A Look at the Ranking System

The new ranking system has been out for a while; finally, people have had some time to settle in, learn how it works and to finally start hammering away at ranked under the guise of the “new” ranked system. But is it really different from before?

Warning: The next part of this article is a brief overview of how the MMR system works. If you already understand how it works, then feel free to skip it.

Previously, we saw the Elo system that everybody loved to hate. “I’m 1400 Elo, but I really should be 1700!” people would cry. “The Elo system isn’t fair, it doesn’t measure individual skill!” After much ranting and hoping for a new ranked system, here we have it, the League system. On the surface, the League system is much different. You no longer have that magical Elo number to turn to, instead you are now fighting for LP. How much LP you gain per game is based off of your MMR(Match-Making Rating) which is more or less the Elo system behind a fancy curtain. If you win against people at a higher MMR than the average MMR of your league, you gain more points, and you gain less if it’s the other way around. You’ll lose more if you’re losing to people below the average MMR of your league. Nothing exciting to see here, it’s remarkably similar to the Elo system, however the old system is now hidden.

Overview Ends Here

Is the new system a good thing?

Much like the Elo system, there have been people who have claimed that the MMR system is flawed, and that it doesn’t actually create a fair and accurate representation of a person’s skill. But why would this be?

Many of the arguments that I have found have largely been revolving around people who are in Division V of a particular tier; IE Silver V, Gold V, Platinum V and so on and so forth. There are some who will claim that the biggest problem is the inability to actually climb after a losing streak. If, for example, you are in Gold V, let us assume that you recently lost six games in a row. Your MMR will have dropped by approximately fifty points, and next thing you know, you’re gaining 6 LP every game you win, while losing 20. This is an apparently insurmountable amount of winning; if this rate of gain and loss were kept up, you’d never get promoted!

It turns out that people who have found their way into a League that it is difficult to win consistently in, it becomes very difficult to make that jump from V to IV. It is a common frustration that “I have gotten 44 promotional series, but lost them all. How do I not deserve to be in a higher division?!”  It might even cause them to look like this:

Well, my friend, while your frustration is certainly valid, one must ask: Why does it seem like people consistently cannot win a promotional series? Realistically, they are no more or less difficult than any other set of three games you have played; you will play with the same levels of people. Sometimes the people on your team are playing for promotion, sometimes they aren’t. Is this a problem?

Perhaps. Perhaps there should be a filter: if you are up for promotion, should the game try to match you with people who are also going up for promotion? That would make sense. Everyone is trying their hardest those games, nobody wants to lose a game that they’ve worked so hard to get to. The promotion series is supposed to be your culmination of work in regards to your time in a particular division. If everybody in a game is up for promotion, then everyone is trying their hardest to win; they all want that promotion, so hopefully this would reduce the trollpicking and reduce toxicity; or would it?

This could easily work either way. People may try their hardest, or they may give up immediately once they see they don’t get their best lane. So while it may sound right away to be a great idea to put everyone who is going for a promotion together, overall you’re just as likely to get people who once again will have even less trust in their teammates due to the importance of a series of matches. Everyone believes themselves to be better than their teammates, and this will inevitably lead to conflict.

Think of it numerically, as well.  There will likely be too few people to actually play a promotion game in any reasonable period of time: in order to make everyone line up, you would need ten people, all of whom managed to get to a promotional series . How many people are up for promotion in your league right now? I doubt it’s more than three, and if you’re one of them, there’s still a good chance that you’re the only one of them online. Even if that isn’t the case, the other person could be just as likely playing a normal game, or an ARM, or simply not playing a game at the moment. The chances of you lining up ten people in a reasonable people of time is minimal at best. You would likely end up delving into the realm of the high elo curse or 10 minutes queue times.

Not only do we enter the realm of people getting frustrated that they don’t get the role that they “deserve”, but if we were to treat everybody that was in a promotional series as one cohesive unit, there would suddenly be a vastly different range of people’s MMR. As it stands right now, it is possible for one person who is a much lower overall MMR to reach a promotion series at the same time as someone that is higher. Because promotion is based off of wins within your division and tier, there is often an MMR differential between people who have managed to reach their promotion series.

Example:

Player A loses a large number of games in a row in Platinum V. Because of these losses, his MMR drops down to 1550, and gains 8 LP every game, while losing will cost him upwards of 25. However, whether it be through luck or improvement, he wins 12 games in a row, and qualifies for a promotion series! Yay for Player A! His MMR after those 12 games in Platinum V is now back up to 1650, and he qualifies for a promotion!

