"Oh, boy, looks like you've gone on quite the winning streak!"
"Better set you up with uncooperative teammates and players with known connection issues to even that right out."
--Riot matchmaking
Today has not been a good day for LoL :c
"Better set you up with uncooperative teammates and players with known connection issues to even that right out."
--Riot matchmaking
Today has not been a good day for LoL :c
Matchmaking algorithm:
Search for all players online 100+ MMR lower than you. Check for multiple reports for AFK, intent feed, verbal abuse, negative attitude or refusing to communicate. Check for 3+ losses in a row. Check for Teemo, Gangplank, Urgot, Fiora or Udyr most played. Place on team with you. Enemy team gets people on winning streaks, with high AOE high CC champs like Orianna, Malphite, Jinx, etc. as most played and who have been honored.
Seems legit.
I don't mean to be rude, but usually if you've gone on a large winning streak your is considerably higher than usual. Usually you lose more games after a winning streak because you are simply playing against better opponents than you are used to.
I wouldn't have made a thread about uncooperative teammates or disconnecting players if i wasn't getting a string of uncooperative teammates and disconnecting players.
Pretty sure Rito doesn't do it on purpose.
But I understand your frustration :/
I don't mean to be rude, but usually if you've gone on a large winning streak your is considerably higher than usual. Usually you lose more games after a winning streak because you are simply playing against better opponents than you are used to.
Actually, it's not the only possibility. It appeaned to me too to be matched to uncoperative player, as my high rank+their low should "balance" that of normal player...
[{quoted}](name=Zirket Was,realm=EUW,application-id=3ErqAdtq,discussion-id=YTsfXbEx,comment-id=00000001,timestamp=2015-05-23T07:06:00.870+0000)
Actually, it's not the only possibility. It appeaned to me too to be matched to uncoperative player, as my high rank+their low should "balance" that of normal player...
After a series of wins this is common for matchmaking to do. It will pair you with players who are all lower than you against a team of opponents who are all higher mmr than any of your team members. It does this to see if you can carry them and deserve to be moved up.
Don't get me started on how stupid it is too put a player the system thinks is ready to advance in a situation where to do so he has to overcome the disadvantage of having a much worse team than his opponents. It makes no sense. It's like saying to Lebron James, ok, you look like the best basketball player in the world, but to prove it you have to beat Steph curry, clay Thompson, Kevin Durant, James harden, and Dwight Howard. Oh and you get to play with 4 pee wee league 8 year olds. Have at it champ.
After a series of wins this is common for matchmaking to do. It will pair you with players who are all lower than you against a team of opponents who are all higher mmr than any of your team members. It does this to see if you can carry them and deserve to be moved up.
Don't get me started on how stupid it is too put a player the system thinks is ready to advance in a situation where to do so he has to overcome the disadvantage of having a much worse team than his opponents. It makes no sense. It's like saying to Lebron James, ok, you look like the best basketball player in the world, but to prove it you have to beat Steph curry, clay Thompson, Kevin Durant, James harden, and Dwight Howard. Oh and you get to play with 4 pee wee league 8 year olds. Have at it champ.
Yeah man, it's a pain. Especially for me, that am a casual player, and play mostly whit friend. It ruin a good hour of bro-playing...
Couldn't agree more. Sadly, this system doesn't place you at your "real elo", but rather the point to where you can no longer carry 4 morons. I guess that is the reason they don't punish people that intentionally feed, they have to gate their elo system somehow.
there is so much wrong with this system and its also the reason why you have more then half your population in bronze rito.
btw rito how did that "soft reset" for season 5 work out? i know that everyone that avidly played stopped it altogether. ow btw what was the point of making me drop from silver 3 and climbing all the way in bronze 1? sure im silver 5 now after spamming nunu jungle which im sure you gonna nerf now that i abused it once to many in a blue moon.
man rito logic is so
ing
This is exactly what's wrong with the current ranking system. To get to promotions you need to win many times. Once you get to promotions, you need to win 2 out of 3 or 3 out of 5. But by the time you get to promotions, the matchmaking system teams you up with idiotic players because "Woops you're winning too much".
