Attention Riot, Solo Q needs to be ranked based upon individual performance.

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

Inciter

Member

10-07-2012

Attention Riot, Solo Q needs to be ranked based upon individual performance while within a Solo Q ranked match.

To start off: there are constant debates about whether or not ELO hell exists, whether or not that is the case, the simple fact that people are discussing the topic as an issue is an indication that there are issues with Solo Q, end of story.
Effectively, every stat viewable in the Post Game Stats panel can be used to evaluate the performance of each individual player on a per game basis. These results and stats should be the criteria that players are evaluated on during Solo Q, not simply whether or not the team wins as a whole. These variables are the EXACT reason why it’s a poor choice to evaluate a player’s ELO off of only win/loss ratios. There are so many factors that contribute to whether or not a player actually performed well, these are the variables that need to be tracked and calculated for a Player ELO Score. Examples:

They could use a grading system based upon a number of variables.
- Kills: Having a higher number of kills indicates efficiency with a particular character.
- Deaths: Low death count indicates good map awareness and knowing when to engage versus not in a particular scenario.
- Assists: Indicates awareness of the importance of helping out your fellow teammates.
- Towers: This may not be the best variable to grade on because it’s based upon whoever gets the final hit, but if you could be graded based upon proximity not necessarily damage done. That way characters like supports can still get credit for participating in a team fight under tower that results in the objective being successfully destroyed. If it is too difficult to grade this variable fairly, or implement code that would do this, then don’t base the rating on this variable.
- Objectives Kills: the number of Dragon/Baron kills you participated in that were attained. This show’s overall objective awareness and team awareness.
- Creep Score: This is an indication of a player’s ability to prioritize Creeps and increase Gold and XP which in turn results in a more fed player that can then hopefully provide more benefit to the team.
- Time Spent Dead: This is a hugely informational variable. If you spend a lot of time dead but have few deaths, it means that you made big mistakes towards mid/late game resulting you being dead during higher levels like level 11+. This means as a player, because you died, your team can’t defend or take objectives as securely as before. But to grade on time spent dead alone, without taking into account deaths and at which point they occurred would be bad. For instance, if a player fed in the beginning but then never died again, they still did poorly because that then fed player rolled over your team. So the way they can evaluate this and keep it balanced is by taking into account the overall duration of the match itself. As the match goes on in duration, it will become more and more acceptable to have a longer Time Spent Dead duration. It’s not an excuse to die of course for no reason, but you understand.

Overall the primary issue with Solo Q is that League of Legends is a Team game, and therefore you are rated as a Team. Which is fine and makes sense if you are playing in a premade team scenario like Ranked 5’s? However in Solo Q you have a random assortment of people slapped together to form a team based upon current ELO for that game mode type.
Unfortunately many players experience a large number of leavers, trolls, and simply players that don’t understand all the dynamics of the game nor belong in ranked all together. Issues start arising even in Champion select because players are playing Solo Q ranked and they don’t even have a large enough Champion selection base to pick roles in the appropriate order: Support/Carry > Jungler > Mid/ Top. Not only that, but people enter Solo Q without the ability to play all the roles with considerable skill, so if they don’t get the exact role they know how to play, it can affect the team. For instance someone joins Solo Q and can only play Mid, and only Ryze, so they think they can play ranked, and instead, bring down the team because of their lack of capabilities, but as a whole, the team is ranked on that performance and graded down as a whole despite that perhaps one person completely failing in a lane feeds to the point that the team loses.

There are counter arguments to this, mid should have warded, Jungler should have ganked more etc etc, overall the issue remains the same and the forums still run rampant. Solo Q needs to be graded on individual performance. Maybe someone didn’t feed at all, maybe they had connection issues with their internet or they have a poor computer: but if that’s the case they shouldn’t be playing ranked. But Riot can’t force people to not play ranked and impose restrictions because that would cut into the competitive scene and draw bad press by restricting players and that could just end in a slippery slope of frustration. If Riot did impose these restriction based upon computer performance or ping rating or other things, that would cut into their Revenue and their ability to continue to support this AMAZING game, so imposing restriction on SOLO Q is just a bad idea, end of story. So that brings us back to the rating system as the cure to this issue.
Some players do deserve to have the ELO lowered because of poor plays, it’s only fair and you can’t have a competitive system without some sort of rating implementation. But what do you do to the player that was supportive, didn’t feed, were cooperative, won their lane etc., oh wait, the team still lost, lower their ELO…..WHAT?!

