similar elo, then they will fight at a ratio of 16:16. I hope you see where I'm going with this - take the elo that you would lose if you lost the match against said opponent and divide it by the amount of elo you would win if you won against that opponent, and that's how much better you are than he is. If you'd lose 16 elo for losing a fight but you'd gain 16 elo for winning that fight, you are in theory (16/16) times better than your opponent. If you would lose 31 elo for losing a fight but would gain 1 elo for winning a fight, you are in theory (31/1) times better than your opponent. I'll shortly be editing in the rest of my response to you, but that line specificallycaught my eye.
MasterGberry wrote
Minecraft matches are very fast. People play 1000s of matches of Minecraft in a season. Chess games are very long and you have the opponent in front of you quite often.
Another point I've seen you make is that a number system not only disrupts a friendly environment but in fact augments a toxicity that otherwise might not exist. The blow of losing 31 elo to a hacker just sets people off. And honestly I would have to agree with you - people obsess over numbers a ridiculous amount. It's a basic human instinct that game designers exploit all the time. The thing is, you must understand that every yard you take away in how much your system adds to a toxic community, you lose a mile in how addictive your game is. Elo being addictive has made it popular, and it's given Badlion's arena gamemode a high player retention. This is a HUGE key to Badlion's success so far, and as far as I can see to let go of it now would be a mistake. And honestly, addictiveness aside, what do you think is worse: losing 31 elo to a cheater or getting deranked to Gold? I can't imagine that the toxicity is much less in the new scenario compared to the old one, if even at all.
If there's anybody that knows about an elo system's flaws, believe you me, it's me. I understand that elo inflation has a whole web of problems connected to it, and that it really sucks lagging out to a dude to lose 31 elo. But something that I think the community time and time again overlooks about the elo system is that elo rates a player based not only on his skill but on his consistency. Why not amply reward the player that wins 100 matches in a row over the guy that loses 2 of them to a fluke? Minecraft doesn't have that high a skill cap - a couple times already we've seen metas in different ladders or gamemode hit their endgame. Tylarzz and Danteh are very closely matched if not perfectly so. So why is it that Danteh constantly outranks Tylarzz, and by a mile at that? It's because Danteh is through and through more consistent, and the elo system rewards him for that. Because although it's not that hard to hit the skill ceiling in this game, even in the most complicated gamemodes, it takes a fuckton of skill to be very consistent. As far as my (admittedly limited) research has shown me, MMR is used in random team based games like League, CS, and Overwatch. Because how consistent can you be when your win/loss performance is based entirely on the random variable that is your teammates' competence? But we have the advantage of having a 1v1 scenario, or if it's bigger parties, the teams are predetermined and non-random. That makes a system like elo where consistency is judged absolutely viable. Why should our rating system not capitalize on this opportunity?
Something that I wanted to learn about to make this argument was the failed attempt to move away from elo for Season 5. I know it was a long time ago, but I think the theory behind the initial change as well as behind the decision to revert could really help us in such a similar situation. I'd be interested in hearing about it.
Posted on 04 August 2016 - 11:06 PM
HIVLIK wrote
MasterGberry wrote...
If Player 1 is 700 elo above Player 2, then Player 1 will gain 1 elo for wins and Player 2 will gain 31 elo for wins. In theory, if 2 players are correctly rated, then out of 32 matches they will end up exactly as they were before the 32 fights. What this tells us is that in 32 matches, if somebody has 700 elo more than somebody else, they will beat them at a ratio of 31:1. If two people have a very similar elo, then they will fight at a ratio of 16:16. I hope you see where I'm going with this - take the elo that you would lose if you lost the match against said opponent and divide it by the amount of elo you would win if you won against that opponent, and that's how much better you are than he is. If you'd lose 16 elo for losing a fight but you'd gain 16 elo for winning that fight, you are in theory (16/16) times better than your opponent. If you would lose 31 elo for losing a fight but would gain 1 elo for winning a fight, you are in theory (31/1) times better than your opponent. I'll shortly be editing in the rest of my response to you, but that line specificallycaught my eye.
