Need help? A guide to ELO and ways to improve your rating.

1234567 ... 8
Comment below rating threshold, click here to show it.

Full Goat Soraka

Senior Member

01-08-2011

Quote:
Originally Posted by robotninjaanna View Post
How do you check your ELO? I just started ranked recently and i wouldn't be surprised if I don't have one yet, but I have no idea how to check it.
Once you get to level 21 or something(not sure about that),you can play ranked games with a premade team. But in a 5v5,you can only play with a premade team of 5,same thing in 3v3. Everyone starts out with a Rating (ELO) of 1200. You gain points by winning Ranked games. You lose points by losing games. But if you want to see your rating,you just go to you Summoner Profile,and look at you Ranked Stats. If your level is high enough,you will be able too see your rating there.


Comment below rating threshold, click here to show it.

DarkPsx

This user has referred a friend to League of Legends, click for more information

Senior Member

01-09-2011

Good to keep this thread going. I'm still in Asia right now so I'm only on occasionally.

The reason I put out my initial post as just a simple outline/guideline is because my belief as a serious competitive gamer comes from a very solid understanding of basic mechanics of a game. The cool part about that is that I apply this to many things in life, not just in games.

---After just proof reading this, I realize this is a kind of corny way of trying to explain ELO. But I had fun writing it, so feel free to just stop reading here if you don't like "wasting your time" ---

Let me take a physics course as an example. Lets just assume there is an ELO in a university physics course, just for fun. Your ranked matches would then come from your homework assignments which would most likely be worth about a plus or minus 15 points per assignment, and test scores which might be able to affect your score by about 50 points per exam.

So:
50% on homework = 0 rating change.
0% on homework = -15 rating change.
100% on homework = +15 rating change.

50% on exam = 0 rating change.
0% on exam = -50 rating change
100% on exam = +50 rating change

To get the highest score in your class, you need to work hard in order to learn the material - all the material, how it correlates, and how to apply it to build on your knowledge of mechanics. To learn all of these things, you must master the basics. How will you learn the more advanced levels of play and dive into more material without the complete knowledge of basics?

Players in the intermediate range (around 1475 - 1825) are like those people in physics class ( and I'm talking as an estimate, my disclaimer is that there are always exceptions) who get away with 70%'s on all their exams, the occasional 100% on homework and maybe they just forgot to turn in a few assignments (and therefore a 0%). If they work hard to get 100% consistently and prove their know the material, there are better chances that they will get 100%'s on future exams because the knowledge builds on top of the past exams, and therefore their personal rating will go up.

However, I know this is a team effort. You need to rely on your teamates to also get high scores in order for the team to do well. But this post strictly toward how to help understand why the basics are important and how it may correlate into a better rating. If you yourself as a player start improving with these ideas, there is a better chance that your team will win, because YOU have improved.

I know this was really corny, and probably a boring read. But knowledge in this game, whether it is helping yourself or somebody else, raises the bar in a competitive game over time. If you look at it from a personal point of view, it is simple to see - you play a carry, you get lucky, you catch the other team in a bad spot, you team wins and get a better rating. If you are able to look at it from the big picture, You win - players ask you how you did this and that, you tell them. They then tell their friends. Their friends might be nice enough to share with others as well. And from that, you have singlehandedly raised the bar in competitive play - adding a not so well known factor into the pool. This is how ideas are shared, knowledge is absorbed AND BUILT UPON, and great/talented players are formed.

Keep this in mind for all your future gaming endeavors - you will be surprised what you can come up with using an approach like this. Suddenly it won't be as much "looking up what the pro's do and doing it", more like "watching what the pro's do and figure out possible ways to creatively express it yourself when you are able to improve upon it".

I would like everyone to continue this thread, no matter what their viewpoint. The collective of all of our opinions are what make this thread more and more useful to our gaming community.

Cheers!


Comment below rating threshold, click here to show it.

Amatzikahni

Senior Member

01-09-2011

Quote:
Originally Posted by lillkilla View Post
fact of the matter is that in this game ELO means nothing about your solo skill level. it goes down when you lose ranked games with a horrendous team, therefore keeping you forever in this elo. if you want your elo to go up you need 5 people that play together extremely well. great guide tho.
Wrong. There are four slots that make up the rest of your team and five slots that make up the opponents'. Therefore, every game has a 20% chance that you will be the deciding factor irregardless of your teammates. Play an infinite number of games and the error reaches zero and your ELO will be determined. Or, just play a few hundred since infinite is not possible. Where you end up will almost entirely be determined by you.


