So, apparently one of my friends has played this damn near sense the open beta. When he found out that i was getting into it he did something that he says he wishs somebody woulda done for him, and bought me $15 worth of the points. So i unlocked jax and gragas.
Can you gift riot points? I tried to do this for a friend, but couldn't figure out how.
Or did he buy you a gift card?
So, apparently one of my friends has played this damn near sense the open beta. When he found out that i was getting into it he did something that he says he wishs somebody woulda done for him, and bought me $15 worth of the points. So i unlocked jax and gragas.
Can you gift riot points? I tried to do this for a friend, but couldn't figure out how.
Or did he buy you a gift card?
Hooray yes you are playing tristana but you do no damage if you spend the entire teamfight backing up, deciding to stay instead, shooting once and then repeating
and then doing that when you should be strictly running after they roll over gragas
Edit: This was more a response to your previous statement, Feldorn, but the whole point of ELO is to match people of a particular skill together, regardless of the number of games they've played. While I could see something along the lines of well, we have 20 people here that are all of similar ELO, then lets try to match up their summoner ranks, because the ranks do provide a small advantage because of runes. But in general, the vast majority of matches (in my experience) in non 5v5 premades aren't decided by those extra <level difference> runes, and including them in the matchmaking formula would just add extra noise/redundancy that would weigh # of games played disproportionally to skill.
chrono_traveller on
The trouble with having an open mind, of course, is that people will insist on coming along and trying to put things in it. ~ Terry Pratchett
Edit: This was more a response to your previous statement, Feldorn, but the whole point of ELO is to match people of a particular skill together, regardless of the number of games they've played. While I could see something along the lines of well, we have 20 people here that are all of similar ELO, then lets try to match up their summoner ranks, because the ranks do provide a small advantage because of runes. But in general, the vast majority of matches (in my experience) in non 5v5 premades aren't decided by those extra <level difference> runes, and including them in the matchmaking formula would just add extra noise/redundancy that would weigh # of games played disproportionally to skill.
maybe not skill, per se, but at summoner level 30, they have logged more hours, and therefore, should have a better understanding about how to play.
i understand that it is based on elo, but elo is also not indicative of skill. what i want, is for the system to try harder to match ELO AND summoner level.
i consider myself a decent player. sometimes i got on a win streak, probably because i got put up against people with 10-17ish summoner levels, therefore, less experienced than myself. then i got on a lose streak, because i'm now on the team with said baddies.
i guess my point was that in many respects, IMHO, summoner level will be more indicative of skill than ELO. mostly because no one has any idea how the system comes up with the arbitrary ELO ratings.
/violence inherent in the system.
//being repressed.
///dennis.
Feldorn on
0
Options
Zen VulgarityWhat a lovely day for teaSecret British ThreadRegistered Userregular
ELO is an easy system to understand. You start at a number, as you win, you gain points on to that number, the larger the difference is, the larger the gain. If you lose, you lose points in the same way.
i understand that it is based on elo, but elo is also not indicative of skill. what i want, is for the system to try harder to match ELO AND summoner level.
It is certainly more indicative of skill than level, almost by definition. Level is just roughly number of games played. ELO gives you a score based on who you lost/won against.
So, my question to you is what does bringing summoner level into the equation along with ELO give you? I'd say, sure with more games played, your ELO/skill will be known to a greater confidence (your ELO will be more stable if, say you've played 200 games vs. 20), but I'm not clear that this would make matches any better. I could see something like "well, we have 20 players here of similar ELO, let then group them into two games based on summoner level", but I think this is a very corner case that probably is very rare (and probably wouldn't make all that much difference in the quality of the games).
chrono_traveller on
The trouble with having an open mind, of course, is that people will insist on coming along and trying to put things in it. ~ Terry Pratchett
well, part of my point earlier was that a poor player can be carried to high ELO.
the other part is that, if matchmaking (MM) is based on ELO, and ELO is based on skill (ELO is more based on wins than skill i feel), then how do you end up with matches that are such blowouts? when you have a good spread of champions, so you aren't outpicked, but still get stomped all over. everyone has bad games, i understand. but i rarely have a game that i feel was a good loss, or that we won because we executed better in teamfights.
many of my wins have been total stomps, where it may as well have been 2v5 because 3 of the other team couldn't see with their pants on their heads. it seems like this shouldn't happen if ELO is indeed a measure of skill, and also who players are matched up.
like i say, i don't want to be matched solely on level, but it would be nice to be matched based on ELO and have level taken into consideration.