Typical Hivlik with going in depth with mineman things
Posted on 04 August 2016 - 11:37 PM
@Archybot can u reset the stats once the rating system, queuing system, and all the other main bugs are fixed?Posted on 04 August 2016 - 11:38 PM
Tyguy95 wrote
@Archybot can u reset the stats once the rating system, queuing system, and all the other main bugs are fixed?
Posted on 04 August 2016 - 11:40 PM
HIVLIK wrote
MasterGberry wrote...
If Player 1 is 700 elo above Player 2, then Player 1 will gain 1 elo for wins and Player 2 will gain 31 elo for wins. In theory, if 2 players are correctly rated, then out of 32 matches they will end up exactly as they were before the 32 fights. What this tells us is that in 32 matches, if somebody has 700 elo more than somebody else, they will beat them at a ratio of 31:1. If two people have a very similar elo, then they will fight at a ratio of 16:16. I hope you see where I'm going with this - take the elo that you would lose if you lost the match against said opponent and divide it by the amount of elo you would win if you won against that opponent, and that's how much better you are than he is. If you'd lose 16 elo for losing a fight but you'd gain 16 elo for winning that fight, you are in theory (16/16) times better than your opponent. If you would lose 31 elo for losing a fight but would gain 1 elo for winning a fight, you are in theory (31/1) times better than your opponent. I'll shortly be editing in the rest of my response to you, but that line specificallycaught my eye.
Posted on 04 August 2016 - 11:44 PM
sewl wrote
Tyguy95 wrote...
I know I would
Posted on 04 August 2016 - 11:47 PM
Tyguy95 wrote
sewl wrote...
I know I would
Last edited on 05 August 2016 - 01:24 AM by MasterGberry
Immense wrote
GBerry you are doing this for your partnership I can almost guarantee it. If you cared about the server and what so many people want, you would give us the ELO system and other things people want. Instead, you are using a system that YOU think is going to work. So what if it "WORKS" like it is supposed to? If it's not what the community wants, and you are doing something because YOU believe it is for the best, in this case that makes you a bad server owner. I have nothing against you, and Badlion has practically been my favorite server for over 2 years now, even when I was banned back in the day for cheating without even knowing I was cheating(May 2014). Do what you want, see how it works for you. If you think this is going to work and it isn't absolutely perfect in a way that no one can complain (which by the way isn't going to happen) then I guarantee as I said above that this is all about that partnership and you simply don't care about what we WANT.
Nope. What is for the partnership and ourselves (we already wanted to do this since season 11 was being made) as well is the micro-server approach versus the 1000 player servers. The rating system is 100% our decision and we have some improvements that I linked in the other post coming to it soon. Players should be quite happy with it soon.
@Immense
Last edited on 05 August 2016 - 01:21 AM by MasterGberry
HIVLIK wrote
MasterGberry wrote...
If Player 1 is 700 elo above Player 2, then Player 1 will gain 1 elo for wins and Player 2 will gain 31 elo for wins. In theory, if 2 players are correctly rated, then out of 32 matches they will end up exactly as they were before the 32 fights. What this tells us is that in 32 matches, if somebody has 700 elo more than somebody else, they will beat them at a ratio of 31:1. If two people have a very similar elo, then they will fight at a ratio of 16:16. I hope you see where I'm going with this - take the elo that you would lose if you lost the match against said opponent and divide it by the amount of elo you would win if you won against that opponent, and that's how much better you are than he is. If you'd lose 16 elo for losing a fight but you'd gain 16 elo for winning that fight, you are in theory (16/16) times better than your opponent. If you would lose 31 elo for losing a fight but would gain 1 elo for winning a fight, you are in theory (31/1) times better than your opponent. I'll shortly be editing in the rest of my response to you, but that line specificallycaught my eye.
