The game doesn't force you to 50% any more than any elo system does.

Rins Thigh Socks·Oct 26, 2017·15 votes·962 views

Any good elo system will match you with players your elo. If you winstreak and break a 50% winrate, you're gonna rise in elo rapidly because you're playing like you don't belong in your current elo.

So the players you'll match with are higher up, not because the game is forcing your winrate down. It's because it's matching you with players the system thinks you're capable of playing against.

So while it is forcing you to a 50% ratio, that's a normal feature of games with these kinds of ranking systems. They do so because it's trying to match you with players in your skill bracket, for higher quality matches (unless you end up in Plat 5 elo but that's another story) and because you learn better that way than when you're with lower skilled players.

14 Comments

Chief KebbieOct 26, 20173 votes

This implies the only thing you can do is match against more difficult opponents. There are 2 option rito has.

Match you against higher elo, easily the best option
Pair you with weaker players and match you against your current elo

Because rito can do both of these things the matchmaking system does not encourage player development. It encourages you to just wait out your crappy teams or bitch and moan. I would argue a metric that feedbacks on performance over winning will encourage player development and discourage toxicity as players would no longer be able to call "elo hell" and would instead be held accountable for their own performance.

All of which I talked about in another thread.

Rins Thigh SocksOct 26, 20174 votes

[{quoted}](name=Chief Kebbie,realm=NA,application-id=3ErqAdtq,discussion-id=RkU8aByp,comment-id=0000,timestamp=2017-10-26T21:01:53.968+0000)

This implies the only thing you can do is match against more difficult opponents. There are 2 option rito has.

Match you against higher elo, easily the best option

Pair you with weaker players and match you against your current elo

The problem with that second point is that if you are the weak player, why are you playing against the stronger players? That hurts you just to slow down a player that's doing crazy well. The only real option between those 2 is number 1, and that's what Riot tends to do. If you go 3+ wins in a row, Riot doesn't want you to shit over lower level players and your progression is going to go a little more rapidly.

Because rito can do both of these things the matchmaking system does not encourage player development. It encourages you to just wait out your crappy teams or bitch and moan. I would argue a metric that feedbacks on performance over winning will encourage player development and discourage toxicity as players would no longer be able to call "elo hell" and would instead be held accountable for their own performance.

Player development is too abstract to be defined. Everyone here thinks it means you shit all over your laner, go 5/0, and singlehandedly win the game.

But that doesn't win you games, particularly in high elo.

Going 1 for 1 and making your opponent lose 1 and a half waves of cs is more skillful than last hitting someone because your jungler came in, but according to the metrics used by a lot of people here, last hitting off your jungler would be considered a better score.

It's not about picking up 1 kill or making a single game winning play. It's about consistent macro that prevents the enemy from fighting. Tryndamere and Singed players regularly end the games with less than 2 KDs, but they carry because they understand how to throw their weight to win games. To argue winning games is all about bitching and moaning while waiting for teams to get their heads out of their asses doesn't help the player develop nor does going full KDA player.

Chief KebbieOct 26, 20171 votes

The problem with that second point is that if you are the weak player, why are you playing against the stronger players? That hurts you just to slow down a player that's doing crazy well. The only real option between those 2 is number 1, and that's what Riot tends to do. If you go 3+ wins in a row, Riot doesn't want you to shit over lower level players and your progression is going to go a little more rapidly.

From the perspective of the weaker players: they are currently enjoying option 1, playing against a more difficult opponent due to a winstreak. So I'm unsure how that disqualifies option 2 from happening.

You wouldn't be shitting over lower level players, you would be matched at your current elo. You would now need to carry adding an additional requirement to your play. This is the "higher level" you now need to reach. Not sure what you're talking about so maybe I missed the point.

Going 1 for 1 and making your opponent lose 1 and a half waves of cs is more skillful than last hitting someone because your jungler came in, but according to the metrics used by a lot of people here, last hitting off your jungler would be considered a better score.