ELO is an easy system to understand. You start at a number, as you win, you gain points on to that number, the larger the difference is, the larger the gain. If you lose, you lose points in the same way.
i get that you gain ELO for winning, lose ELO for losing. just not sure how it decides how much to give you.
the way i see it, i don't arbitrarily pick a game to join, they system does that for me, there should never ever be a large difference in ELO between the teams.
I could understand a level 30 and non-level 30 factor. I think that should be included.
I am sick of getting matched with people under level 30 and obviously under 100 games played against level 30's with 300 wins. It keeps fucking happening.
I'm trying to be patient with it, but it's getting quite annoying.
I limed the parts that i have been talking about for those who are arguing with me.
blah blah, the whole point of ELO is to match people of a particular skill together, regardless of the number of games they've played. blah.
i re-read some of this and picked out something that i don't think i have properly provided a rebuttal for. (not trying to be condescending here)
i fully understand the point of ELO, but with the current iteration of the matchmaking, i feel like it is a horrible failure. if i am indeed getting matched with and against people with similar ELO ratings, then i don't think it works.
ELO is a measure of performance, not skill, and thus takes level into consideration implicitly. A given ELO score could apply to a more skilled lower level player or a less skilled higher level player.
ELO is a measure of performance, not skill, and thus takes level into consideration implicitly. A given ELO score could apply to a more skilled lower level player or a less skilled higher level player.
well, my problem must be totally with the matchmaking algorithm then.
because in the games i get matched in, when i see someone who goes 1/9/3 on fiddle (due to tower diving and trying to solo the other team), and is summoner lvl 13, i can't help but think that he is not more skilled than the lvl 30 summoners in the game.
EDIT: Smurf Accounts. Guys who have 300 games played and a level 30 account, on a level 12 account with under 100 games played. They are probably winning more then they are losing and thus getting shot up in to higher ELO brackets based on there performance. They would get there faster then some one who has no idea how to play and starts at level 1.
ELO is a measure of performance, not skill, and thus takes level into consideration implicitly. A given ELO score could apply to a more skilled lower level player or a less skilled higher level player.
well, my problem must be totally with the matchmaking algorithm then.
because in the games i get matched in, when i see someone who goes 1/9/3 on fiddle (due to tower diving and trying to solo the other team), and is summoner lvl 13, i can't help but think that he is not more skilled than the lvl 30 summoners in the game.
The reality of matchmaking is that there are only a finite number of players in the queue at any given time. Inevitably, you'll end up with some games that are not ideal.
the other part is that, if matchmaking (MM) is based on ELO, and ELO is based on skill (ELO is more based on wins than skill i feel), then how do you end up with matches that are such blowouts? when you have a good spread of champions, so you aren't outpicked, but still get stomped all over. everyone has bad games, i understand. but i rarely have a game that i feel was a good loss, or that we won because we executed better in teamfights.
many of my wins have been total stomps, where it may as well have been 2v5 because 3 of the other team couldn't see with their pants on their heads. it seems like this shouldn't happen if ELO is indeed a measure of skill, and also who players are matched up.
like i say, i don't want to be matched solely on level, but it would be nice to be matched based on ELO and have level taken into consideration.
I'd say part of the problem is how (especially before the patch) one player who feeds for whatever reason (He's trying out a new champion, different build, had a bad day, doesn't like your haircut) tends to lead to snowballing because they will get ganked mercilessly by a competent lane opponent. And there's really no matchmaking system to account for this.
I mean, there's no way to avoid having stomps. Two NFL teams can come into a game with similar records against similar opponents, both healthy, fully scouted each other out, and still it can turn into a rout if the ball bounces a certain way. And this is with a game thats been around a hundred years. How good can you do to match people when they're blind picking, have very limited communication with each other, and have never worked together.
