Possible solutions to Ranked Solo

Comment below rating threshold, click here to show it.

PogoPogoPogoPogo

Senior Member

01-23-2013

Quote:
Originally Posted by Smushynumnum View Post
Can you please reply to my questions, you seem to be dodging a bunch of them - reading post now.
Yea, no. I'm not replying to your questions. You're not even reading my posts before you reply to them, as you've just admitted here.


Comment below rating threshold, click here to show it.

Smushynumnum

Member

01-23-2013

Sorry Pogo if I offended you from quoting the ranking system from Halo 3, it was for matter of comparison. You seem to be dodging a bunch of my questions, very unfortunate. If you'd like to continue to the discussion feel free to let me know, I'm interested on the input


Comment below rating threshold, click here to show it.

adc

Member

01-23-2013

Quote:
Originally Posted by Smushynumnum View Post
Regarding your post - "And if my strategy isn't a superior strategy for winning, then over the course of those 100 games, I'll win fewer games than my teammates that consistently go 5/2/0 (to my 0/10/2 consistently). The person with the best strategy for destroying the enemy Nexus will have the most Elo after a large sample size of games, regardless of their KDR."

Typically what I've seen with support that perform optimally, is that their K/D ratio is not typically as high, however their assists certainly are. I think someone who performs 3/3/5 with 0 turret kills, 50 minions, should not be rewarded nearly as much as someone who is 0/3/26, with 2 turret kills, 100 minion kills. Would you not agree that the 2nd player influence the win of the match much more so than the first?
Yes, they did influence the win more. And if they consistently influence the win more, they're going to win more often than the player who does not influence it quite as much - thus causing their Elo to rise, while the other player plateaus.

I don't think you understand how a matchmaking system works. It's not supposed to rank you accurately within 5 games. You have to play hundreds of games for it to accurately rank you. And if you're a better player than where you are currently at, you'll rise because you consistently do well. Similarly, if you're a worse player, you'll drop because you consistently do badly.

We shouldn't be rewarding players for getting a good KDR, or turret kills, or whatever. I can push 10 turrets, but if I let minions finish every single one, I get less Elo? The things that cause wins are intangibles. I played a game where I went 9/3/7. I struggled the entire game, and just generally had a really bad game. But my team made up for it. Even though the best score was 5/5, and the others were like 2/6 and 4/5, they carried me, not the other way around. 5/5 Lee caught people all day, he stole baron, etc. etc. I just autoattacked. Yet you seem to think I should gain more Elo than them, despite me not being the reason we won.


Comment below rating threshold, click here to show it.

Smushynumnum

Member

01-23-2013

I read your posts clear as day. You think destroying the nexus is the only thing that matters. I have to disagree on one very important point - I d o not think all members of a winning or losing team should be rewarded equally in terms of elo should a great number of variables influence the outcome of the game. Namely - afking, K/D/Assists, tower kills, feeding, negative attitude, etc.

I don't see how you can argue that - should everything be exactly the same - all members of one team are 5/5/5 - 100 minion kills precisely the same, and one player's attitude is poor, do you not agree that the other players should be rewarded more so?


Comment below rating threshold, click here to show it.

Smushynumnum

Member

01-23-2013

Quote:
Originally Posted by AD Bottom View Post
Yes, they did influence the win more. And if they consistently influence the win more, they're going to win more often than the player who does not influence it quite as much - thus causing their Elo to rise, while the other player plateaus.

I don't think you understand how a matchmaking system works. It's not supposed to rank you accurately within 5 games. You have to play hundreds of games for it to accurately rank you. And if you're a better player than where you are currently at, you'll rise because you consistently do well. Similarly, if you're a worse player, you'll drop because you consistently do badly.

We shouldn't be rewarding players for getting a good KDR, or turret kills, or whatever. I can push 10 turrets, but if I let minions finish every single one, I get less Elo? The things that cause wins are intangibles. I played a game where I went 9/3/7. I struggled the entire game, and just generally had a really bad game. But my team made up for it. Even though the best score was 5/5, and the others were like 2/6 and 4/5, they carried me, not the other way around. 5/5 Lee caught people all day, he stole baron, etc. etc. I just autoattacked. Yet you seem to think I should gain more Elo than them, despite me not being the reason we won.


Agreed about the consistency - and influence of future elo. What about duo queue though, do you not agree that this is an unfair advantage for solo-only players?

As for long term playing, why not have the option to have your elo be truly effected so that your skills are reflected faster than simply the outcome of an individual game. How is considered "fair" when one performs far by none more optimally compared to their teammates, yet is only rewarded the same? Similarly, why should one be penalized for their teammates afk? Should some of the elo in their loss not be re-cooped?


Comment below rating threshold, click here to show it.