What Riot needs to do is convert the current ELO system for Solo Q and base it on individual performance. They already have many tracking systems currently implemented into the game. They have Post Game Stats, LolReplay, and track a large number of stats like Creep Score, Deaths Kills Assists, Time Spent Dead etc. These stats can be used as Variables on a scale system to indicate how well a player performed in a match. They already have a system similar to this implemented in Dominon resulting in a player score, although Dominion’s rating system also has a number of issues so let’s not go there . Overall I’m stating that Riot already has the information at their disposal to improve their rating system, so why haven’t they done it.?


Comment below rating threshold, click here to show it.

I C Y U N V this

Member

10-07-2012

This wouldn't help to be honest. When people need to group, one person is going to try to split push. When the inhibitor tower is getting pushed 5v3, 2 people will prioritize a dragon or a tier 1 tower.

The law of averages do play out pretty well in the long run. I complain a lot about games where i'm 4v5 or 3v5 or have a 0-15 players who doesn't know anything. But the enemy team has that too at times. The biggest thing in raising elo is communicating objectives, commanding teams, and not giving up if you're good at the formers. For the 10 impossible to win must surrender at 20 games, chances are at least 2-3 are winnable. If you're winning your lane, help other lanes. Or tell the jungler to help other lanes and take your tower soon to join in. Help secure dragons if you're far ahead too.

I'm not saying you're one, but the people who always brag about individual performances are usually the ones who want to tell his team that they suck and lower their morale. He's usually the one to say not to comment on what they should do in team fights or even if they should be doing teamfights. He's usually the one who doesn't keep timers. He's usually the one that lives in hindsight and call people stupid for initiating when they shouldn't be. But trust me, it never hurts to assume everyone is stupid and point out the obvious that you think everyone should know. If a player is really good individually, the only thing he needs to learn is to talk to his team and tell his team what to do if they're lost. That's how those people will raise elo.

And like i said. Yes you get 5v3 games where you can't possibly win. But you win your fair share of those games in the long run too.


Comment below rating threshold, click here to show it.

Zzenkaido

Member

10-07-2012

So what you're saying is that I should start practicing darius if this goes through?


Comment below rating threshold, click here to show it.

Epjest

Senior Member

10-07-2012

Too long didn't read.

Your premise was too poor to begin conjecture.


Comment below rating threshold, click here to show it.

perry mason1

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

Senior Member

10-07-2012

Yes, let's promote kill stealing and poor teamwork in favor of "individual" elo.


Comment below rating threshold, click here to show it.

IsOnFirstBase

Senior Member

10-07-2012

I didn't even read the post and I agree.

The same way I think I should win every hand of poker I ever get dealt. I'm the best poker player in the world. Only reason I ever lose is because the dealer feeds the other players better cards than me.







Hope someone can catch that ball.....


Comment below rating threshold, click here to show it.

Chilipapa

Senior Member

10-07-2012

Quote:
Originally Posted by Inciter View Post
Attention Riot, Solo Q needs to be ranked based upon individual performance while within a Solo Q ranked match.

To start off: there are constant debates about whether or not ELO hell exists, whether or not that is the case, the simple fact that people are discussing the topic as an issue is an indication that there are issues with Solo Q, end of story.
Effectively, every stat viewable in the Post Game Stats panel can be used to evaluate the performance of each individual player on a per game basis. These results and stats should be the criteria that players are evaluated on during Solo Q, not simply whether or not the team wins as a whole. These variables are the EXACT reason why itís a poor choice to evaluate a playerís ELO off of only win/loss ratios. There are so many factors that contribute to whether or not a player actually performed well, these are the variables that need to be tracked and calculated for a Player ELO Score. Examples:

They could use a grading system based upon a number of variables.
- Kills: Having a higher number of kills indicates efficiency with a particular character.
- Deaths: Low death count indicates good map awareness and knowing when to engage versus not in a particular scenario.
- Assists: Indicates awareness of the importance of helping out your fellow teammates.
- Towers: This may not be the best variable to grade on because itís based upon whoever gets the final hit, but if you could be graded based upon proximity not necessarily damage done. That way characters like supports can still get credit for participating in a team fight under tower that results in the objective being successfully destroyed. If it is too difficult to grade this variable fairly, or implement code that would do this, then donít base the rating on this variable.
- Objectives Kills: the number of Dragon/Baron kills you participated in that were attained. This showís overall objective awareness and team awareness.
- Creep Score: This is an indication of a playerís ability to prioritize Creeps and increase Gold and XP which in turn results in a more fed player that can then hopefully provide more benefit to the team.
- Time Spent Dead: This is a hugely informational variable. If you spend a lot of time dead but have few deaths, it means that you made big mistakes towards mid/late game resulting you being dead during higher levels like level 11+. This means as a player, because you died, your team canít defend or take objectives as securely as before. But to grade on time spent dead alone, without taking into account deaths and at which point they occurred would be bad. For instance, if a player fed in the beginning but then never died again, they still did poorly because that then fed player rolled over your team. So the way they can evaluate this and keep it balanced is by taking into account the overall duration of the match itself. As the match goes on in duration, it will become more and more acceptable to have a longer Time Spent Dead duration. Itís not an excuse to die of course for no reason, but you understand.

