Boards›Gameplay

Riot please fix your matchmaking, and give me a job.

PyrotekNick·5/15/2017, 2:47:10 AM·4 votes·388 views

Finally made it to my Silver I promos, get these two back-to-back matches from the 'ol Riot: http://matchhistory.na.leagueoflegends.com/en/#match-details/NA1/2500283646/39343699?tab=overview http://matchhistory.na.leagueoflegends.com/en/#match-details/NA1/2500278908/39343699?tab=overview

Seems like your matchmaking system is broken, since in my first match there, we (we being 4 Silver IIs, and a Silver I) got pitted against a Gold V, Silver I, and 3 Silver IIs. Second match, (my team was Gold IV, Gold V, Silver I, and 2 Silver IIs) against 3 Gold V, Silver I, and Silver II. I worked hard to get to this promo. Worked hard to get out of bronze 3, 2, and 1. Twice, if memory serves. Worked hard to burn out of Silver 5-3. Busted my ass to get out of Silver 2. Earn a shot at Silver I and get handed two guaranteed losses.

I've put up two posts now regarding alternative methods for evaluating Elo/MMR. By op.gg's estimation I've got somewhere between a Silver I and Gold V MMR. But when the LoL client matches me up with/against people, it seems to only look at how skilled my opponents should be, not how skilled my teammates should be as well. I haven't even gotten a token, "we can't talk about information like that" from you guys.

First match (from op.gg estimated MMRs): My team's MMRs: Teflon badger: 1378 OmegaBeaver: 1378 AngelOfDeath lwk: 1397 kimchibicboi: 1396 PyrotekNick: 1449 --My team's avg MMR: 1399.6 Other team's MMRs: Bronze Feed: 1428 toolacademy: 1428 Debbie Destroyer: 1477 ChangChangalang: 1373 fûckboy69: 1386 --Other team's avg MMR: 1418.4

Second match (from op.gg estimated MMRs): My team's MMRs: I Gank You Spank: 1352 Masterofeuphoria: 1416 renzf: 1379 DestroyerMardux: 1306 PyrotekNick: 1449 --My Team's avg MMR: 1380.4 Other team's MMRs: Veda Suvar: 1353 Mgegeda: 1417 POTGOD420: 1460 legendaryhotdog: 1460 soranounmei: 1426 --Other Team's avg MMR: 1423.2

Why is there a 40+ point difference in a matchmaking system that could pull from a silver-bracket body that has the most depth of all the brackets in the whole client? Why is there a 20 point difference, while I'm working out of this bracket? This bracket is the most well-populated. It should be the most easily match-made player bracket in the game.

But I get stuck with people who flat out ignore advice like, "lucian/mao will have the advantage early, don't hard engage" (karma/ashe went in at lvl 3 for an all-in that resulted in a mao/lucian double kill). Or have teammates that refuse to buy and use control wards. Or queued into ranked on Ahri with a total of 7 games played on Ahri this season (OmegaBeaver), 6 of which were losses, all of them in ranked.

I don't particularly care for an in-depth answer, but if y'all are reworking your matchmaking system, give me a job, I have ideas and I'm currently an engineering student. I can help write the equations that output MMR values in response to player behavior and/or win/loss ratio. The current system is archaic, and ineffective. It needs some TLC, and mebbe an overhaul. Setting some poor sod up for failure in his promos is pretty shitty, even if it isn't intentional. As an additional consideration for the Riot employee(s) that may view this, it'd be awesome, even if no overhaul of the matchmaking system is employed, to reward good behaviors with a small (1-5) LP bump. It would encourage the things that Riot wants to see from their community during games (farming, warding, healing, shielding, etc.) and protect many players from things like the afk types who plague League as a whole. Imagine playing a ranked game, having your jg afk, but going through the game with minimal deaths and sub-healthy farm. No reward for your farm, but you didn't continue the feeding frenzy, and managed to snake a kill in somewhere as well, perhaps you see a couple LP returned to you as compensation for your good play and it will help struggling players who're adapting to good behavior to advance themselves (just an idea).

Preemptively, (PyrotekNick) I have 397 games of league played this season (more by the end of the night most likely). I have a 52% win ratio overall, 47% win rate in my last 20 games. I usually queue jg/supp but had to break up the monotony of that queuing lifestyle (hence the top and mid roles that show up). All that to say, I've been back in bronze a few times this season, and there is a large populous in the silver bracket, and I'm getting frustrated that my adaptation to better gameplay mechanics and behavior is only rewarded with more frequently difficult opponents without the teammates to match.

Thank you for your time, -PyrotekNick.

