League of legends poor matchmaking

Virtually every human being on the planet is far more likely to notice mistakes and bad play from teammates than they are their own, and furthermore tend to excuse their own mistakes as momentary lapses while instead believing other people's mistakes are indications of their actual skill level. This is a well known cognitive bias known as the attribution error. The second is that you do have 5 people on your team.

No, not 4 other people - see above. Sometimes you are the one doing something stupid, messing up, playing poorly, getting caught out, rotating right through enemy territory, going on tilt, whatever.

League of Legends Matchmaking Explained, Myths Debunked

So essentially you have 5 chances in any given game for someone to be messing up. Sometimes in critical ways that result in serious snowballing. It's worth noting that Heroes has serious comeback mechanics so it's not as severe as it might be for LoL but there is still snowballing and getting an advantage is powerful. Being wrecked is much more visceral than the joy of dominating your opponent, and the really bad moments can unfortunately be more memorable than the great ones. Give this a try.

For about 20 matches or so, write down the results of the match and why you think it went that way.

ELOHELL. Why League of Legends Matchmaking sucks

Take note of when a particular ally or enemy! I don't mean one bad decision; I mean consistent irreconcilable failure to make good decisions or perform mechanically. The kind of thing that indicates they really should be in a lower league. Then come back and tell us how often this sort of match actually happened to you, using data rather than general impression.

I think you'll find that it's much rarer than "every single game. Yeah because the MMRs should rate players not heroes. A plat player should know by now that you don't dive towers unless yada yada. But you do dive towers. Our hypothetical inexperienced Diablo failed because he couldn't yet recognize those circumstances. Maybe you've got a Medivh ready to get you out safely. Maybe you're fully-stacked and there's only one tower so you know you can do it safely. Maybe there's a Naz with you to draw tower fire onto his zombies.

Maybe their Kael has 19 stacks on Convection and it's worth the death to reset them. Maybe there are two minions with you that can tank the shots just long enough. An experienced tank would see these situations and make the correct decision: An inexperienced tank wouldn't understand quickly enough why it worked for them. He would remember that he saw other Diablos successfully tower-dive, fail to realize that 10 stacks wouldn't make him bulky enough like the ones he saw, try it, and die.

Remember, all this decision-making is happening on the fly in instants while players are also trying to think about a hundred other things. Experience makes it easier to make the right decision in an instant. Anyways, it is always the exception. Plat is not an inexperienced anything, playing tanks instead of Chromie doesn't mean he suddenly forgot towers killed you.

One-trick Plats absolutely exist. They play their Chromie in placements, win a lot because they're good at Chromie, later decide to try out Diablo, and suck with him. None of that is the matchmaker's fault. So you're a one-trick Chromie and somehow don't know that towers kill you?

Then you shouldn't be playing the game with human opponents, because you obviously can't grasp the simplest gameplay concepts. Sorry, warsage, this is not an acceptable argument, nor is it acceptable in-game behavior. Most if not all of this is applicable to HOTS. The problem is that most of the problems HotS has come from the points where Blizzard decided to go a different way than Riot.

As example the seeding from quick match. Riot sets everyone to the same point and the only difference is whether you get tested upwards or downwards for your first few games based on your normal mmr. Also they heavily limit the maximal rank you can reach gold 1 with faster climbs afterwards.

Welcome to Reddit,

In LoL they take the old spread of players and reduce it to bronze to plat from bronze to gm after the placement games. That means you will be to low in rank normally divisions down is the normal rate unless you have great placements. After placements all restrictions are lifted and due to the increased climbing everyone is back to their old ranks in quite a few games if they really belong there. In HotS you can get placed in the highest rank after your placements. And the reason why this is necessary is that climbing is horrible here. You don't jump divisions, you don't jump promotion games, you don't get times as many points for a win as you would normally just 1.

All this means Blizzard can't place you too low because you would need endless time to get back up and also they use your rank for matchmaking which Riot never did which was kinda fun because you could play against the highest ranks as Bronze 5 player through tactic selection dodging which was the reason for the introduction of promotion game skipping. This should be fun. So on one side we have Blizzard.

Blizzard has all the data, and that is literally millions of new games to look at every single day. Well, that seems like an awfully good data set to figure out with great confidence if drafting from QM I assume that is what you mean? On a more serious note, you will be happy to hear that the other points have been adressed for the upcoming season. Anecdotally we all see the idiots abusing QM matchmaking from time to time. We also all played with people who were way too terrible for where they were placed.

It's too bad Riot has such a small amount of games for them to have determined seeding ranked with casual formats was a bad idea. If Riot was a big company like Blizzard and LoL had even half the player base of Heroes, their data would probably matter more. I guess we should just ignore it though since LoL is such a small unknown game.

But Riot did not do it the Blizzard way, doesn't that mean that Riot has data showing it's fine not to seed from Normals?

Want to add to the discussion?

And considering how much bigger League is than Overwatch, HotS, Starcraft and Hearthstone put together don't you think they have more data? Maybe they have a different implementation, Maybe it is because it's a different game. My guess is as good as yours. You ignore thousands of people complaining about placements, and the very real fact that their matchmaking is broken. Blizzard dev's have admitted to PRA and placement matches not working as intended.

Blizzard has the data, yes, and it doesn't say very good things about their system by comparison. Not sure if you know how these things work, but massive games like this aren't fixed with a few strokes. Their whole system has the QM seed baked into it, don't act like they would change it if they knew it was the better solution.

They are aware of these issues and can really only mitigate it by now. Blizzard seems to have the weird idea that randomness is always fun which goes against the actual psychological foundations.

Their system is designed to give you random points and take random points from you. The problem is that the random points taken feel bad because you get punished for no reason and the random points gained don't feels good because you didn't deserve them.


  • League of Legends Matchmaking Explained, Myths Debunked.
  • vedic match making software.
  • ?
  • Log In to GameFAQs.
  • ihk azubi speed dating köln.

Riot does the opposite. They make randomness less punishing and less rewarding and can compensate for that by giving or taking more points later on. If you lose anyway it feels bad anyway and if you are climbing and improving getting rewarded way more for the actual improvement and not for random fluctuations feels way better. I made the climb multiple times in LoL and HotS and in LoL it drags a bit for the first games around and then you shoot up insanely fast which is great.

In HotS it feels like a drag all the way through the ranks and random lose streaks feels way more punishing as well. Yeah they did and it works. Randomness isnt inheretnly bad or good almost every competitive game you watch has some randomness to it. Its about how you design your game so that randomness highlights skill. One of the most competitive sports is Poker and that game has a ton of rng.

Yet if you watch Poker you know that better players almost always win. The goal is to create a game where RNG doesnt win you a game by itself and give players opportunities to control RNG to a degree. Best example here are Dragons In LoL.

Riot's in-depth talk on matchmaking & MMR. Worth a read imho : heroesofthestorm

Its random which dragon will spawn but the game tells you that way in advance so you have plently of time to decide what you want to do with it. If you want to look at video games look no further than CSGO. That game is often decided by random spread and yet is one of the most popular esports out there. But a game with 0 rng can get solved and become really boring and stale as a result. Poker has 52 cards and are dealt according to the same rules each time.

Any given hand in poker has a fixed probability of winning based on the current game state.

In previous versions of Counterstrike there was no randomness to the bullet spray pattern. They added it in because players got too good at predicting and compensating for the fixed spread patterns. What you want in a game is variation not randomness.