Lp variance from performance

Th3 Rift Hero··1 votes·1,674 views

Looking at what things can possibly help winning a game, it may be possible to improve the ranked mode with a system that could reward players doing well no matter what teammates they have. For this we would need some metrics of skills.

  • K/D/A
  • Kill contribution
  • Damage to objectives (except for jungle since it have more easy access to objectives)
  • Creep score (except for support and jungle role since they dont have lane minions to farm)

We could tune the lp with these stats by doing an average of the lp that would be won or lost in each category:

  • K/DA, Kill contribution: We could use a ladder for the team...

On win -

1st : + 26 2nd : + 23 3rd: + 20 4th: + 17 5th: + 14

On lose - 1st : - 14 2nd: - 17 3rd: - 20 4th : - 23 5th : - 26

-Damage to objectives : We could use a ladder for the team but on 4, since it is not calculated for jg...

On win -

1st : + 26 2nd : + 23 3rd: + 17 4th: +14

On lose -

1st : - 14 2nd : - 17 3rd: MMR - 23 4th: MMR - 26

  • Creep score : We could also use a ladder for the team but on 3 ranks, since its not calculated for the support and jg...

On win -

1st : MMR + 24 2nd : MMR + 20 3rd: MMR + 16

On lose - 1st : MMR - 16 2nd: MMR - 20 3rd: MMR - 24

So, let's look at a game tested on both sides for an example

https://imgur.com/a/ydcmDln

Winning side :

K/D/A

Yorick : Perfect (1) +26 Fizz : 4 (3) +20 Ekko : 6 (2) +23 Yasuo : 6 (2) +23 Elise : 3 (4) +17

Kill Contribution

Yorick : 25% (4) +17 Fizz : 40% (2) +23 Ekko : 60% (1) +26 Yasuo : 30% (3) +20 Elise : 30% (3) +20

Damage to objectives

Yorick : 10616 (1) +26 Ekko : 3860 (3) +17 Yasuo : 2900 (4) +14 Elise : 4832 (2) +23

Creep score

Yorick : 86 (3) +16 Ekko : 117 (2) +20 Yasuo : 127 (1) +24

Overall Yorick : +21 Fizz : +22 Ekko : +22 Yasuo : +20 Elise : +20

Losing side

K/D/A

Irelia : 1.33 (2) -17 Shaco : 2 (1) -14 Katarina : 0 (5) -26 Miss Fortune : 0.50 (4) -23 Sona : 1.20 (3) -20

Kill contribution

Irelia : 57% (2) -17 Shaco : 29% (3) -20 Katarina : 0% (4) -23 Miss Fortune : 29% (3) -20 Sona : 86% (1) -14

Damage to objectives

Irelia : 3520 (1) -14 Katarina : 0 (4) -26 Miss Fortune : 1060 (3) -23 Sona : 1586 (2) -17

Creep score

Irelia : 133 (1) -16 Katarina : 68 (3) -24 Miss Fortune : 101 (2) -20

Overall

Irelia: -16 Shaco: -17 Katarina : -25 Miss fortune : -22 Sona : -17

In order to reward consistency, it may also be possible to add 1 lp a win/remove 1 lp a lose on win streaks and add 1 lp a lose/remove 1 lp a win on lose streak.

15 Comments

DrCyanide2 votes

Stat padding becomes a huge problem with such a system.

Take your screenshot for example. It's in Kat's best interest to ignore the attackers in the base (since they're already going to lose) and try to out CS Shaco, since that will mean she loses less LP. Ekko and Yasuo have incentive to set Yorick up to die so that they can claim the top KDA spot, and Yorick doesn't have an incentive to go anywhere near the enemy team since he can keep his KDA and get more CS farming in a side lane.

In short, the most effective way to climb the ladder would become not about who's best at the game or who wins the most, but who's best at betraying their allies. The idea looks good on paper, but just falls flat in execution.

Th3 Rift Hero1 votes

{quoted}

Stat padding becomes a huge problem with such a system.

Take your screenshot for example. It's in Kat's best interest to ignore the attackers in the base (since they're already going to lose) and try to out CS Shaco, since that will mean she loses less LP. Ekko and Yasuo have incentive to set Yorick up to die so that they can claim the top KDA spot, and Yorick doesn't have an incentive to go anywhere near the enemy team since he can keep his KDA and get more CS farming in a side lane.