If the system could reward more than 30 LP per win and deduce less than 12 per loss, maybe it would be less egregious to try and grind out of the low brackets when your MMR is clearly way too high.
This is exactly what's wrong with the current ranking system. To get to promotions you need to win many times. Once you get to promotions, you need to win 2 out of 3 or 3 out of 5. But by the time you get to promotions, the matchmaking system teams you up with idiotic players because "Woops you're winning too much".
If the system could reward more than 30 LP per win and deduce less than 12 per loss, maybe it would be less egregious to try and grind out of the low brackets when your MMR is clearly way too high.
The system DOES change lp gains and losses relative to ones mmr vs ones division. For example, I had an account last year that decayed from platinum. I played it back up to plat in 5 games from silver. Why? My mmr was so high that despite the fact it was still pairing me against plat players all I had to do was win 50% to get back in a flash, gaining 60+ lp per win and skipping whole divisions at a time. This should be every players goal. To accumulate mmr dramatically higher than the division they are playing in within the first 50 ranked games ever played on that account before lp and mmr gains stabilize at 50 ranked games played. Before this point mmr gains will be dramatically inflated, diminishing from the first game to the fiftieth.
Which is the issue really. I was Gold I, going into Platinum when the last season ended. I went into ranked as soon as it started and it placed me into goddam bronze right off the bat. I learned my lesson though, no more ranked at the start of a season.
Which is the issue really. I was Gold I, going into Platinum when the last season ended. I went into ranked as soon as it started and it placed me into goddam bronze right off the bat. I learned my lesson though, no more ranked at the start of a season.
The start of the season is the worst possible time to play ranked. The greatest variance in matchmaking occurs when all players ranked mmrs are compressed toward an average region, which occurs with the soft resets riot issues at the beginning of every season now. They used to be hard resets... You got challenger in the first season? Well you gotta start right back at 1200 mmr at the start of season 2... The chaos... I have no words for it. Or the amount of complaints that rolled in because of it. Soft resets ever since. I relate this so that you can remember, once upon a time, it was even worse. Things improve. Just very, VERY, slowly with riot at the helm. I've been demanding champion mastery since early in s2 and only just now got it. And it's basically the exact system I outlined almost 4 years ago. Change happens here, but only at a glacial pace.
The start of the season is the worst possible time to play ranked.
Any time you play ranked, is the worst time to play ranked. That's why I don't. XD
"Better set you up with uncooperative teammates and players with known connection issues to even that right out."
--Riot matchmaking
Today has not been a good day for LoL :c
I'm just going to leave this here... Cough Cough
Pretty sure the Gold Master Yi is a smurf just look at that KDA and 17 ranked wins. Also look at that nice number Team MMR, there is no huge gap in that.
this is not how it should be done though. this system keeps that mastery Yi player in gold when in reality he should be in diamond if you dare to put him in such a match. and don't tell me that he gets better "rewards" the loss of such a game is just as devastating as any normal loss which outweigh's the benefit tenfold
this is not how it should be done though. this system keeps that mastery Yi player in gold when in reality he should be in diamond if you dare to put him in such a match. and don't tell me that he gets better "rewards" the loss of such a game is just as devastating as any normal loss which outweigh's the benefit tenfold
The yi isn't being kept in gold, he had 17 total ranked games played and had roflstomped his way to diamond mmr without having the system skip him ahead yet. If he won that game he likely skipped all the way up to plat 4. You guys are goin crazy over that screen shot but all it shows is a very high level player playing on a fresh 30. Matches like this occur at a pretty common rate, and more than likely the team with the "gold" master yi has a significant advantage as he's probably the most skilled player in the game.