PogoPogoPogoPogo

Senior Member

01-23-2013

Quote:
Originally Posted by Smushynumnum View Post
Typically what I've seen with support that perform optimally, is that their K/D ratio is not typically as high, however their assists certainly are. I think someone who performs 3/3/5 with 0 turret kills, 50 minions, should not be rewarded nearly as much as someone who is 0/3/26, with 2 turret kills, 100 minion kills. Would you not agree that the 2nd player influence the win of the match much more so than the first?
But what if I'm 0/10/0, with 8 turret kills, 2 inhibitor kills, and I'm the only champion on my team that damaged the Nexus?



But ultimately, you want to get into this extraordinarily complex game of assigning value to every possible thing that shows up on the end game scoreboard and give a fraction of an Elo point, positive or negative to it.


How about things that don't show up on the scoreboard? Things like Nunu's bloodboil? If we're looking at anything besides simply win or lose, we have to account for absolutely EVERYTHING.

Do you want a team of 10+ statisticians and LoL experts to spend 6 months analyzing how to appropriate award Elo for Nunu's bloodboil? They'd have to determine the absolute best possible use for it at all times, the worst possible use for it at all times. They have to determine things like when it was good for Nunu to use it versus when it was just a waste of mana. They have to determine whether Nunu is bloodboiling the right target.

And then, once they've perfectly analyzed absolutely every single possible scenario, they've got to get a big team of coders and programmers and go in and spend a massive chunk of time programming and debugging the code that assigns Nunu Elo for his ability to use Bloodboil.


And then that same process has to be repeated for the other 439 abilities currently in the game, not counting passives.


And then we've got to go through all the same stuff for things like item purchases. We have to go through all the same stuff for things like the order I leveled my skills. We have to go through all the same stuff for ward placement, potion usage. We have to go through the same stuff for decisions like deciding between staying in lane and soaking xp, or pushing lane and hitting turret, or pushing lane and roaming to get a gank.

And there's tons and tons and tons and tons and tons of other little micro-decisions players make at every instant of the game that has to be considered.



And then we have to hope that not only are all the micro-decisions accounted for, but that Riot got people who knew what they were doing, because otherwise if I'm Voyboy #1 top NA, and I'm winning 65% of my games because I'm the best top NA, but I'm winning less Elo then other tops per win, because even though my strategy and decision making is superior to other tops, it doesn't perfectly match what's programmed into the game.




AND THEN, 5 years later once that's all programmed in and perfected, we have to hope that no champions were redesigned in that time period (or all the analysis on that champion has to be redone). We have to hope that no new champion is released (because we have to fully analyze that champion). We have to hope there are no changes at all, nerf or buff to any champions. We have to hope there are no changes to anything on the map (creep health, jungle monster health, damage, gold the monsters/minions give, turrets or how they work, turret range, etc. etc. etc.) and we have to hope there are no item changes, etc., etc., etc.




Do you see how unfeasible this is? Look, a simple W/L system isn't absolutely perfect, but it's going to be a lot more accurate than anything short of an essentially PERFECT system that accounts for absolutely everything.


Comment below rating threshold, click here to show it.

PogoPogoPogoPogo

Senior Member

01-23-2013

Quote:
Originally Posted by Smushynumnum View Post
Similarly, why should one be penalized for their teammates afk? Should some of the elo in their loss not be re-cooped?
It is re-cooped. When you play the 99 other matches out of a 100 game set, and because you're not an AFKer, you get fewer AFKers on your team than the AFKer.

AFKing is clearly not a winning strategy, so if the AFKer continues that strategy over the course of 100 matches, while you continue a very strict strategy of not-AFKing, you'll win more games than the AFKer and have more Elo overall.


Comment below rating threshold, click here to show it.

Smushynumnum

Member

01-23-2013

Quote:
Originally Posted by PogoPogoPogoPogo View Post

But ultimately, you want to get into this extraordinarily complex game of assigning value to every possible thing that shows up on the end game scoreboard and give a fraction of an Elo point, positive or negative to it.

Nope, not necessarily. I am simply pointing out possible solutions, it wouldn't have to be everything. Too many different points going on at once, I apologize. These can all be taken with a grain of salt. Ones that I think should be undeniable however, are penalization for afking, or feeding, in terms of elo.


Comment below rating threshold, click here to show it.

captainhammhamm

Senior Member

01-23-2013

what about the supports? whose job it is NOT to get kills and if they die, its not as big a deal as if, say, the ADC dies. What you suggest is flawed because it penalizes the support even if your team wins.


Comment below rating threshold, click here to show it.

adc

Member

01-23-2013

Quote:
Originally Posted by Smushynumnum View Post
Agreed about the consistency - and influence of future elo. What about duo queue though, do you not agree that this is an unfair advantage for solo-only players?

As for long term playing, why not have the option to have your elo be truly effected so that your skills are reflected faster than simply the outcome of an individual game. How is considered "fair" when one performs far by none more optimally compared to their teammates, yet is only rewarded the same? Similarly, why should one be penalized for their teammates afk? Should some of the elo in their loss not be re-cooped?
How does duo queueing affect anything? Yes, it brings a higher skilled player down to a lower level, but that's not a guarantee they'll do well. It also brings a lower skilled player up to a higher level (though that's not a guarantee they'll do badly).