In short, the most effective way to climb the ladder would become not about who's best at the game or who wins the most, but who's best at betraying their allies. The idea looks good on paper, but just falls flat in execution.

tbh if they dont contribute to fight and dont take objectives they will get penalized so they wont get as much lp or lose that few lp

DrCyanide2 votes

Then you'll have a split pushing Ezreal Draven Ashe Jinx Soraka Karthus that's attacking a Tier 1 tower as the nexus dies and getting one last Ult off before the fight ends, because they can get the objective while everyone else just instantly dies.

No matter how you weigh the stats they'll fall short, because often times the right play isn't the one that contributes the most to your stat line.

Th3 Rift Hero1 votes

{quoted}

Then you'll have a split pushing Ezreal Draven Ashe Jinx Soraka Karthus that's attacking a Tier 1 tower as the nexus dies and getting one last Ult off before the fight ends, because they can get the objective while everyone else just instantly dies.

No matter how you weigh the stats they'll fall short, because often times the right play isn't the one that contributes the most to your stat line.

with this you assume that how lp is calculated would be explained by riot and also why do you bother talking in this thread when you didnt play rank since s7 where ranked was still good?

LTK KoRo1 votes

He made some good and well-thought points, and you're attacking him because you ran out of arguments XD

Great job OP.

Th3 Rift Hero1 votes

[{quoted}](name=LTK KoRo,realm=EUW,application-id=3ErqAdtq,discussion-id=uVTFzQVN,comment-id=00010000000000000000,timestamp=2019-01-02T20:07:45.605+0000)

He made some good and well-thought points, and you're attacking him because you ran out of arguments XD

Great job OP.

not really attacking, but i looked his account and there is no ranked history in season 8, but yeah if how is it calculated isnt said by riot there would be no real problem

DrCyanide1 votes

Sorry, I missed the part about playing ranked myself earlier.

I got to Gold in Solo, Flex 5, and Flex 3 in 2017, hitting my goal for the season, but then was burnt out. IRL stuff was keeping me busy anyway, so I took a break for a bit, leaving the boards/patch notes/everything, tried playing the game again, but just wasn't able to get back into it. Part of it was how much had changed in the time I'd been gone and not keeping up with it, but part of it was the game just felt worse to play. So I haven't forced myself to play, but I've kept up with the Boards/patch notes, looking for changes that look like they'll improve the game and make it more enjoyable again.

Regardless, this is hardly a new suggestion for a new problem. It's been suggested at least since the Mastery system went live, and will continue to be suggested as long as people think they're being matched with teammates who suck while they're awesome (it's never suggested when someone gets carried in a game themselves).

Th3 Rift Hero1 votes

{quoted}

Sorry, I missed the part about playing ranked myself earlier.

I got to Gold in Solo, Flex 5, and Flex 3 in 2017, hitting my goal for the season, but then was burnt out. IRL stuff was keeping me busy anyway, so I took a break for a bit, leaving the boards/patch notes/everything, tried playing the game again, but just wasn't able to get back into it. Part of it was how much had changed in the time I'd been gone and not keeping up with it, but part of it was the game just felt worse to play. So I haven't forced myself to play, but I've kept up with the Boards/patch notes, looking for changes that look like they'll improve the game and make it more enjoyable again.

Regardless, this is hardly a new suggestion for a new problem. It's been suggested at least since the Mastery system went live, and will continue to be suggested as long as people think they're being matched with teammates who suck while they're awesome (it's never suggested when someone gets carried in a game themselves).

tbh with this system you could get carried but you will win less lp if you really carried you will win more, but yea since season 8 start, youre almost at the total mercy of your team, i've never seen such streaks and disgusting game outcomes in season 7 tbh thats why ranked needs a real improvement if riot don't want to fix the mess they did in season 8

ZephyrDrake1 votes

{quoted}

not really attacking, but i looked his account and there is no ranked history in season 8, but yeah if how is it calculated isnt said by riot there would be no real problem