Overall the primary issue with Solo Q is that League of Legends is a Team game, and therefore you are rated as a Team. Which is fine and makes sense if you are playing in a premade team scenario like Ranked 5ís? However in Solo Q you have a random assortment of people slapped together to form a team based upon current ELO for that game mode type.
Unfortunately many players experience a large number of leavers, trolls, and simply players that donít understand all the dynamics of the game nor belong in ranked all together. Issues start arising even in Champion select because players are playing Solo Q ranked and they donít even have a large enough Champion selection base to pick roles in the appropriate order: Support/Carry > Jungler > Mid/ Top. Not only that, but people enter Solo Q without the ability to play all the roles with considerable skill, so if they donít get the exact role they know how to play, it can affect the team. For instance someone joins Solo Q and can only play Mid, and only Ryze, so they think they can play ranked, and instead, bring down the team because of their lack of capabilities, but as a whole, the team is ranked on that performance and graded down as a whole despite that perhaps one person completely failing in a lane feeds to the point that the team loses.

There are counter arguments to this, mid should have warded, Jungler should have ganked more etc etc, overall the issue remains the same and the forums still run rampant. Solo Q needs to be graded on individual performance. Maybe someone didnít feed at all, maybe they had connection issues with their internet or they have a poor computer: but if thatís the case they shouldnít be playing ranked. But Riot canít force people to not play ranked and impose restrictions because that would cut into the competitive scene and draw bad press by restricting players and that could just end in a slippery slope of frustration. If Riot did impose these restriction based upon computer performance or ping rating or other things, that would cut into their Revenue and their ability to continue to support this AMAZING game, so imposing restriction on SOLO Q is just a bad idea, end of story. So that brings us back to the rating system as the cure to this issue.
Some players do deserve to have the ELO lowered because of poor plays, itís only fair and you canít have a competitive system without some sort of rating implementation. But what do you do to the player that was supportive, didnít feed, were cooperative, won their lane etc., oh wait, the team still lost, lower their ELOÖ..WHAT?!

What Riot needs to do is convert the current ELO system for Solo Q and base it on individual performance. They already have many tracking systems currently implemented into the game. They have Post Game Stats, LolReplay, and track a large number of stats like Creep Score, Deaths Kills Assists, Time Spent Dead etc. These stats can be used as Variables on a scale system to indicate how well a player performed in a match. They already have a system similar to this implemented in Dominon resulting in a player score, although Dominionís rating system also has a number of issues so letís not go there . Overall Iím stating that Riot already has the information at their disposal to improve their rating system, so why havenít they done it.?
And killstealing becomes a reality, support role goes away for good, jungler decides he'd rather be a split pusher(well, everyone does do they can get tower kills), and the game devolves into petty bickering.

Sounds legit.


Comment below rating threshold, click here to show it.

Zen Fury

Senior Member

10-07-2012

There is no way to do this in a fair fashion. There are way too many factors to consider and too many ways the metrics can get heavily skewed or be misinterpreted entirely.

How do you handle judging jungle vs. support vs. mid? What if the team decides to dual jungle or try some other random strategy? There is no simple model for judging the characteristics of play to generate elo fairly.

Or to put a finer point on it, what if you were trying to assign elo to individual members of a sports team. Is there any objective way to compare a QB to a linebacker? A nose tackle to an offensive guard? What about when they decide to do a crazy play where they hand off to a lineman to push the ball that one extra yard? How does an elo model handle that?

Your proposal doesn't really make any sense, and would turn a team game into an "every-man-for-himself" game. If that is what you want, there are plenty of FPS games out there for you to chose from.


Comment below rating threshold, click here to show it.

Fosterisgod

Senior Member

10-07-2012

That must suck to waste a bunch of time writing an essay just to get downvoted. QQ


Comment below rating threshold, click here to show it.

jakab

Senior Member

10-07-2012

What a terrible idea. How about you learn to play with your team instead?


12