I'm not saying that they're doing things great or that it couldn't be improved, but I haven't seen any argument that "somehow" including level would make better matches than "I was in some games where we had this low level guy that totally lost the game for us".
chrono_traveller on
The trouble with having an open mind, of course, is that people will insist on coming along and trying to put things in it. ~ Terry Pratchett
The problem with Elo for a game like this is that it's a team based game. It works for teams that have consistent members but it doesn't work when teams are a mishmash of people who got their previous Elo +/- from other mishmashes of people.
Elo works better with less variables to account for (summoner level, runes, champion balance, blind pick, etc. etc. etc.) so right now it doesn't work exactly as it's meant to IMO.
It does work to seperate low/mid/top tier FOR THE MOST PART, though.
ELO is a measure of performance, not skill, and thus takes level into consideration implicitly. A given ELO score could apply to a more skilled lower level player or a less skilled higher level player.
well, my problem must be totally with the matchmaking algorithm then.
because in the games i get matched in, when i see someone who goes 1/9/3 on fiddle (due to tower diving and trying to solo the other team), and is summoner lvl 13, i can't help but think that he is not more skilled than the lvl 30 summoners in the game.
It could just as easily been him having a bad day, or maybe he's used to playing Tryndamere, etc. But I've seen my share of level 20+ players being just as idiotic.
I had a level 30 Gangplank last week who adamantly decreed that while there was a golem or lizard up, he was not going to enter into a team fight. All we would get was his ult.
And this is what my point is. We could trade stories back and forth about how level X player lost the game by being stupid, but thats all it is. It doesn't indicate anything.
chrono_traveller on
The trouble with having an open mind, of course, is that people will insist on coming along and trying to put things in it. ~ Terry Pratchett
The problem with Elo for a game like this is that it's a team based game. It works for teams that have consistent members but it doesn't work when teams are a mishmash of people who got their previous Elo +/- from other mishmashes of people.
Elo works better with less variables to account for (summoner level, runes, champion balance, blind pick, etc. etc. etc.) so right now it doesn't work exactly as it's meant to IMO.
It does work to seperate low/mid/top tier FOR THE MOST PART, though.
I'd agree with this. And really for random solo queue games, I don't think you're going to get any better.
chrono_traveller on
The trouble with having an open mind, of course, is that people will insist on coming along and trying to put things in it. ~ Terry Pratchett
I'll just put in that my W:L ratio jumped drastically after gaining access to T3 runes and better mastery setups. All they'd have to do is make one more bracket between the newbie island cutoff and level 20.
I'll just put in that my W:L ratio jumped drastically after gaining access to T3 runes and better mastery setups. All they'd have to do is make one more bracket between the newbie island cutoff and level 20.
Mine hasn't been affected all that much, as far as I can remember, though I don't remember when exactly I hit level 20 (how many games played do you have?). I remember my 100th game, I was at something like 55:45 (1.22), at 100 losses I was at 130:100 (1.3), and I'm currently at 170:130 (1.31).
Edit: Though, these numbers are somewhat unreliable for a number of reasons. I've done a fair mix of practice/solo queue/premade games, so its pretty noisy data. Also, I think lower level players tend to experiment more with different characters (I know I did), thus even when you get a competent lower level player, he may try experimenting with a new champ, which could lead to less desirable results. I think this would be alleviated by the ranked game option that they've been working on more than adjusting how to estimate player's skill.
chrono_traveller on
The trouble with having an open mind, of course, is that people will insist on coming along and trying to put things in it. ~ Terry Pratchett
Might have something to do with the fact that I came back from beta a week and a half after the game had launched, and before Newbie island even existed (I think). I was getting reamed by level 30s that had preordered the game (and exp bonus) from level 1 onwards.
Might have something to do with the fact that I came back from beta a week and a half after the game had launched, and before Newbie island even existed (I think). I was getting reamed by level 30s that had preordered the game (and exp bonus) from level 1 onwards.