Why would there be no real problem? You really think that people won't realize or catch on that the better their stats the more LP they win? Like how stupid do you think people are? People data mine everything, they test all sort of things. You really are underestimating how people look up information especially in this day and age

Th3 Rift Hero1 votes

{quoted}

Why would there be no real problem? You really think that people won't realize or catch on that the better their stats the more LP they win? Like how stupid do you think people are? People data mine everything, they test all sort of things. You really are underestimating how people look up information especially in this day and age

tbh i would rather have some people trying to get good stats than getting unfair lp gains and loses, and anyway, people having good stats help to win the game

DrCyanide1 votes

It doesn't matter if Riot explains how they calculate it or not. As soon as someone notices that it's not based off of the MMR (which enough people track) they'll start trying to figure out what it is based on. It's not like it would take a genius to notice "hey, on the games I get an S+ on I get a ton of LP, and games I get a D on I get hardly any LP, but it didn't use to do that," and thus the connection would be drawn to the game stats.

They could have no clue what the real stats are, but they'd still try and manipulate it. Kill Steals would be more prevalent (after all, they don't know they're treated the same as assists). Players would AFK farm losing games, and some would buy and place 20 wards in the fountain to try and increase their ward score. Some may try to tank turret shots in a futile split push to increase their damage taken numbers.

Th3 Rift Hero1 votes

{quoted}

It doesn't matter if Riot explains how they calculate it or not. As soon as someone notices that it's not based off of the MMR (which enough people track) they'll start trying to figure out what it is based on. It's not like it would take a genius to notice "hey, on the games I get an S+ on I get a ton of LP, and games I get a D on I get hardly any LP, but it didn't use to do that," and thus the connection would be drawn to the game stats.

They could have no clue what the real stats are, but they'd still try and manipulate it. Kill Steals would be more prevalent (after all, they don't know they're treated the same as assists). Players would AFK farm losing games, and some would buy and place 20 wards in the fountain to try and increase their ward score. Some may try to tank turret shots in a futile split push to increase their damage taken numbers.

most people would probably just assume they need to play well and if this happen this will probably be rare cases imo

Unker1391 votes

First, factoring stats means that those stats become something to play for besides the win. A support or jungle killing minions lowers the lp gains of the laners even if it is the correct play for the situation. Kill stealing becomes real because you are also stealing lp (reducing lp gains is the same as losing lp)

Second, the system you are suggesting assumes that there is a specific way to win and you want riot to define that for you. Any strategy that does not focus on maximizing kda, cs and objective damage slows lp gains and is inferior even if you can win more often with it.

Finally, it does not factor in a players ability to adapt to less than ideal circumstances. My game yesterday had the adc choose a jungler and left me 1v2 as a support. My stats don't reflect it, but I managed to hold the first tower by myself for almost 15 mins before I was too far behind. I said as much in chat and while I backed, the team grouped and killed the bot lane and took a tower and a dragon...none of which i got credit for. I don't care about the credit, but if stats are part of lp, then I would have been punished because of my teammates choices.... the same issue your suggestion is trying to eliminate.

Th3 Rift Hero1 votes

{quoted}

First, factoring stats means that those stats become something to play for besides the win. A support or jungle killing minions lowers the lp gains of the laners even if it is the correct play for the situation. Kill stealing becomes real because you are also stealing lp (reducing lp gains is the same as losing lp)

Second, the system you are suggesting assumes that there is a specific way to win and you want riot to define that for you. Any strategy that does not focus on maximizing kda, cs and objective damage slows lp gains and is inferior even if you can win more often with it.

Finally, it does not factor in a players ability to adapt to less than ideal circumstances. My game yesterday had the adc choose a jungler and left me 1v2 as a support. My stats don't reflect it, but I managed to hold the first tower by myself for almost 15 mins before I was too far behind. I said as much in chat and while I backed, the team grouped and killed the bot lane and took a tower and a dragon...none of which i got credit for. I don't care about the credit, but if stats are part of lp, then I would have been punished because of my teammates choices.... the same issue your suggestion is trying to eliminate.

though if riot would implement this without saying how it works people will still try to play their best to win and probably wont abuse except if they are very smart at guessing the process, also ks wont count since the other player still get the assist, this is counted as (kills + assists)/deaths

More from Gameplay

More by Th3 Rift Hero