 

Meanwhile, player B entered Platinum 5 with an MMR of Approximately 1720, and remained even for a while, but has managed to overall go positive. He wins more than he loses, and eventually reaches his Promotion series, but does so with an MMR of 1800. Both of these players have reached the magical 100 LP, but do they deserve to be matched in the same games? Player B has clearly been doing better overall; he won more games in a much shorter period of time, and did so without destroying his MMR and putting himself in a position to need to recover it. Does this mean that the two players are, overall, of equal skill?

This skill devaluation will lead to a reduction in prestige for being in this particular league: if people who are being matched with Silver and Gold people are in Platinum V, who cares if they’re Platinum? Just because they achieved it once, doesn’t mean it is a mean representation of their skill: it is a peak representation.

If they both win their promotional matches, than that is certainly what it would seem. Player A would move right on up to Platinum IV with a rating that many people would have at the start of their Platinum V careers. Some would say that the person who has such a low MMR would not deserve the promotion, and there is a system in place for this. If you manage to get yourself an extremely low MMR, even if you successfully win your 2 out of 3 promotional series, you won’t be promoted. This is in order to encourage the idea that there is a certain consistent quality of play at each division; a person who has such a low MMR has clearly proven that they do not deserve to be at a higher Division than they are at.

So here we stand. How well does it work? I want to expose any biases I may have here: I had a peak Elo of 1582 in Season 2, and when the league system was implemented, I had an Elo of approximately 1550. At the introduction of the new system, I was placed at Silver I, and since then have accelerated past Silver, and as of writing this I am in Division I of Gold. I have never believed in the concept of “Elo Hell” and do believe that the whole shpeal about being able to carry yourself to an Elo actually representative of your skill level is true. I find there to be overwhelming evidence of this, and people who disagree tend to turn to sheer anecdotal evidence and a culmination of their own inability to climb. This is my viewpoint, and this is the angle from which I am approaching it; if you disagree with this initial assessment, I hope I haven’t lost whatever attention of yours I may have had.

I think the greatest flaw that I personally see in this new systems is the inability to be demoted through tiers. That is, once you hit Silver V, you’ll never be Bronze again! On the surface this seems great; you no longer have to worry about retaining that ranking that you worked so hard for: You’ll always get to keep your silver! But what if your play consistently worsens, and you continue to slip into a pattern of playing that is worse than it used to be? What if, after winning your promotion series, and finding your way into that next division it turns out to be too much for you to handle, and you are consistently losing 60% of your games?

Things like this can happen. No one plays at a level indicative of their top skill at all times; but at the same time, if someone scrambles into a new division, but clearly proves that they cannot perform there via being stuck with a win rate in the mid 40%, there is no repercussion. It becomes harder to climb, sure, but the mentality can easily develop that “I got here, so I deserve this” when that may not necessarily be the case. There have been rumblings of suggestions claiming that it should be possible to be demoted from your current tier, and while I hate more than anything “riot pls” styles of posts, I am inclined to agree.

Locking someone in the fifth division of a tier means that you imply that they play as well as anyone else in that tier at any given time. Giving the option of a demotion from a tier completely gives people incentive to try to continue to improve and get better. However, it should be very difficult to be demoted. I’m talking very difficult. I’m talking 45% or lower win rate of a high number of games; this would make it so that you aren’t punished with a full tier demotion simply if you go on an unlucky losing streak. Whether it seems like it or not, it is a very large jump to move from tier to tier, and it should be reflected in a possible demotion.

This all boils down to the question of: Is the new system working any better than the Elo system? I certainly think so. There are certainly some flaws, and you can certainly get unlucky now and again with your promotion series, but if you are able to make it once, then you should be able to make it again. If you can’t, then chances are you don’t deserve to be there: it’s the same deal if you find yourself losing promotion series games over and over. Some will say that they are stuck in the classic “Win every game, lose promotion series, win every game, lose promotion series” loop, but in actuality, you are just as likely to win a promotion game as you are any other game. The cries of “you need a 70% win rate to go up tiers!” are largely false, and based off of faulty math. If you have a 70% win rate, I promise you that you will fly through tiers like no other. There may be occasions in which you have to do a promo series now and again, but you will rise. Period.

The new system seems like a positive thing; while it’s certainly possible to get yourself stuck in a division, it’s really no different than if you were to be stuck at 1400 Elo. The only difference now is that it’s harder to see exactly how far away you are from your goal if you find yourself losing in Division V. Other than that, it serves its purpose of being able to see how far along you are to your goal, as well as having easier to achieve and smaller checkpoints, instead of simply Bronze, Silver, Gold, Platinum and Diamond. While it isn’t perfect, it’s certainly a large step in the right direction.