suggestion : ranked solo queue peer evaluation

12
Comment below rating threshold, click here to show it.

Sparth

Junior Member

08-21-2010

Due to many frustrating ranked solo queue matches and seeing my rating constantly decrease due to people leaving, going afk, or just plain not wanting to fight as part of a team I have been doing some brainstorming about how this could be fixed. And I'm happy to announce that an idea has come to mind.

The idea came to me from my undergraduate degree. I did a business degree so there would always be a lot of team projects to do with classmates. And one of the key components a teacher would use in grading would be peer feedback. In other words, if everyone on the team disliked an individual's contribution to the team, their grade would be altered downwards. Similarly, if peers felt an individuals contribution was above average, then that individuals grade would be pushed upwards.

I think this would be a good feature to implement into solo queue. As it is, solo queue rating is purely a reflection of how good your past teams were. It has absolutely nothing to do with how you as an individual perform. As such, your rating is purely based on luck and who you get matched up with and against during a queue. Getting some kind of adjusted rating, based on peer evaluation, should add a certain amount of individuality to someone's rating.

In practice I foresee some pretty amazing results. For example, if the whole team rates one person extremely poorly (for leaving the game for example). That individual takes all the negative rating hit and the 4 other players stay at the same rating as they were before the game started.

More specifically, I envision a winning team gaining a maximum of 50 points and a losing team losing a maximum of 50 points. How those points are distributed amongst the winning or losing team members depends on how their peers rate their performance. So, if one individual did very well, they could recieve +20 rating while the other 4 divide the remaining 30 points amongst one another.

I believe it would be a very simple feature to implement. All that would be needed is an added screen between the exit of the game and the entrance into chat. Teammates would be able to rate their 4 other teammates on a scale of 1 to 10, 5 being average, 10 being exceedingly good and 1 being completely negative participation.


Comment below rating threshold, click here to show it.

Kaeoz

This user has referred a friend to League of Legends, click for more information

Senior Member

08-21-2010

2 duo queue duos get matched up with an extra. The two duo queues are really tryhard. The 5th picks someone like teemo even though the other 4 beg not to. Teemo carries the game. All 4 rate him down because they thought he was an idiot for picking teemo and could've cost them the game. Now he gets 2 points when everyone else gets 12. Lovely right?


If you don't get the idea.... LOTS OF ROOM FOR ABUSE.


Comment below rating threshold, click here to show it.

Sparth

Junior Member

08-21-2010

Well a couple things:

-First

If someone is so insistent on doing something that it pisses off their entire team to the point of all 4 wanting to give an individual a 1 he probably should not be doing what he's doing. Why be stubborn to the point of picking Teemo if it's going to piss off your entire team. This is the point of this, to encourage people to take the opinions of their teammates into account.


-Second

At least it'll encourage the 2 duo queues to actually finish the game. Right now, I wouldn't put it past those duo queues to just leave or go AFK.


In conclusion:

Currently, the system as it is is full of abuse potential. So saying my suggestions has some potential for it is hardly an argument against it and pro status quo.

This system at the very least, should discourage people from acting stupid.


Comment below rating threshold, click here to show it.

Remunial

Senior Member

08-21-2010

I don't like the idea. Mainly because I almost always play some kind of tank or support champion. Some people actually understand the job of these types of champions, however, some people are still ignorant....mainly the people who only play carries. Sure, they just ended the game with the most kills. Does that mean they're good? Maybe, maybe not. It could mean they have skill, it could mean they just got lucky last hits or KS's during team fights, or it could mean that their teamates set up situations to make sure the carry gets the kill.

Say this carry went 13-4-2 during the game. Now, let's say I (as the tank) went 2-4-13 during the game. People might look at the carry and say "ZOMG SO MANY KILLS!! SO GOOD PLAYER!!" and give him the majority of the points for winning. Now, even if I initiated correctly, held enemies in place for that carry to finish, and even intentionally left low HP champions alone if I know they can't escape. My team may or may not care or even notice that i'm doing my job well. They will see my score compared to the carry and I will get less points for the game even though I contributed a lot to the win.


Comment below rating threshold, click here to show it.

Sabnitron

This user has referred a friend to League of Legends, click for more information

Senior Member

08-21-2010

I like this idea for two reasons:

A) It's an actual constructive post about the rating system, instead of OH GOD I LOST THREE GAMES RIOT FIX YOUR BROKEN GAME.

B) A peer review system that actually affected rating would encourage a more positive atmosphere. For instance, even if one person does decently, if they spend the entire game screaming at the rest of their team and generally being a jerk, they're likely to get a poor rating. The converse would likely also be true - someone does mediocre, but is friendly, helpful, and a team player, they are more likely to get a positive rating. And really, who would you rather play with? I'd take the second player over the first any day of the week.