You guys need to throw this screen shot out. You will confuse people. There is very little issue with matchmaking or the ranked system at that mmr range. That screen shot actually shows the system working ideally, a player with advanced mmr playing with players of equal skill despite whatever the stupid league system says his skill is. The system is only inaccurate for low level play, and you won't be able to point to a screen shot like this where one obvious element stands out. You'll have to parse through the individual game stats to find any information of value, or have watched the game.
so your TL;DR is...
more then half the ranked population is fuked
so your TL;DR is...
more then half the ranked population is fuked
Basically. Although I would say the entirety of the ranked population is fucked. The problems with the ranking system and matchmaking just affect high level games so little that the problem is almost non existent. It's part of what makes a lot of high level players feel entitled and gives them the mindset that if they got there then nothing is wrong. I'm no different myself. I railed about the system all seasons 2 and 3 then I hit diamond and was mum on the subject for most of season 4 until I decided to level up a smurf for end of season rewards and was incapable of doing it. After a very lengthy discussion with caristinn, hate daddy, Loken stoneheart, jamaree, yenn, and several other high level players many of us arrived at the conclusion that to dismiss the problem would be to do a disservice to the community. For some reason the community over values the opinions of high level players, and we figured if any change was ever to occur some of us would have to be visible on the scene, and be vocal about the issue, despite the ire we knew it would draw from other high level players who want to keep that sacred ground all to themselves. Make no mistake, we are not liked for championing these intended changes, we've upset the status quo at the highest levels and offended a great many parties.
I can't do anything but stay the course though, ever since I became aware of the reality, I've felt that it would only be right if I worked to improve it. There is no reason I have one account in master tier and 13 that are not except for pure chance. I got lucky one time in 14 tries and only after more than 10,000 games.
I think you should make a thread about this topic yourself, because it would start an interesting discussion. I'd like to see some actual statistics about the correlation of in game behaviour and rank. Probably most people could advance only by playing more often, practicing in normals, playing on a better computer and not raging, afk-ing, even though I heard that high ranks can be even more toxic than low ones.
I think you should make a thread about this topic yourself, because it would start an interesting discussion. I'd like to see some actual statistics about the correlation of in game behaviour and rank. Probably most people could advance only by playing more often, practicing in normals, playing on a better computer and not raging, afk-ing, even though I heard that high ranks can be even more toxic than low ones.
No, the higher you climb the less toxic the game is, and the less offended players are by chat in game. There are an equal proportion of troll across all levels, but in general the toxicity level in master and diamond is very very low. These players are all focused more on a win than anything else. Things get tense, sure, sometimes arguments even happen, but it's usually over strategy and what players think is or isn't effective in the match, it's a lot less likely to be over how well some one is playing or who is acting like what.
Id love to see some more data on this subject myself, but I've left you with at least my impression of how things are.
It's common knowledge at this point that ranked at low levels is almost 100% luck based in regards to matchmaking and it's inability to provide adequate matches.
You can duo q to take out some of that element of chance. That is all you can do. And it must be done with another player of quality higher than his division or else the duo matchmaking difficulty penalty will really hurt.
i got an irelia who afk'd mid game to smoke a cigarette
I got two who said "oh I gotta go bake", one said cake, one said bread. I thought it was a joke, they instantly DC'd. During my promos after I went on a winning streak naturally.
[{quoted}](name=Sadistic Veigar,realm=NA,application-id=3ErqAdtq,discussion-id=YTsfXbEx,comment-id=0002,timestamp=2015-05-23T06:52:56.593+0000)
i got an irelia who afk'd mid game to smoke a cigarette
Cigarette Butt Irelia skin confirmed.
During my promos, I got:
(Overall, only one game was lost because of my skill).
Then, I got carried the next game, and after that, I did average on the next one (won too).
Ranked is mostly luck, but it's favoured to you if you are better than your actual rank.
Its clear its not random. If you have a high mmr they whack you with bad teammates.
Wow OP, i'v been saying this for years. Riot match making is trash.
I have probably more than 500 games in ranked and I get matched with people who are under 25 games played at ranked.
How it works... why is it like this? who knows.
the system DOES pair you with lower mmr teammates/against higher mmr opponents to try and "balance" out your win streak. it tries to make everyone have a 50% win rate. It feels and is horseshit but ayyyy whatcha gonna do about it? nothing....nothing at all.
No it doesn't. It pairs you with higher MMR opponents AND teammates. Sometimes your team is better than theirs, sometimes theirs is better than yours. The system evaluates the MMR, it can't evaluate how good each person will perform in that exact game.