MasterGberry wrote...
Another point I've seen you make is that a number system not only disrupts a friendly environment but in fact augments a toxicity that otherwise might not exist. The blow of losing 31 elo to a hacker just sets people off. And honestly I would have to agree with you - people obsess over numbers a ridiculous amount. It's a basic human instinct that game designers exploit all the time. The thing is, you must understand that every yard you take away in how much your system adds to a toxic community, you lose a mile in how addictive your game is. Elo being addictive has made it popular, and it's given Badlion's arena gamemode a high player retention. This is a HUGE key to Badlion's success so far, and as far as I can see to let go of it now would be a mistake. And honestly, addictiveness aside, what do you think is worse: losing 31 elo to a cheater or getting deranked to Gold? I can't imagine that the toxicity is much less in the new scenario compared to the old one, if even at all.
If there's anybody that knows about an elo system's flaws, believe you me, it's me. I understand that elo inflation has a whole web of problems connected to it, and that it really sucks lagging out to a dude to lose 31 elo. But something that I think the community time and time again overlooks about the elo system is that elo rates a player based not only on his skill but on his consistency. Why not amply reward the player that wins 100 matches in a row over the guy that loses 2 of them to a fluke? Minecraft doesn't have that high a skill cap - a couple times already we've seen metas in different ladders or gamemode hit their endgame. Tylarzz and Danteh are very closely matched if not perfectly so. So why is it that Danteh constantly outranks Tylarzz, and by a mile at that? It's because Danteh is through and through more consistent, and the elo system rewards him for that. Because although it's not that hard to hit the skill ceiling in this game, even in the most complicated gamemodes, it takes a fuckton of skill to be very consistent. As far as my (admittedly limited) research has shown me, MMR is used in random team based games like League, CS, and Overwatch. Because how consistent can you be when your win/loss performance is based entirely on the random variable that is your teammates' competence? But we have the advantage of having a 1v1 scenario, or if it's bigger parties, the teams are predetermined and non-random. That makes a system like elo where consistency is judged absolutely viable. Why should our rating system not capitalize on this opportunity?
Something that I wanted to learn about to make this argument was the failed attempt to move away from elo for Season 5. I know it was a long time ago, but I think the theory behind the initial change as well as behind the decision to revert could really help us in such a similar situation. I'd be interested in hearing about it.
No Hivlik I am 100% correct. I very well understand how the math and the logic behind the ELO system work. It helps to compare 2 people playing a match against each other. It does NOT help to create a proper leaderboard where you know what percentage you are in. Once again argue with me on the correct points here, the matchmaking, and the poor representation of the rating system, not how ELO is the best thing in the world. MMR does not even do as good of a job in teamed matchmaking like you are trying to claim it does. It does not account for individual skill/performance at all in any major competitive game (which it should).
Last edited on 05 August 2016 - 01:30 AM by zillionth
idc we have Uhcs stillPosted on 05 August 2016 - 01:28 AM
MasterGberry wrote
KarliXon wrote...
I am gonna just write a short essay in response, this is too long and too repetitive for me to spend more time repeating what has already been said publically.