Comment below rating threshold, click here to show it.

Sparth

Junior Member

08-21-2010

Quote:
Originally Posted by Remunial View Post
I don't like the idea. Mainly because I almost always play some kind of tank or support champion. Some people actually understand the job of these types of champions, however, some people are still ignorant....mainly the people who only play carries. Sure, they just ended the game with the most kills. Does that mean they're good? Maybe, maybe not. It could mean they have skill, it could mean they just got lucky last hits or KS's during team fights, or it could mean that their teamates set up situations to make sure the carry gets the kill.

Say this carry went 13-4-2 during the game. Now, let's say I (as the tank) went 2-4-13 during the game. People might look at the carry and say "ZOMG SO MANY KILLS!! SO GOOD PLAYER!!" and give him the majority of the points for winning. Now, even if I initiated correctly, held enemies in place for that carry to finish, and even intentionally left low HP champions alone if I know they can't escape. My team may or may not care or even notice that i'm doing my job well. They will see my score compared to the carry and I will get less points for the game even though I contributed a lot to the win.
I for one would not rate people based on their K/D ratio. This is the point of the peer evaluation. The first idea that had come to my mind would be a rating adjusted based on K/D. But I quickly dismissed it because of the reasons you highlighted.

The reasons for which a K/D ratio adjusted rating is bad are the same reasons a peer adjusted rating is good. Because at least here there is the potential of a tank type role of being very highly rated. This is why I like the idea I put forth. Further, if you feel like the carry you are helping is unappreciative, you can rate that person lower at the end of the game.

Sure the potential for what you describe is there. However, I think it would be far more likely that people rate individuals with positive attitude well and those with negative attitude poorly.


Comment below rating threshold, click here to show it.

Luderik

This user has referred a friend to League of Legends, click for more information

Senior Member

08-21-2010

Quote:
Originally Posted by Remunial View Post
I don't like the idea. Mainly because I almost always play some kind of tank or support champion. Some people actually understand the job of these types of champions, however, some people are still ignorant....mainly the people who only play carries. Sure, they just ended the game with the most kills. Does that mean they're good? Maybe, maybe not. It could mean they have skill, it could mean they just got lucky last hits or KS's during team fights, or it could mean that their teamates set up situations to make sure the carry gets the kill.

Say this carry went 13-4-2 during the game. Now, let's say I (as the tank) went 2-4-13 during the game. People might look at the carry and say "ZOMG SO MANY KILLS!! SO GOOD PLAYER!!" and give him the majority of the points for winning. Now, even if I initiated correctly, held enemies in place for that carry to finish, and even intentionally left low HP champions alone if I know they can't escape. My team may or may not care or even notice that i'm doing my job well. They will see my score compared to the carry and I will get less points for the game even though I contributed a lot to the win.
Well... the idea is to rate more accurately not only for 1 game, but for ALL games that this guy does. Ok, you play with 4 other teammates that think the carry ZOMG HES SO PRO. He gets higher rating, plays against better players and... If he's really good, he's gonna get more rating, if it was just luck, then he's gonna feed on the other games and he's gonna get dropped. A lot.

I think it's the more accurate way to decide. I mean, who in the winning team deserve most credit, who in the loosing team failed his teammates.

From my point of view, rating is the best way to match people with equivalent skills qith each others. If a guy goes 0-7-2 with yi after 5 minutes, then maybe he needs to get a HUGE drop to play with people at his skill level. If a player goes 3-0-2 after 15 minutes, but gets owned because his teammates fed the enemy carry, then maybe this guy deserves a lesser drop. Not because he has a better kdr, but because he could not show his potential, since when he entered the mid game, it was clearly unfair due to feeding.

Maybe teams should rate the other team too. This gives an objective point of view. Maybe the 0-5-10 shen was very good, but when he ulted his teammates ran away and he got killed in 1v3 instead of 3v3. His teammates are like "stupid shen suiciding". But you think "stupid, you ran away when shen ults... poor him"

I find this idea to be the best system ever. It could stop :

-OMG YOU ****ING NOOB PLEASE UNINSTALL RIGHT NAO
-Ksing... because if one guy keeps ksing a team he's gonna get downvoted
- /all I'M THE ONLY GOOD PLAYER ON THIS TEAM!, YI IS A ****ING IDIOT
-People could just stop always playing carries...
-Don't worry guys, I got this *locks in corki and feed*
-on the forum : QQ I LOST 200 RATING ****ING NOOBS MATCHMAKING BROKEN GET RIDS OF LESS THAN 8k WINS PLEASE

-People would stop complaining about matchmaking and about how people suck at their rating. If you get downvoted, then you gotta ask yourself some serious questions and stop blaming others. If you have a 5 loss streak, but you loose like 30 ELO, what's the big deal. If you are the amazing player you think you are, one win should get back those 5 losses.