Each time your team is better, you'll be more likely to win and thus just continue on your winning spree. But each time their team is better, you're more likely to lose and when this happens then suddenly "OMG Rito complotting against me cuz I won 5 timez cuz Im da best I alweys needz to winz"
[{quoted}](name=Dr Poro,realm=EUW,application-id=3ErqAdtq,discussion-id=YTsfXbEx,comment-id=00140001,timestamp=2015-05-23T19:06:41.045+0000)
No it doesn't. It pairs you with higher MMR opponents AND teammates. Sometimes your team is better than theirs, sometimes theirs is better than yours. The system evaluates the MMR, it can't evaluate how good each person will perform in that exact game.
Each time your team is better, you'll be more likely to win and thus just continue on your winning spree. But each time their team is better, you're more likely to lose and when this happens then suddenly "OMG Rito complotting against me cuz I won 5 timez cuz Im da best I alweys needz to winz"
No. Let me stop you after your second sentence. It does not pair you with higher mmr teammates in the scenario we are talking about (achieving a long enough win streak that the mmr accumulation triggers the system's belief that you are ready to advance) in this specific situation it pairs you with teammates who will be of lower average mmr than you against teams of opponents of higher average mmr. It's designed to do this... It's on purpose, it's the way the system checks to see if you deserve to advance.
Let me also say this so I can do away with some of the misconception in this thread. The fact that these pairings are not equal at low level play actually means almost nothing. The variance within low level tiers is so dramatic that any random silver 5 player can be better than a gold 1 in actual skill and any random silver 1 could be worse than bronze 5. The variance itself is the issue. Not the way the matchmaking decides to set you against a tough team. That part functions just as intended. It's the fact that below diamond level a players ranking bears no relation to his actual skill level that is the inherent problem. The ranking is a reflection of chance, and a players diplomatic skills as well as his versatility. In other words, players who are better at communicating, more capable of playing any role effectively, and who are lucky with matchmaking rank better. And truthfully, communication and versatility mean nothing at low levels anyways, since most games can be won with strange champ picks and coordinated teams below diamond simply don't exist but once every hundred games. Thus, we have the most influential factor on a players ranking remaining. Chance.
This is basically a summarized version of a very, VERY long discussion I had with Caristinn about a year and a half ago. Independently we both arrived at this same conclusion using the same data set.
There are no real good solutions to this problem that don't create problems of their own. That was a whole NOtHER very, very long discussion. The best we could come up with was that individual efforts needed to be considered along with team efforts to creat a composite ranking or a secondary ranked skill aspect. Champion mastery only just came out, but if riot follows our years and years worth of suggestions then eventually a players mastery score and his roles played will affect the way the system ranks and pairs him.
A silver 5 playing on his main is almost never going to be better than a gold 1, unless you are to make the more dramatic statement that plat and bronze is equal, since silver 5's also play with bronze 1's and plat 5's also play with gold 1's.
A silver 5 playing on his main is almost never going to be better than a gold 1, unless you are to make the more dramatic statement that plat and bronze is equal, since silver 5's also play with bronze 1's and plat 5's also play with gold 1's.
That's fairly accurate. Correct. There's very little to separate a platinum player from a silver player. You hit the nail on the head.
don't think thats the case, because I'm plat and I spectate my silver friends all the time. The silvers say this only because they think they're being held down by bad teammates when it is themselves that are bad.
don't think thats the case, because I'm plat and I spectate my silver friends all the time. The silvers say this only because they think they're being held down by bad teammates when it is themselves that are bad.
Everyone has an opinion. Just be aware that what I'm sharing here is a consensus among several master tier level players, myself included but only one of a multitude. The opinion you express is common, it's held by many others just as mine is.
In the end all I really hope is that everyone can agree that the current systems need improvement. It was silly to bring a scoring system designed for individual competition into the structure to function as leagues ranking system when team based games have numerous aspects that can affect outcomes and not all relate directly to the player to whom the ranking is issued.
Slowly, we are seeing riot accept their mistakes and work to fix them. Hopefully champion mastery works out great and we can move forward using it as an additionally measurement of skill.
if you were master tier, I don't think you'd be posting from a 600+ normal win silver 4 account.
if you were master tier, I don't think you'd be posting from a 600+ normal win silver 4 account.