Comment below rating threshold, click here to show it.

b EZ

Member

01-09-2011

I wish it was illegal to pick 4 solo mid chars and an OT. I cant choose characters for people, therefor I cant win games when people don't know how to pick characters.


Comment below rating threshold, click here to show it.

Onisake

Senior Member

01-10-2011

Quote:
Originally Posted by DarkPsx View Post
Good to keep this thread going. I'm still in Asia right now so I'm only on occasionally.

The reason I put out my initial post as just a simple outline/guideline is because my belief as a serious competitive gamer comes from a very solid understanding of basic mechanics of a game. The cool part about that is that I apply this to many things in life, not just in games.

---After just proof reading this, I realize this is a kind of corny way of trying to explain ELO. But I had fun writing it, so feel free to just stop reading here if you don't like "wasting your time" ---

Let me take a physics course as an example. Lets just assume there is an ELO in a university physics course, just for fun. Your ranked matches would then come from your homework assignments which would most likely be worth about a plus or minus 15 points per assignment, and test scores which might be able to affect your score by about 50 points per exam.

So:
50% on homework = 0 rating change.
0% on homework = -15 rating change.
100% on homework = +15 rating change.

50% on exam = 0 rating change.
0% on exam = -50 rating change
100% on exam = +50 rating change

To get the highest score in your class, you need to work hard in order to learn the material - all the material, how it correlates, and how to apply it to build on your knowledge of mechanics. To learn all of these things, you must master the basics. How will you learn the more advanced levels of play and dive into more material without the complete knowledge of basics?

Players in the intermediate range (around 1475 - 1825) are like those people in physics class ( and I'm talking as an estimate, my disclaimer is that there are always exceptions) who get away with 70%'s on all their exams, the occasional 100% on homework and maybe they just forgot to turn in a few assignments (and therefore a 0%). If they work hard to get 100% consistently and prove their know the material, there are better chances that they will get 100%'s on future exams because the knowledge builds on top of the past exams, and therefore their personal rating will go up.

However, I know this is a team effort. You need to rely on your teamates to also get high scores in order for the team to do well. But this post strictly toward how to help understand why the basics are important and how it may correlate into a better rating. If you yourself as a player start improving with these ideas, there is a better chance that your team will win, because YOU have improved.

I know this was really corny, and probably a boring read. But knowledge in this game, whether it is helping yourself or somebody else, raises the bar in a competitive game over time. If you look at it from a personal point of view, it is simple to see - you play a carry, you get lucky, you catch the other team in a bad spot, you team wins and get a better rating. If you are able to look at it from the big picture, You win - players ask you how you did this and that, you tell them. They then tell their friends. Their friends might be nice enough to share with others as well. And from that, you have singlehandedly raised the bar in competitive play - adding a not so well known factor into the pool. This is how ideas are shared, knowledge is absorbed AND BUILT UPON, and great/talented players are formed.

Keep this in mind for all your future gaming endeavors - you will be surprised what you can come up with using an approach like this. Suddenly it won't be as much "looking up what the pro's do and doing it", more like "watching what the pro's do and figure out possible ways to creatively express it yourself when you are able to improve upon it".

I would like everyone to continue this thread, no matter what their viewpoint. The collective of all of our opinions are what make this thread more and more useful to our gaming community.

Cheers!
that's...not exactly how ELO works. you're right. it's corny. and it's far over-simplified. what LoL uses isnt' strictly an ELO system. i think it's just called that because it's the closest analogy. in fact, i don't ever remember any official riot staff ever calling it ELO. i think it's purely a community term.

