|
On December 26 2011 19:37 wooscums wrote:Show nested quote +
At the end of a season, an active player's MMR will be carried over to the next. If they do not play any matches in the current season, however, and are not placed, the MMR will not carry over into a third season, effectively starting them fresh in the system.
Hello I'm a NA master protoss but for the past 4-5 months I have been studying abroad in Vietnam, so I haven't really been able to play ladder at all (internet here is really bad). Anyway, going from the statement above, how does "fresh" MMR work because since I've been here, 2 seasons have passed without me playing a ladder game and according to the OP, my MMR will not carry over. When I come back on Dec. 31st and play my first ladder game in 2 seasons, will I have to play 5 placement matches again? or will 1 just determine my new MMR since my old one is not carried over? thanks!
Try it and report to us!
EDIT: I haven't played for 2 seasons as well, and it appears that I need 5 games for my placement - most likely showing that the MMR is reset.
|
OP, thanks for this thread. Very clear and useful information! I used to think that Battle.net monitored in-game statistics like APM, supply over time, and unspent resources. ;P
|
On December 26 2011 19:27 Azzur wrote:Show nested quote +On December 26 2011 18:57 Macpo wrote:I have a question, sorry if it has been answered before, but i couldn't find it in the OP. It's about the meaning of being in the higher or lower tier of a league. I basically have the impression it does not necessarily mean that you are good or bad; rather that your are getting better or worse (or to tell it differently, that MMR and ranking are two very different indicators, ranking is not a proxy of MMR). For instance, let's assume that a player A has a MMR at a value of X in master. he wins ten games and loses ten. he therefore has a number of points P (around the bonus pool he got) . Now, let's assume a player B in masters, who has a MMR of X+ Y. He is better than player B, proportionnally to Y. let's assume that he also wins ten games and loses ten. As he is paired with better players,there is no reason for him to win more games than a lower player, so this is very likely to happen. His number of points P' therefore is very close to P, and in the league, the rankings of players A and B are very similar ( For instance, they are both 40th or so in their league). So, player B could be much better than player A (actually the value of Y doesn't matter at all at first sight), but due to the fact that he plays against better players, they can be ranked at the same level. Rank indicates that you are getting better or worse, not that you are good or bad. The only situation in which I would see this not work is at the very top and bottom of player pool, as then you cannot be paired against equivalent ranked players, and therefore being good (in case of high master) or bad (in case of lower bronze) implies that you have a win/loss ratio tendancially different from 50% on the long term, and therefore in itself implies a position at the top or bottom of your league. Am I right? Another mystery to me, maybe linked to the first one: How can someone with 10 win 18 losses or 19 win 32 losses be ahead in rankings of someone who is 18 wins and 11 losses? I am curious  (and this very fact kind of seems to invalidate the previous analyses: 2 identical win loss ratios don't imply necessarily identical rankings) Thanks I bolded the part where you have a mis-understanding of the ladder system. If player B has a high MMR, at the start of the season, since the number of points has not converged to MMR yet, the amount won/loss per game will be disproportionate compared to player A (i.e. player B will get more per win, and lose less per loss). As for the rank in their division, that is quite arbitrary - however, it is known that Master league is not tiered, so you can compare their points for who is better. Basically, win/loss is not a good measure of skill (unless the player is very high master or in GM). As an aside, you'll see many players who spout nonsense about their good win ratios, when someone who is far better than them (but not pro-level) can have negative win ratios. Also, if you're interested, it has been shown that these players who are just below pro-level will frequently start off with negative win/loss records. As for your question: Show nested quote +Another mystery to me, maybe linked to the first one: How can someone with 10 win 18 losses or 19 win 32 losses be ahead in rankings of someone who is 18 wins and 11 losses? I am curious (and this very fact kind of seems to invalidate the previous analyses: 2 identical win loss ratios don't imply necessarily identical rankings) As explained above, someone with 10-18 could have played pro's and hence they received alot of points for each win and don't lose many for each loss. In the OP, the number of points won/loss for game is determined by a player's current points vs opponent MMR.
Thanks for your answer
Still there is a few things which are unclear to me^^.
1. I can understand that you received a lot of points for win and lose very little for losses when you are paired against (much) better players. Then the issue for me is : why would you be paired against better opponents for 50 games, if you lose about 66% of them? I am sure there is a reason, but I can't really see it. I guess it has to do with the sentence I bolded in your explanation, which I don't get ("since the number of points has not converged to MMR yet"). I don't see why number of points would converge to MMR (and be good proxy indicator).
2. I very much understand that win/loss ratio is not a good indicator, as two very different players can have very similar ratios, and reversely (ratio, especially in the beginning of a season, indicating more the variation of the level, than the level itself). But then I don't get why points would be a better indicator, as points are closely related to win/loss ratios in the long term, as you are tendancially paired with players of your MMR, and therefore have an average 12 points for win, and 12 points for losses (but this brings back to point 1). Basically you can make a few extra points here and there when getting better, because you play against (slightly) favored opponents; but being good in general doesn't imply that you get more points for your win, and therefore will not be reflected in your points. Hence, having more points means that you are getting better, not that you are good.