You get real defensive about 1 observation and a no name plat's opinion for a master tier player.
. That was a whole NOtHER very, very long discussion. The best we could come up with was that individual efforts needed to be considered along with team efforts to creat a composite ranking or a secondary ranked skill aspect. Champion mastery only just came out, but if riot follows our years and years worth of suggestions then eventually a players mastery score and his roles played will affect the way the system ranks and pairs him.
There is no way to measure this. I can go 0-9-9 as a midlaner and still have been the 100% deciding factor in winning the game. If I want a good "score" right now, I have to stop tanking and try to fish for easy kills while AFK farming. This is RITO's system, a company that has millions of dollars. They are either not allowed to do their jobs or incompetent.
There is no way to measure this. I can go 0-9-9 as a midlaner and still have been the 100% deciding factor in winning the game. If I want a good "score" right now, I have to stop tanking and try to fish for easy kills while AFK farming. This is RITO's system, a company that has millions of dollars. They are either not allowed to do their jobs or incompetent.
You realize of course that champion mastery takes into account more than kda. In fact, kda isn't all that important in it's calculations. It also takes into account farm rate, warding and vision control, obj control, obj secured and kill contribution %. And I sure hope you realize that champion mastery is precisely the way to do it, it great indicator of personal individual skill and mechanics and can be used in conjunction with mmr to get a more accurate idea of a players skill.
You realize of course that champion mastery takes into account more than kda. In fact, kda isn't all that important in it's calculations. It also takes into account farm rate, warding and vision control, obj control, obj secured and kill contribution %. And I sure hope you realize that champion mastery is precisely the way to do it, it great indicator of personal individual skill and mechanics and can be used in conjunction with mmr to get a more accurate idea of a players skill.
Jaymc1130, you covered the majority of what the new Mastery system calculates. My question is doesn't it also compare your overall stats towards others of your skill level?
Jaymc1130, you covered the majority of what the new Mastery system calculates. My question is doesn't it also compare your overall stats towards others of your skill level?
Correct in a sense. Champion mastery takes into account your individual contributions and weighs them vs the average for all players playing that champ in that role. My understanding is that it's not limited to players within your tier, but the average if all players.
Well then I don't see why they don't use the new mastery system to help them determine whether or not someone should go up or down a tier. This would greatly help roles like; "jungle" or "support" that are in a less carry role. It really sucks playing a support role and going 4-1-27 as Nami, and losing the game because you get a troll, or two on your team.
It would also help determine if someone is just having a bad game, or if they're just trolling.
Well then I don't see why they don't use the new mastery system to help them determine whether or not someone should go up or down a tier. This would greatly help roles like; "jungle" or "support" that are in a less carry role. It really sucks playing a support role and going 4-1-27 as Nami, and losing the game because you get a troll, or two on your team.
It would also help determine if someone is just having a bad game, or if they're just trolling.
Those things will happen in time. Let's give riot the rest of this season to work out any kinks with mastery first, then we can work on integrating it with the ranked and matchmaking systems.
I'll just point out, if anyone has played Bloodline champions (I did back before it became shit STEAM exclusive), they primarily used a system similar to champion mastery.
Each round/game the player was rated based on their contribution to the match in multiple categories. Your ranking went up or down based on whether you won or lost, but the amount was based on that contribution.
For example (I don't remember exact values): Games were 3v3 or 2v2 (excluding tournaments which used a bracket system) Win / loss has a base increment of 10. There are 5 categories: Damage dealt, damage taken, CC applied, buffs applied, amount healed. (I think there were more) Ranking highest to lowest in a category when winning would grant bonus points: 6,5,4,3,2,1 Ranking highest to lowest in a category when losing would cause you to loose additional points: 6,5,4,3,2,1
Most champions were pretty well balanced around those categories. I think the Inhibitor (tank) was the only one with a kit that was capable (not likely) of getting highest in every category ["reduce damage done by enemies" or "increase damage they take" abilities, 2 stuns + silence, a target shield, and a self heal].
This is the style (not exact) system that I think LoL should transition towards.