ELO is suppose to be a normal curve. what you've described is a right shift curve.that may not be what is decribed in LoL. in fact i would bet it's actually a left shift. as in most cases it's a left shift for chess and go. also, it's possibel to not gain or lose ranking in a real ELO system. in league, you always gain points after a match, or you always lose points.

we dont' have access to all of the ranking Data. for the most part, you're going to see a 'left shift' in ELO, meaning the median is below the average. and the actual average is below the for-seen average. IE, i would bet that the median ELO ranking is around 850, adn the average is around 1k for the total distrubtion. meaning the majority of players are 'unranked.'

ideally, because when you first enter ranked you're at this ELO, the average is somewhere between 1100-1300. or about 1200. that being said, an 'intermediate' player should be around this ranking. (dont' know where you got 1400+ as intermediate, that's absurd)

you dont' want to rank all of your intermediate players. only the 'upper' echelons of intermediates shoudl be in the rankings. otherwise there's just too much data to report. i don't knwo the total size of the LoL player base. but between smurf accounts and the people tha tonly play 4 or 5 ranked games, there's a ton of data that needs to be filtered. because some people on the ranking board don't play consistantly enough to be a valid statistic.

so our actual 'intermediate' range is below 1200.

things are split into standard deviations, and categories. you gain/lose the most points when you first enter ranked. i think this is a bad idea, because draft selction is far, far, far different than blind selection. there's no time to adjust to the learning curve. there's no 'practice' phase. there's no draft selection for normal games, so there's no real way to practice, other than custom games. and that's not always a good choice, because the spread is way too big, and lots of people there are going to be trying new champs.

ELO, in its most successful form, is used in chess and go. which are 1v1. because this is 5v5, your individual rating doesnt' mean a lot. ELO should really be made for teams. because you're not against any single individual when you're playing, there's no way to determine how well you're actually doing. aside from KDA, minion kills, and tower kills. but even this can be tricky. there are no assists for tower kills. which are essential for winning the game. a tower kill is worth far more than a champion.

minions are like pawns, champs are like knights, and towers are more like queens. except there's also the added complexity of scaling. our knights eventually become stronger than our queens. and you can also do extremely well by killing only minions (Sivir). but i still think it's a valid analogy, because you can lose a chess game without losing your queen, and you can demorilize your opponent enough in LoL and win without taking any towers. of course this is extremely rare.

in a nutshell: the other places ELO is used there are a set number of points on the board. in GO, there are only so many teritories to capture, and chess, there are only so many pieces you can take

for LoL, the number of pieces varies because of spawning based on time. that being said, you could be awarded points based on the duration of the match, and the total potential points available from board, or map, objectives.

of course the problem with this is KSing. there is also no mechanic to get an assist on a tower kill. so if you're pushing a tower, and your minion happens to steal the last hit on a tower, you lose potential points. it also doesnt' provide any rewards for support champs. as support champs like Soraka or Sona dont' take towers, don't take map objectives like golem or lizard or dragon. they are there to help the other players.

As Soraka, i don't ult to get assists. i could, but that's not always when my global heal is needed. more often than not, it acts as an 'oh ****!' button. so support champs are pretty much doomed to having low scores, especially ones like Soraka and Sona.

a fix for this is to allow kills and assists to be worth the same amount of ELO. but even then that's more of a bandaid. creating a good balanced ranking system for a game as complex as LoL is a duanting task. there are too many things to keep in mind. a champ like Soraka helps mainly with her healing and debuffing. but if you add healing into the point system, summonres who use mundo have an advantage, as he's the best healer in the game. you could essentially stack health, dive a team, get low, and then just ult to inflate your rating.

it's harder for us to understand, because we
1) don't have all the data riot does. and they aren't likely to release it. (for good reason)
2) the rules and formula for calculating rank aren't defined publically. (that i'm aware of, and for good reason)
3) there are players that are over-ranked and players that are under-ranked. throwing off any manual collection of data on our end to try to reverse engineer the formula.

what we do know is this: you win, you gain points. you lose, you lose points. that's really all their is to it.

so essentially, if Riot wanted to use a real ELO system. it would have to be team based, or restricted to 5v5 and 3v3 arranged teams with 5 or 3 players respectively with a clan tag to hold the win/loss ratio for the team, rather than the individual. that would mean 'free-lance.' summoners would not be able to hold a ranking. or it woudl be some agregate of the ranking for all of the teams they've played for

the soloqueue is mainly for practice. as the ranking system really only puts you near the rating you're suppose to be (within say 200).

there's also a mentality that there's somethign wrong with being unranked. the e-peen and e-gina people get too worked up. if anything you need to encourage people that being unranked isn't a bad thing. it just is what it is. yes, there are problems. yes, there are plenty of people that are under-ranked. but if you're really serious about the competitive scene: join a clan.