I would agree. By evaluating your success in game on the win/loss scale however, rito does not reward this behavior. So the only reward is the in-game success of the kill. So people only care about that. By evaluating you on your performance (not your perceived performance, rito's honest evaluation based on the data they have from thousands of games) Rito could reinforce the CS'ing over the singular pursuit of kills. To reinforce this they have to provide feedback on it and the matchmaking system is an ideal way to do this. By laying the onus of promotion on YOUR play alone you would no longer be liable for weaker teammates. This would invalidate the second option as a concern as it no longer impacts your ability for advancement and the toxic players would no longer be able to say 'I'd be plat except I always get weak team'.

DuskDaUmbreonOct 26, 20172 votes

Thank you! Finally, someone understands this and can put it into words properly!

GripaAviaraOct 27, 20172 votes

No, no NO

This is not how elo system works!

Rarely in an elo system you get 50% win rate. You need to be very very evenly matched and play the same players over and over again... If the system would pend towards 50%, nobody would climb. Also, win streaks/ or losing streaks would be incredibly rare more than 4.

That being said, if the system would truly be an Elo system , the discrepancies in game would not be that huge.

I used to play a lot of Heroes of Newerth , where they had a true ELO system, and the games would feel way more balanced in terms of players. The discrepancies here are stupid.

Please stop sending false information if you didn't study the topic, or at least basic game math

PaladinNOOct 27, 20171 votes

50 %

I did the math on my ARAM games, just for the fun of it and as it would be a large enough number to provide reliable data. I didn't really know what to expect...but I don't know what to think of the results either.

ARAM matches: 7048 Total 3595 Wins (51,01 %) 3453 Defeats (48,99 %)

item 3301 cidence? Somehow, I am inclined to say "no" - and then proceed to cook up conspiracy theories about fixed balancing.

But it's interesting. I really didn't think it would be that close.

...And downvoted for making an observation and personal opinion. How typical.

KoggyStyleOct 27, 20171 votes

7.4k arams with 50.5% wr

atleast one modes balenced lel.

option paralysisOct 26, 20171 votes

I wouldnt mind playing against better players, which I did when I was in d2, played against challengers. But what I do mind is that I get worse teammates. I had a d5 autofill going vs. d1 main. Tell me who won that game ? Or when I went against challengers having d2's on my team. All this shit while I had a 60% win rate. Then I dropped to d5 and stopped playing ranked. Its just me but I dont like playing a game where I lose in champ select.

Lucky MonikerOct 26, 20171 votes

The main problem with an ‘ELO’ system is that it was designed with 1v1 in mind. The average ELO may look about normal (sometimes) but when you put 1 higher ELO and balance it against 5 medium ELOs it doesn’t always pan out ‘equal’ and can lead to a very ‘unfun’ game. Not to mention the ‘auto fill vs main’ argument presented already....

KythersOct 27, 20171 votes

I wonder what the IQ cutoff is for being hardstuck in silver

ToBeSomebodyOct 27, 20171 votes

The ranking system is going to have to do one of two things:

Punish you for being good Punish you for not being good

It's a delicate balance, and it's hard to justify either way. Riot chose the former.

There are systems in place to allow the better players to climb faster (division skipping and even promo skipping), but generally speaking you're not necessarily going to make it to gold/platinum/diamond even if you are that level of player. The system as it exists does require a heavy amount of investment for the average player to reach their peak, but this also helps prevent lower-skilled players from having too many poor experiences being matched against players they shouldn't.

Darkdemon653Oct 27, 20171 votes

Regardless of how the system works, there's a difference between reaching a point where someone naturally wins 50% of the time and the system intentionally giving people free wins and losses in an attempt to force a 50% win rate.

Dr DogOct 27, 20171 votes

I don't know about others but the problem isn't getting paired against high elo players, its the fact that on top of that you never get teams that match their level, today alone I've vsed plats and dimonds as a b2 player BUT the problem is that I got paired with a level 18 and another time a level 27 (you can check my op.gg and fact check it as it was just today) with mostly bronze and silver for the rest of my teams so ya it does feel like they intentionally try to force you down

EbololOct 27, 20171 votes

League in general is a game that isn't really in your own hands. You're above average play can only shift your overall winrate a few percent, but majority of it relies too much on your teammates. The best players only have winrates of 52-56% usually, and I've seen D1 players with a sub-50% winrate.