-In every game, 5 people win and 5 people loose. End. If you can spread the amout of ELO gain/loss by rating people from how good they played, then your 50% statistical w/l ratio should get you high ELO in no time. If you suck, then you can have 50 wins 30 loss, but still be under 1200.

And finally...


How

many

games

do

you

have

a pro nidalee, a pro twitch, a pro tristana and a pro AP ezreal saying "We need tank" at you, the 5th pick.

EVERYBODY THINKS "in solo queue, you have to be able to carry your team."

WRONG!

If everybody carries, nobody can tank, support, etc. seriously...

Playing a good tank some games could help you not drop if your carries sucks. They will get downvoted...

In short :
- Possible to climb up without always playing a carry
- Stop trash talking
- Not after one single game, but after many games, it would rate more accurately.


Comment below rating threshold, click here to show it.

Sparth

Junior Member

08-21-2010

Quote:
Originally Posted by Luderik View Post
Well... the idea is to rate more accurately not only for 1 game, but for ALL games that this guy does. Ok, you play with 4 other teammates that think the carry ZOMG HES SO PRO. He gets higher rating, plays against better players and... If he's really good, he's gonna get more rating, if it was just luck, then he's gonna feed on the other games and he's gonna get dropped. A lot.

I think it's the more accurate way to decide. I mean, who in the winning team deserve most credit, who in the loosing team failed his teammates.

From my point of view, rating is the best way to match people with equivalent skills qith each others. If a guy goes 0-7-2 with yi after 5 minutes, then maybe he needs to get a HUGE drop to play with people at his skill level. If a player goes 3-0-2 after 15 minutes, but gets owned because his teammates fed the enemy carry, then maybe this guy deserves a lesser drop. Not because he has a better kdr, but because he could not show his potential, since when he entered the mid game, it was clearly unfair due to feeding.

Maybe teams should rate the other team too. This gives an objective point of view. Maybe the 0-5-10 shen was very good, but when he ulted his teammates ran away and he got killed in 1v3 instead of 3v3. His teammates are like "stupid shen suiciding". But you think "stupid, you ran away when shen ults... poor him"

I find this idea to be the best system ever. It could stop :

-OMG YOU ****ING NOOB PLEASE UNINSTALL RIGHT NAO
-Ksing... because if one guy keeps ksing a team he's gonna get downvoted
- /all I'M THE ONLY GOOD PLAYER ON THIS TEAM!, YI IS A ****ING IDIOT
-People could just stop always playing carries...
-Don't worry guys, I got this *locks in corki and feed*
-on the forum : QQ I LOST 200 RATING ****ING NOOBS MATCHMAKING BROKEN GET RIDS OF LESS THAN 8k WINS PLEASE

-People would stop complaining about matchmaking and about how people suck at their rating. If you get downvoted, then you gotta ask yourself some serious questions and stop blaming others. If you have a 5 loss streak, but you loose like 30 ELO, what's the big deal. If you are the amazing player you think you are, one win should get back those 5 losses.

-In every game, 5 people win and 5 people loose. End. If you can spread the amout of ELO gain/loss by rating people from how good they played, then your 50% statistical w/l ratio should get you high ELO in no time. If you suck, then you can have 50 wins 30 loss, but still be under 1200.

And finally...


How

many

games

do

you

have

a pro nidalee, a pro twitch, a pro tristana and a pro AP ezreal saying "We need tank" at you, the 5th pick.

EVERYBODY THINKS "in solo queue, you have to be able to carry your team."

WRONG!

If everybody carries, nobody can tank, support, etc. seriously...

Playing a good tank some games could help you not drop if your carries sucks. They will get downvoted...

In short :
- Possible to climb up without always playing a carry
- Stop trash talking
- Not after one single game, but after many games, it would rate more accurately.
Thank you for the support.

I also like your idea of letting players rate people on the opposite team.


Comment below rating threshold, click here to show it.

Suijinheki

Senior Member

08-21-2010

Sounds like a good idea although i do worry somewhat about trolls <.< and the like; they're everywhere i swear


Comment below rating threshold, click here to show it.

Luderik

This user has referred a friend to League of Legends, click for more information

Senior Member

08-21-2010

Yeah in fact, I had the idea too but you proposed it first. Lots of advantages in that system. I can't see any fails (like I said, not for one game only but for many games).

For premades... lets say they all get the same, the average of the 2 for solo queue. And they can't vote each others in solo.


12