Comment below rating threshold, click here to show it.

DarkPsx

This user has referred a friend to League of Legends, click for more information

Senior Member

01-10-2011

It is oversimplified, because i never believed that someone needed to know the full meaning of ELO in order to improve their gameplay in that way. We do not need to dump a ton of statistics on top of the community to give them reason to believe that they have no hope - it is all player percieved. If anyone would really like to know how ELO works, feel free to look it up on Wikipedia! http://en.wikipedia.org/wiki/Elo_rating_system


Comment below rating threshold, click here to show it.

Falcedge

Junior Member

01-10-2011

I really like this thread, onisake and DarkPsx make some really good points. But aren't you guys over extending the ELO part of the game? it is after all a game first with competitve portion for those who want to take it further. Although you guys both make really good points, you should'nt argue about a game that's played for fun.


Comment below rating threshold, click here to show it.

Onisake

Senior Member

01-11-2011

Quote:
Originally Posted by Falcedge View Post
I really like this thread, onisake and DarkPsx make some really good points. But aren't you guys over extending the ELO part of the game? it is after all a game first with competitve portion for those who want to take it further. Although you guys both make really good points, you should'nt argue about a game that's played for fun.
for those that enjoy competitive play, it's a different kind of fun. if you dont 'get it, you're not ready for true competitive play.

like i said. League doesn't use ELO. it's just called ELO for some reason. i explained it more thoroughly to show that it isn't an ELO system. we just call it that for some reason. that part eludes me. but w/e

if you think the statistics are unimportant, and data is unimportant, you really, really aren't ready for competitive play. giving a simplified version, is again, something more for general discussion. not high level competitive play.

ESPECIALLY since your original topic was about how to improve your ELO. you can't withold data or information.

as the great lord kelvin once said: To measure is to know. if we can't measure it. it becomes unknown, and therefore beyond our control.

real ELO is measurable. everyone who is serious about chess or go, knows how ELO works. it's defined. Riot's system is undefined. likely because it is not robust, and easy to exploit. IE: their stats team isn't very good. there's nothing wrong with that. this is an f2p game. they dont' have the resources blizzard does, or any other larger company.

again, competitive play means you're ranked. if you're ranked, you already know everything you mentioned. your guide is more for people who are unranked. it doesn't really belong hear until you expand on it.


Comment below rating threshold, click here to show it.

TheColorRed

Junior Member

01-11-2011

I just read this entire argument between matchmaking and this is what I have concluded: Darkpsx has actually contributed something decent for the community to follow as a guideline for all players, while Onisake is seemlessly trolling Darkpsx and trying to overcomplicated the entire system. In all honesty, I have thought of this as a decent argument until Onisake has tried to argue against ELO. Yes Onisake, this is a modified ELO system. Not many institutions still implement ELO straight up, because there is a K factor for certain rankings as well as a type of soloment rank fluctuation, as well as the fact that logical distribution was implemented in place as a more reliable source of ranking. If you really understood the true meaning behind ELO, you would know it is more of a formality in respect to Arpad Elo, the person who formed the caluclations for such a rating. Most multiplayer games today use a modified version of the ELO system, which we commonly just refer to as the ELO system itself because it follows the basics of ELO.

For everyone in the community that is still hazed because of the argument that is going on in this thread, you only need to know a few things to clarify this big maze of information:

1. In the first post, Darkpsx gives the community a general guideline for players to follow in order to improve their knowledge of the game and therefore extend on a player's level of skill in a game. He is making a claim that good sportsmanship, a very wide knowledge of game mechanics, and the urgency to seek help from other peers is a key to improving your competitive gaming experience. This is completely logical, and the reason is because most players are content in the level they play and have little room for growth, mentally, without that extra kick that something needs to be changed.

As a counterargument, Onisake goes ahead and adds the fact that it's not just this little information that cuts it for competitive play - which is completely true - but goes on and tries to prove that certain carries are so groundbreaking that no matter what you have done to learn and improve your gameplay, it has little to no effect on the outcome of the game. To an extent, this can always be argued - in fact, there are many things in competitive play that can be argued, because to scale with player skill, a random variable can always be pushed to its max potential (figuratively) and the game is thrown off as "carried".