I don't think that raw W:L ratios help much with analysis. If we could maybe have 100 or 1000 histories of, say, solo queuers W:L and their associated level, it'd be interesting to see if there is a appreciable level bump.
chrono_traveller on
The trouble with having an open mind, of course, is that people will insist on coming along and trying to put things in it. ~ Terry Pratchett
0
Options
Zen VulgarityWhat a lovely day for teaSecret British ThreadRegistered Userregular
edited March 2010
Feldorn is not part of the inside jokes. :whistle:
Posts
It's like a riskier blitz grab except you can go through minions and the cooldown is longer.
also a good one can also put the other enemies further and just bring in one or two so it works so beautifully
I don't think I'll ever be able to build up like that again.
Great game all around, battles were decided in five seconds with all the aoe.
Yeah, Zilean's good support in the right hands though.
Check my last game:
Look at those creep kills. w00!
also, getting back into the swing with corki, has some good wins against players who were not nearly as experienced last night.
really, my team was, 30, 30, 30, 30, 16, and you could tell really easy who the 16 was looking at scores.
enemy team was, 30, 30, 11, 14, 16, and they didn't stand a chance.
i'm still totally unsure why MM seems to totally disregard summoner level.
Can you gift riot points? I tried to do this for a friend, but couldn't figure out how.
Or did he buy you a gift card?
He got me a giftcard.
a fat gragas on a team with no DPS
Hooray yes you are playing tristana but you do no damage if you spend the entire teamfight backing up, deciding to stay instead, shooting once and then repeating
and then doing that when you should be strictly running after they roll over gragas
i know, and as many say, it is reterrrrded.
Because summoner rank isn't indicative of skill?
Edit: This was more a response to your previous statement, Feldorn, but the whole point of ELO is to match people of a particular skill together, regardless of the number of games they've played. While I could see something along the lines of well, we have 20 people here that are all of similar ELO, then lets try to match up their summoner ranks, because the ranks do provide a small advantage because of runes. But in general, the vast majority of matches (in my experience) in non 5v5 premades aren't decided by those extra <level difference> runes, and including them in the matchmaking formula would just add extra noise/redundancy that would weigh # of games played disproportionally to skill.
I think you're thinking of Cho'gath or Nunu, not Jax.
maybe not skill, per se, but at summoner level 30, they have logged more hours, and therefore, should have a better understanding about how to play.
i understand that it is based on elo, but elo is also not indicative of skill. what i want, is for the system to try harder to match ELO AND summoner level.
i consider myself a decent player. sometimes i got on a win streak, probably because i got put up against people with 10-17ish summoner levels, therefore, less experienced than myself. then i got on a lose streak, because i'm now on the team with said baddies.
i guess my point was that in many respects, IMHO, summoner level will be more indicative of skill than ELO. mostly because no one has any idea how the system comes up with the arbitrary ELO ratings.
/violence inherent in the system.
//being repressed.
///dennis.
Reagan has 300+ wins and is rank 30
He's terrible.
It is certainly more indicative of skill than level, almost by definition. Level is just roughly number of games played. ELO gives you a score based on who you lost/won against.
So, my question to you is what does bringing summoner level into the equation along with ELO give you? I'd say, sure with more games played, your ELO/skill will be known to a greater confidence (your ELO will be more stable if, say you've played 200 games vs. 20), but I'm not clear that this would make matches any better. I could see something like "well, we have 20 players here of similar ELO, let then group them into two games based on summoner level", but I think this is a very corner case that probably is very rare (and probably wouldn't make all that much difference in the quality of the games).
the other part is that, if matchmaking (MM) is based on ELO, and ELO is based on skill (ELO is more based on wins than skill i feel), then how do you end up with matches that are such blowouts? when you have a good spread of champions, so you aren't outpicked, but still get stomped all over. everyone has bad games, i understand. but i rarely have a game that i feel was a good loss, or that we won because we executed better in teamfights.
many of my wins have been total stomps, where it may as well have been 2v5 because 3 of the other team couldn't see with their pants on their heads. it seems like this shouldn't happen if ELO is indeed a measure of skill, and also who players are matched up.
like i say, i don't want to be matched solely on level, but it would be nice to be matched based on ELO and have level taken into consideration.
also, stop picking on Reagan.
i get that you gain ELO for winning, lose ELO for losing. just not sure how it decides how much to give you.
the way i see it, i don't arbitrarily pick a game to join, they system does that for me, there should never ever be a large difference in ELO between the teams.