Chess =/= Minecraft. Minecraft matches are very fast. People play 1000s of matches of Minecraft in a season. Chess games are very long and you have the opponent in front of you quite often. There is also no way to cheat in chess. ELO does not make sense for video games that are on a large scale. Since most people here are probably un-educated: Badlion is the one that brought ELO to Minecraft. We are the original creators of any form of a skill based ranking system in Minecraft. We are always going to do what is best to create a competitive and community friendly environment at the same time. A number that people obsess over is not the way to do it. Who defines what is good and bad in ELO? Suddenly being a certain number in ELO at one ladder and not another makes a difference to how things are calculated. It is not the communities responsibility to determine who is good and who is bad in terms of skill level. It is ours. ELO does not provide this. Go look at Season 1 and now go look at Season 12. 1000 ELO difference. Not a good skill based system at all. Danteh is not 1000 ELO better than HattoWolf was back in the day. ELO cannot be translated to skill. I have 100 ELO more than this person, therefore I am 100x better than him. It just means you have a number that is 100 higher, but it cannot actually be converted to a skill level system. I am Diamond 1 vs my Platinum 3 friend. This has a comparison of skill since it is a better measurement of skill for someone who does not understand anything about ELO or the Badlion community. "Oh I a tier (Diamond vs Platinum) higher than my friend and I am also two divisions (1 vs 3) higher than my friend." In addition we are the ones determine the various skill levels and will have public information about the skill levels (e.g. Diamond being top 2%) and such. There is absolutely no benefit to ELO over our new system besides people just being stubborn people and believing that ELO is the best thing in the world, when the reality is you should be bitching about (1) the placement matches [already being fixed] and (2) the matchmaker [already also being fixed]. I made it very clear in this post: https://www.badlion.net/forum/thread/129730 that the real problems are being worked on and not the "fake problems" that the community thinks are the issues when in reality they aren't. The new ranking system helps to fix things that ELO does not account for such as winning/losing streaks, not punishing severely for messing up one match, etc. The reason the MMR system we are using was created was to fix the problems that ELO cannot properly handle or solve. Overall the new rating system provides a more user friendly environment rather than obsessing over a number that doesn't really mean much to the new coming players. As I stated earlier, we are the ones that brought ELO to Minecraft, we are now going to be the ones to bring in a new ranking system into it as well that does a better job than ELO.
From what we are seeing publicly people are satisfied with the KB changes made a few days ago. The number of complaints have dropped significantly about the knockback and the gameplay so your #3 doesn't belong in this thread.
As far as chronic cheaters we have some upcoming changes to the ban system and the appeal system that will be cracking down on people cheating harder very soon. People are going to find out very fast that if you are going to (1) ruin the experience of others and (2) waste our time to track you down that all of your accounts will be locked out and a blanket IP ban will be issued very fast.
As far as SG 2.0 maps go they are going to be randomly picked when u join the queue instead of voting for them. That will solve that issue.
We are always hard at work here at Badlion trying to make things the best we can. As explained in the post I linked above there has been a lot of work going on in the background too and I have four developers dedicated to fixing up ArenaPvP Season 13 still at the moment so we can get it all in the 110% state that it should be. I have other things that need to be worried about too and I am dedicating so many of our resources at the moment to ensure that Season 13 is fixed so that we can get back to our 7-8k peaks and eventually grow further. I guarantee by the end of this month at the latest Badlion will be at it's best again rather than the state it is now.
Why don't you just make a poll and do what the community wants not what the admins want because the community is what keeps the server alive and if you dont supply them with what they want then they will stop playing. You can just say "Oh, its not a big deal if they dont like it then they dont have to play" but players liked Season 12 much more, you had alot more players back then and now look. Most of the community hates it and your losing players. Thats all im going to say.
- PizzqWithAQ
Posted on 05 August 2016 - 01:51 AM
This is 100% my opinion and I'm just speaking my mind even though I pretty much just play arena for fun nowI was 2399 last season and I loved the hit detection I loved the kb and everything about elo. Badlion season 13 is new to everyone but I think people like to stick to what they know and not something totally different like the rating system. I just think the whole community doesn't like big changes in the server. I have played on this server for a while now and it's been really fun but as of lately arena has gotten boring their was just something in elo that made numbers exact so u knew how close u were to beating your record or applying for a clan and saying that u were 2.3k elo instead of saying Emerald III. I don't think anyone would listen to me because I'm not very known but I don't think the community likes any of this but the positive part is it brought a lot of players from hypixel ( some that hack and don't know how good the anti cheat is lol)
I don't think this reply made any sense it's like 1:51 am but if u know what I'm talking about then good.
Posted on 05 August 2016 - 02:05 AM
200th post :D
Thread is locked