2. Darkpsx then tries to explain that the guide is not meant to account for sharp changes in random variables, and that his belief of improving your rating is simply to improve and foster a more efficient and meaningful playstyle. This is a logical approach to the post - the players who are highly rated are generally BETTER PLAYERS in general. Whether they are creative in their use of game mechanics or simply have a better team effort due to a better understanding of the game, they are higher rated than just a player who didn't know any better and never knew that a double cs lead means that they could be in trouble.

The counterargument by Onisake was that this was incomplete (which Darkpsx understood), and that the random variable pushed to its max potential is too much of a reliant factor in how rating fluctuates. This, on top of a random 4 members who either do their best to support the carry or be clueless and go on their own, are not so big of a factor as long as there are a few reckoning forces on the team that truly understand their game. This argument is also correct - but is less of a consistent argument. This is simply because the random variable (carry) on one team will not always be the dominant force in a game, whereas this random variable was pushed to its max potential, the other team may have a BAD carry but 4 EXCELLENT supports who are able to push the carry down. Therefore, we rely on 1 random carry and 4 players who can't seemingly figure out how to stop this carry in order to counter its effects.

In this case, we must keep our eye on consistency.

3. The last manner was Dark's attempt to simplify ELO and how it is roughly calculated and its purpose behind the way it acts, from the beginning of when you are first rated, to the end (rating change is then minimal plus your matchmaking can further reduce or increase your loss and gain per match). Honestly, the way it was analogized, this was a great example. It was put in terms that many players in the community can understand. This is where I feel Onisake has lost it completely, and its either trying way too hard to find an argument or is just trying to prove his knowledge in order to discredit this explanation of ELO. Fortunately for most of the community, you are hearing from me, a statitistic graduate that this is indeed a modified ELO system, and that it was explained rather well. Don't try to think about the system more than it is - the original intent of the post was simply to inform players that perfecting the basics can increase your chances of winning based on the bigger picture - the fact that a player is then increasing his skill potential, he has just promoted his potential to win in the bracket that he is in, where players may share similar skill.


Comment below rating threshold, click here to show it.

Onisake

Senior Member

01-11-2011

Quote:
Originally Posted by TheColorRed View Post
I just read this entire argument between matchmaking and this is what I have concluded: Darkpsx has actually contributed something decent for the community to follow as a guideline for all players, while Onisake is seemlessly trolling Darkpsx and trying to overcomplicated the entire system. In all honesty, I have thought of this as a decent argument until Onisake has tried to argue against ELO. Yes Onisake, this is a modified ELO system. Not many institutions still implement ELO straight up, because there is a K factor for certain rankings as well as a type of soloment rank fluctuation, as well as the fact that logical distribution was implemented in place as a more reliable source of ranking. If you really understood the true meaning behind ELO, you would know it is more of a formality in respect to Arpad Elo, the person who formed the caluclations for such a rating. Most multiplayer games today use a modified version of the ELO system, which we commonly just refer to as the ELO system itself because it follows the basics of ELO.

For everyone in the community that is still hazed because of the argument that is going on in this thread, you only need to know a few things to clarify this big maze of information:

1. In the first post, Darkpsx gives the community a general guideline for players to follow in order to improve their knowledge of the game and therefore extend on a player's level of skill in a game. He is making a claim that good sportsmanship, a very wide knowledge of game mechanics, and the urgency to seek help from other peers is a key to improving your competitive gaming experience. This is completely logical, and the reason is because most players are content in the level they play and have little room for growth, mentally, without that extra kick that something needs to be changed.

As a counterargument, Onisake goes ahead and adds the fact that it's not just this little information that cuts it for competitive play - which is completely true - but goes on and tries to prove that certain carries are so groundbreaking that no matter what you have done to learn and improve your gameplay, it has little to no effect on the outcome of the game. To an extent, this can always be argued - in fact, there are many things in competitive play that can be argued, because to scale with player skill, a random variable can always be pushed to its max potential (figuratively) and the game is thrown off as "carried".

2. Darkpsx then tries to explain that the guide is not meant to account for sharp changes in random variables, and that his belief of improving your rating is simply to improve and foster a more efficient and meaningful playstyle. This is a logical approach to the post - the players who are highly rated are generally BETTER PLAYERS in general. Whether they are creative in their use of game mechanics or simply have a better team effort due to a better understanding of the game, they are higher rated than just a player who didn't know any better and never knew that a double cs lead means that they could be in trouble.

