Boards›Gameplay

MMR and rank should incorporate individualized success

Zelnick·11/9/2018, 11:49:40 PM·5 votes·3,145 views

Speaking as someone who was diamond for 2 seasons, I feel more confident about my abilities than a few years ago. But, trying to grind through to it again at the last minute after being unable to play ranked during prior months leads me to think less of the team-oriented nature of the game and more on individualized success, not just early game but throughout the game, whereas originally I was focused more on teamwork over a longer period.

Throughout say, 75% or more the games I lose, I still have the most damage if not the second most, I still have the highest or second highest kdr or occasionally the third highest, I still have an average or above average vision score, I usually get kills early or mid game and go 2/0 or 3/1 if not better, I ping whenever my lane roams and ping for other people's lanes when they forget or for taking objectives, I roam when I push ahead a wave and see another pushed back after clearing wards, etc. But somehow, given that multiple people who have played with me think I'm just extra cursed with being unlucky for being paired with bad players after bad player, it feels like no amount of growth, planning or adapting matters in any capacity, it's just the luck of draw.

Are there things I could be doing better? As always, there are many things anyone could be doing better, but what about what players do right now? Solo-queue just doesn't support team-oriented behavior, and the meta always changes to disproportionately favor specific archetypes in solo-que, thus defeating the purpose of league's culmination of over 100 characters. Give your team a competent plan that you've seen work multiple times in the past? Well, it still doesn't help because even with a mic, solo queue players will do what they want regardless. With this in mind, it only follows that after 8 years, the mmr and rank algorithm should be more sophisticated to account for individual success against comparative odds. The elo rating system is literally decades old, surely Riot has at least one new idea on that front.

40 Comments

ZephyrDrake11/10/2018, 1:03:59 AM7 votes

Let's punish team players and reward people who just want to pad out stats.. great logic you have there

Madsin2511/10/2018, 4:50:57 AM1 votes

Have you ever watched high elo one trick zyras/anivias?

You can learn a lot from doing that.

haaaaaaalp11/10/2018, 5:26:26 AM1 votes

The system can’t really do better. We saw what happened when riot tried to measure individualized success. As flawed as win/ loss is, that stat is the best measure of individual success riot has. Riot probably had dozens of other systems but given that the best they came up with with the current grading and stats, it is highly doubtful they will come up with one correlates better with individual performance than elo.

haaaaaaalp11/14/2018, 2:07:57 AM1 votes

{quoted}

What investment? No one has spent money on this thread, people have barely spent any time on it. Once again, you're purposely conflating two different concepts with no evidence to support it. Riot never stated any intention for chest-reward grading system to substitute elo, nor is there any reason to suspect it could be. I offered suggestions for creating a more accurate solo-queue measurement, and before even taking a second to consider them, you shamelessly strawmanned what was offered with something that doesn't even make sense.

"The grading system doesn't do what no one intended it to do therefore everything is hopeless" is not a cogent rhetoric.

It is a per-requisite, no one needed to outright state any intentions of replacing the elo model. In order to replace the elo model with one that measures individual skill, riot has to be able to create a measure of individual skill. Riot has already shown that they fail at the per-requisit step.

You can't just assume a more accurate solo-queue measure than elo exists. Riot has to create one and they have shown that cannot do that.