I am sick of getting matched with people under level 30 and obviously under 100 games played against level 30's with 300 wins.
I'm trying to be patient with it, but it's getting quite annoying.
I limed the parts that i have been talking about for those who are arguing with me.
i re-read some of this and picked out something that i don't think i have properly provided a rebuttal for. (not trying to be condescending here)
i fully understand the point of ELO, but with the current iteration of the matchmaking, i feel like it is a horrible failure. if i am indeed getting matched with and against people with similar ELO ratings, then i don't think it works.
well, my problem must be totally with the matchmaking algorithm then.
because in the games i get matched in, when i see someone who goes 1/9/3 on fiddle (due to tower diving and trying to solo the other team), and is summoner lvl 13, i can't help but think that he is not more skilled than the lvl 30 summoners in the game.
EDIT: Smurf Accounts. Guys who have 300 games played and a level 30 account, on a level 12 account with under 100 games played. They are probably winning more then they are losing and thus getting shot up in to higher ELO brackets based on there performance. They would get there faster then some one who has no idea how to play and starts at level 1.
The reality of matchmaking is that there are only a finite number of players in the queue at any given time. Inevitably, you'll end up with some games that are not ideal.
What's your W/L record?
I'd say part of the problem is how (especially before the patch) one player who feeds for whatever reason (He's trying out a new champion, different build, had a bad day, doesn't like your haircut) tends to lead to snowballing because they will get ganked mercilessly by a competent lane opponent. And there's really no matchmaking system to account for this.
I mean, there's no way to avoid having stomps. Two NFL teams can come into a game with similar records against similar opponents, both healthy, fully scouted each other out, and still it can turn into a rout if the ball bounces a certain way. And this is with a game thats been around a hundred years. How good can you do to match people when they're blind picking, have very limited communication with each other, and have never worked together.
I'm not saying that they're doing things great or that it couldn't be improved, but I haven't seen any argument that "somehow" including level would make better matches than "I was in some games where we had this low level guy that totally lost the game for us".
Elo works better with less variables to account for (summoner level, runes, champion balance, blind pick, etc. etc. etc.) so right now it doesn't work exactly as it's meant to IMO.
It does work to seperate low/mid/top tier FOR THE MOST PART, though.
It could just as easily been him having a bad day, or maybe he's used to playing Tryndamere, etc. But I've seen my share of level 20+ players being just as idiotic.
I had a level 30 Gangplank last week who adamantly decreed that while there was a golem or lizard up, he was not going to enter into a team fight. All we would get was his ult.
And this is what my point is. We could trade stories back and forth about how level X player lost the game by being stupid, but thats all it is. It doesn't indicate anything.
I'd agree with this. And really for random solo queue games, I don't think you're going to get any better.
Mine hasn't been affected all that much, as far as I can remember, though I don't remember when exactly I hit level 20 (how many games played do you have?). I remember my 100th game, I was at something like 55:45 (1.22), at 100 losses I was at 130:100 (1.3), and I'm currently at 170:130 (1.31).
Edit: Though, these numbers are somewhat unreliable for a number of reasons. I've done a fair mix of practice/solo queue/premade games, so its pretty noisy data. Also, I think lower level players tend to experiment more with different characters (I know I did), thus even when you get a competent lower level player, he may try experimenting with a new champ, which could lead to less desirable results. I think this would be alleviated by the ranked game option that they've been working on more than adjusting how to estimate player's skill.
Mooman and Zen are also terrible.
I don't think that raw W:L ratios help much with analysis. If we could maybe have 100 or 1000 histories of, say, solo queuers W:L and their associated level, it'd be interesting to see if there is a appreciable level bump.
Zomg! Zen actually used a period! ;-)
Wait, what inside jokes?
:whistle: