### Welcome to the Forum Archive!

Years of conversation fill a ton of digital pages, and we've kept all of it accessible to browse or copy over. Whether you're looking for reveal articles for older champions, or the first time that Rammus rolled into an "OK" thread, or anything in between, you can find it here. When you're finished, check out the boards to join in the latest League of Legends discussions.

### Loyal player begging for an answer from the devs

Comment below rating threshold, click here to show it.

hellsingsage

Junior Member

Stop complaining about what you think is wrong about the game and enjoy it. It's better for you and for the community that this fine game serves with diligence.

Cheers.

P.S. If anyone out there could be a gent and follow a buddy of mine working his way up.
http://www.twitch.tv/g1_metasamus

Comment below rating threshold, click here to show it.

sab9999

Member

i think you just suck

Comment below rating threshold, click here to show it.

1400orBUST

Member

Quote:
PogoPogoPogoPogo:
First, the equation for what, exactly?
Second, what do the variables stand for?

Performance rating = [((Total of opponents' ratings) + 400 * (Wins - Losses)) / Games].
is an explanation of the formula I previously referenced and is the core of Elo's work which assumes any game lost was completely within your control to win or loose versus a single other opponent.

Comment below rating threshold, click here to show it.

Member

Quote:
1400orBUST:
Performance rating = [((Total of opponents' ratings) + 400 * (Wins - Losses)) / Games].
is an explanation of the formula I previously referenced and is the core of Elo's work which assumes any game lost was completely within your control to win or loose versus a single other opponent.

You're insane. That's to estimate a player's performance rating at any given chess tournament. So if I won 5 games against 5 2k Elo players, my performance rating would be:

(10000) + (400 * 5) / 5, or 2400 Elo.

This has no bearing on my actual rating or who I'm matched against in League - for all you know, I could be a 1k Elo player and get crazy lucky.

Comment below rating threshold, click here to show it.

Arinthe

Member

Quote:
1400orBUST:
Performance rating = [((Total of opponents' ratings) + 400 * (Wins - Losses)) / Games].
is an explanation of the formula I previously referenced and is the core of Elo's work which assumes any game lost was completely within your control to win or loose versus a single other opponent.

Take a step back for a second, take a deep breath. AD bottom has been very patient with you in this thread and explained everything very well. Consider the possibility that you don't understand things as well as you think you do. Consider the possibility that AD bottom is correct in his arguments.

Comment below rating threshold, click here to show it.

paintolar

Senior Member

Maybe, just maybe, the matchmaking system couldnt find 9 other people that matched (or were ver closeto) your elo/win:loss. And instead of having you wait 30 min in queue, it matched you with people who were available?

Comment below rating threshold, click here to show it.

1400orBUST

Member

Quote:
paintolar:
Maybe, just maybe, the matchmaking system couldnt find 9 other people that matched (or were ver closeto) your elo/win:loss. And instead of having you wait 30 min in queue, it matched you with people who were available?

That is a very real possibility, Most matchmaking systems do have a time tolerance built in and the longer they go without finding a match the larger a variance in skill ratings they will consider.

Comment below rating threshold, click here to show it.

PogoPogoPogoPogo

Senior Member

Quote:
PogoPogoPogoPogo:
How creative.

Match-making has been *****ed about since the dawn of online-gaming. In fact, I'm certain that complaints about match-making probably existed before match-making itself even existed.

First and foremost, THIS.

Seriously.

Match-making has been *****ed about since the beginning. People today still think it's as bad or worse as it ever was before there was any sophistication at all to the match-making.

People complain about match-making in every game that has a match-making service and this is applicable across all genres.

Why? Well, it's not because there's some secret conspiracy for all gaming companies to keep match-making a low priority. In fact, if I'm Riot and I'm doing what I'm doing and trying to put eSports right up there in watchability with real sports, I'm going to have to be extremely confident in my match-making system. Why? Because I want my salaried gamers to actually be significantly better gamers then the average player.

The NFL wouldn't be interesting to watch if it consisted of average high school players rather than the best of college players (which are the best of high school players). Similarly, LoL's professional tournaments wouldn't be interesting to watch if it were a bunch of 1200 Elo guys playing against each other. No one would take it seriously in the slightest.

...and you can't get the cream to rise to the top without a solid match-making system.

But why do players complain so much about match-making systems in online games? I'll tell you why... it's because players are inconsistent. Video games are just that way. Even at the elite levels, players like Crs Nyjacky don't have a perfect CS every game.

The same applies to sports actually. I mean, consider college basketball. The aim of the NCAA tournament is to appropriately seed the teams according to how the performed over the course of the season. Yet despite this, we can sometimes see extremely low-seeded teams perform very well.

The #2 team losing to the #15 team happens. The #2 team doesn't complain about match-making though, even if the #3, #4, and #5 teams all got paired against easier opponents (and the #2 team should have been paired with the easiest of those 4).

The fact of the matter is, it is absolutely 100% impossible for any match-making system to account for inconsistency in player skill level.

Comment below rating threshold, click here to show it.

PogoPogoPogoPogo

Senior Member

By the way, as for your question about match-making pairing a 3W-20L player with a 20W-30L player, and how that's possible, or how those two players could both be at the same Elo...

First off, we don't need to assume they started at the same Elo. At the end of season 2, if the 3-20 player had a significantly higher Elo than the 20-30 player had at the end of season 2, then with those records, even gaining the same identical amount of Elo for each win and loss, the 3-20 could easily be at the same Elo as you.

And by the way... his losses would have been coming against better opponents than your losses.

Comment below rating threshold, click here to show it.

1400orBUST

Member

The point of the original post has gotten a bit sidetracked, my objection to AD bottoms examples is that they simply aren't statistically possible as an explanation of the real numbers in this case. The numbers he used were quite literally pulled out of thin air. It's easy to make up some random example with fictional numbers and have it sound like it makes sense. But in this case they call the skill ratings for LoL Elo ratings when they aren't actually based on Elo's work which leads to a great deal of misunderstanding. I referenced the performance rating simply in that if you consider all the games played so far this pre-season as a single tournament and you figure my performance rating against the two players in question's ratings you would see that there was no way in an actual Elo system we would have been considered a fair match. If I wouldn't be considered a fair opponent for these players how is it they were considered fair teammates for me. I'm all for teh game at large and I do both praise the Devs for the work they do and try to support them financially by purchasing riot points when possible. So to suggest that I should just be more grateful and stop complaining is ridiculous. We're talking about ranked play not a blind pick game. The fact is that every game is logged, every players performance is in a database that can easily be searched for anomalies in statistics that would clearly show outliers in the statistical modeling which would allow them to easily identify players who are intentionally throwing games or whose skill level or ability to choose champions is so poor so as to appear intentional. It would be infinitely simple to segregate these players from the population at large without even telling them you had done so. The trolls wouldn't even realize what had happened until their 20th game with 9 other trolls.