5 Comments

Sasogwa5/15/2017, 7:38:13 AM3 votes

..? That's barely different. If you can't carry your team it means that you don't deserve a higher mmr than what they give you. Also, I don't know if you've heard of how Riot handles winstreaks, there's a really good algorithm to sort out smurfs to make them stay in an elo the lowest amount of time possible. But there's kind of a distortion between league and MMR ESPECIALLY if you're on winstreaks (smurf being a good example). It has happened to me to see some things like gold 5 smurfs being matched with plat or similar stuff. Leagues kind of lag behind MMR, that's why you sometimes skip divisions and even promos. At the same time, someone on a lose streak won't climb down that easily and can be of a MMR that's lower than the league he belongs in. That being said, the MMR is hidden, and calculations made by other programs are ESTIMATIONS. Estimations can easily be wrong by 30-40 points. Your argument is therefore invalid. And eh, I've seen much much worse in matchmaking in normal games. The games you show don't even make me raise an eyebrow, the MMR is really close and you're complaining about nothing.

If you want some real good matchmaking wtf here you are: http://i.imgur.com/snRlzmX.jpg (funniest things are that I won lane against Riven with my uncompleted T2 runes Nasus, and Wukong played the worst in our team)

Quepha5/15/2017, 7:45:53 AM2 votes

The average estimated MMR's are not that far off even though they're also op.gg estimates which are not accurate. If anything your example goes above my expectations for the matchmaking system.

It's crazy to me that you've played almost 400 games this season but are throwing a hissyfit over getting two losses in a row. Do you realize that, were matchmaking 100% perfect, you would get 2 losses in a row in 25% of all pairs of games that you'd ever play?

Who is Taliyah5/15/2017, 7:51:24 AM2 votes

There is like actually no difference between silver II - Gold IV it's totally normal to get matched with gold 4 - 5 and even a bit higher as silver 1 and silver 2.

FourVsFive5/15/2017, 7:27:10 AM1 votes

You were vs a Lucian 2 times in a row is the problem. He stupid strong. be kinda hard to match perfect elo rating with peoples role preferences no?

PyrotekNick5/15/2017, 2:42:06 PM1 votes

FourVsFive: You are correct, Lucian IS stupid strong, so are Twitch, Ivern and several other champions...can't ban everything that is strong (or the other champs that aren't particularly strong, but people in Silver brackets don't know how to deal with). Naturally, the 10 ban system will assist with that particular problem.

Sasogwa: So, generally speaking, when airing a complaint, it is desirable to provide a solution. Which I have. I understand that its easier to critique anything that is put out there, since you have no investment contained (in this particular complaint), however with your comment, you claim that the entire argument is invalid, but there are traceable statistics that show that the match was set up for failure from the outset. Player warding over just their last 20 games shows my team to have a disturbing deficit, the mid-lane Ahri had only 7 games played on that champ and queued into ranked on that champion, and several others. I bring this up because while TaliyahToMaster is right in a general sense about player skills being ~semi-equal, the reason many of the folks who made it to Gold are there, is due to macro-mechanics that the execute well. Where in Silvers, and Bronze the players may have an understanding of their champion and micro-mechanics, their macro decisions are generally speaking, appalling.

Quepha: I'd like to see the math that says I should have a 75% win rate, and would like you to hook me up with more wins to accommodate that claim. This isn't a 'hissyfit' over two games. This is a several month long complaint, that still has not gotten a glance from a Riot employee, even to tell me that I should take my ideas and shove 'em. The MMR matchup is fairly close by the numbers, and still worlds apart by skill, as shown by CS, wards, KP% and other statistics that aren't estimated, they're easily trackable by op.gg.

TaliyahToMaster: I understand that I'm calling out the bracket differential as the cause of my loss. However, the disparity between champion skill capability, as you noted, is negligible. What I said above to Sasogwa holds weight here too, the reason these players are in Gold and not in Silver, is due to their macro-mechanics being executed well. In Silvers, and Bronze the players may have an understanding of their champion and micro-mechanics, their macro decision making is more likely to result in a loss (bad baron plays, not warding, etc.).

To all of you, I've adapted to the lack of map-awareness in Silvers by over-compensating. I buy an excessive amount of control wards, I try to maintain vision of the entrances to my jungle when I'm jungling, and cover my lane when I'm laning. I've taken the time to watch the pros to see where effective locations are to place wards that will last for the maximum duration and still provide adequate coverage. That has no bearing on the outcome of my games, when the rest of my team purchases a total of zero control wards (support included) to protect their flanks, or places wards in standard locations that are easily anticipated. Or, goes into a brawl bot lane within 2 minutes of being warned that they'll lose that trade, or has only 7 games played in season 7 on the champion they're using mid lane...the list goes on.

There was more to those matchups being a failure from the outset than a single lane feeding or going afk, which happens frequently. I've offered solutions to these problems, that also reinforce the behavior Riot has started to look at encouraging (healing, shielding and vision control, among other things). So before you go hating on the complaint because I'm throwing a, "hissyfit" take the time to notice that a solution was provided, so future players don't have to experience widespread bad behavior without any reason to do anything other than emulate what they see.