The counterargument by Onisake was that this was incomplete (which Darkpsx understood), and that the random variable pushed to its max potential is too much of a reliant factor in how rating fluctuates. This, on top of a random 4 members who either do their best to support the carry or be clueless and go on their own, are not so big of a factor as long as there are a few reckoning forces on the team that truly understand their game. This argument is also correct - but is less of a consistent argument. This is simply because the random variable (carry) on one team will not always be the dominant force in a game, whereas this random variable was pushed to its max potential, the other team may have a BAD carry but 4 EXCELLENT supports who are able to push the carry down. Therefore, we rely on 1 random carry and 4 players who can't seemingly figure out how to stop this carry in order to counter its effects.

In this case, we must keep our eye on consistency.

3. The last manner was Dark's attempt to simplify ELO and how it is roughly calculated and its purpose behind the way it acts, from the beginning of when you are first rated, to the end (rating change is then minimal plus your matchmaking can further reduce or increase your loss and gain per match). Honestly, the way it was analogized, this was a great example. It was put in terms that many players in the community can understand. This is where I feel Onisake has lost it completely, and its either trying way too hard to find an argument or is just trying to prove his knowledge in order to discredit this explanation of ELO. Fortunately for most of the community, you are hearing from me, a statitistic graduate that this is indeed a modified ELO system, and that it was explained rather well. Don't try to think about the system more than it is - the original intent of the post was simply to inform players that perfecting the basics can increase your chances of winning based on the bigger picture - the fact that a player is then increasing his skill potential, he has just promoted his potential to win in the bracket that he is in, where players may share similar skill.
good summary. except you missed my point entirely: it doesnt' belong here. it belongs in general guides. if this is a huge eye opener for you, you didn't belong in competitive play. end of story. this is a 'how to improve your play in LoL if you have no idea how to play guide.' not a 'how to improve your ranking.' guide. some basics should be said at every level of play. sportsmanship, awareness of the playing feild.

DarkPSX is tryign to contribute something. i'm not faulting him for that. i'm faulting him for where it is and the simpliciticy of the content. anyone having real problems isn't going to get anything out of this. he's missing the fundamentals that make rank play different from normal games. all of his advice applies to normal games. there is nothing specific to ranked play. that's the biggest fault with this guide.

His ELO explanation is bad for several reasons. the big one, which you failed to notice mr statistician, i only indirectly pointed out: ELO is a distributive curve. it ranks everyone in the system relative to eachother. in his explanation it was possible for everyone in the class to get max rating. simply by getting 100% on everything. that's not likely, but it's possible. in a real ELO system. that's impossible. because you only go up by beating people, and that makes them go down. if his tests, quizes and HW were all head-to-head assignments where one person wins and one person loses, it woudl make more sense. what he explained is NOT how elo works. your university must not be acredited. i'm tempted to ask where, so i know where NOT to higher statisticians from. but that would be pompous. dont' call me a troll then completely miss something fundamental about a system. his explanation is terrible. (although i admit, i am trolling a bit) the only consistancy between what he said and how ELO works is if you're bad. you have a low rating. if you're good. you have a good rating.

DarkPSX needs to add more to this guide. there is nothing that explains the difference between draft pick and blind pick (the biggest reason people have terrible ratings). nothing to explain why some champs are better than others (how come you dont' see Soraka, Sona, or urgot in high level games). nothing to explain specific counter to specific problems newer players have (mechanics behind grevious wound, qss, and why DPS carries are better than caster carries). the guide is incomplete. if DarkPSX understands that, he should fix it. not blindly throw counter arguements at me and go off on a rant about how ELO works. i fell into that trap, and just started spouting information. it was irrelevant and not construtive to this topic. my bad.

the things he need to add are

an explanation of how draft picking works. and the strategies that shoudl be employed before the match even starts.

an explanation of certain key items you see in high level play that are often absent from low level play

the importance of a core build that branches in several directions

the fundamental differences between caster carries and DPS carries and why DPS > casters and why certain champs are consistently picked for higher level play and why some aren't.

you will never, ever get above 1200 ELO playing only Soraka. you could potentially get to 1600 playing only Ashe. there's a reason for that. that needs to be explained.


1234567 ... 8