Obviously I am making some mistake here but I am just not sure where it is!
thanks
|
On December 26 2011 19:59 Macpo wrote:Show nested quote +On December 26 2011 19:27 Azzur wrote:On December 26 2011 18:57 Macpo wrote:I have a question, sorry if it has been answered before, but i couldn't find it in the OP. It's about the meaning of being in the higher or lower tier of a league. I basically have the impression it does not necessarily mean that you are good or bad; rather that your are getting better or worse (or to tell it differently, that MMR and ranking are two very different indicators, ranking is not a proxy of MMR). For instance, let's assume that a player A has a MMR at a value of X in master. he wins ten games and loses ten. he therefore has a number of points P (around the bonus pool he got) . Now, let's assume a player B in masters, who has a MMR of X+ Y. He is better than player B, proportionnally to Y. let's assume that he also wins ten games and loses ten. As he is paired with better players,there is no reason for him to win more games than a lower player, so this is very likely to happen. His number of points P' therefore is very close to P, and in the league, the rankings of players A and B are very similar ( For instance, they are both 40th or so in their league). So, player B could be much better than player A (actually the value of Y doesn't matter at all at first sight), but due to the fact that he plays against better players, they can be ranked at the same level. Rank indicates that you are getting better or worse, not that you are good or bad. The only situation in which I would see this not work is at the very top and bottom of player pool, as then you cannot be paired against equivalent ranked players, and therefore being good (in case of high master) or bad (in case of lower bronze) implies that you have a win/loss ratio tendancially different from 50% on the long term, and therefore in itself implies a position at the top or bottom of your league. Am I right? Another mystery to me, maybe linked to the first one: How can someone with 10 win 18 losses or 19 win 32 losses be ahead in rankings of someone who is 18 wins and 11 losses? I am curious (and this very fact kind of seems to invalidate the previous analyses: 2 identical win loss ratios don't imply necessarily identical rankings) Thanks I bolded the part where you have a mis-understanding of the ladder system. If player B has a high MMR, at the start of the season, since the number of points has not converged to MMR yet, the amount won/loss per game will be disproportionate compared to player A (i.e. player B will get more per win, and lose less per loss). As for the rank in their division, that is quite arbitrary - however, it is known that Master league is not tiered, so you can compare their points for who is better. Basically, win/loss is not a good measure of skill (unless the player is very high master or in GM). As an aside, you'll see many players who spout nonsense about their good win ratios, when someone who is far better than them (but not pro-level) can have negative win ratios. Also, if you're interested, it has been shown that these players who are just below pro-level will frequently start off with negative win/loss records. As for your question: Another mystery to me, maybe linked to the first one: How can someone with 10 win 18 losses or 19 win 32 losses be ahead in rankings of someone who is 18 wins and 11 losses? I am curious (and this very fact kind of seems to invalidate the previous analyses: 2 identical win loss ratios don't imply necessarily identical rankings) As explained above, someone with 10-18 could have played pro's and hence they received alot of points for each win and don't lose many for each loss. In the OP, the number of points won/loss for game is determined by a player's current points vs opponent MMR. Thanks for your answer Still there is a few things which are unclear to me^^. 1. I can understand that you received a lot of points for win and lose very little for losses when you are paired against (much) better players. Then the issue for me is : why would you be paired against better opponents for 50 games, if you lose about 66% of them? I am sure there is a reason, but I can't really see it at the moment. I guess it has to do with the sentence I bolded in your explanation, which I don't get ("since the number of points has not converged to MMR yet"). I don't see why number of points would converge to MMR (and be good proxy indicator). 2. I very much understand that win/loss ratio is not a good indicator, as two very different players can have very similar ratios, and reversely (ratio, especially in the beginning of a season, indicating more the variation of the level, than the level itself). But then I don't get why points would be a better indicator, as points are closely related to win/loss ratios in the long term, as you are tendancially paired with players of your MMR, and therefore have an average 12 points for win, and 12 points for losses (but this brings back to point 1). Basically you can make a few extra points here and there when getting better, because you play against (slightly) favored opponents; but being good in general doesn't imply that you get more points for your win, and therefore will not be reflected in your points. Obviously I am making some mistake here but I am just not sure where it is! thanks
First, you must consider short term and long term.
Short term
- In the start of the season, the player is reset to 0 points but their MMR is leftover from the previous season.
- Thus, because of this disparity (points vs MMR), a high MMR player will receive alot for a win, but lose little for a loss. For the technical explanation of this, MMR is used for matchmaking but the win/loss of points for a game is determined by points against opponent MMR. Since there is a disparity between points and MMR, in the short-term, a high MMR player will receive alot for a win and lose little for a loss.
- Because of this, if the amount of bonus pool = 0, points is "only" a decent indicator. Until it stabilises, it will not be totally accurate.
- In the short term, it is not unusual for a player to have a lop-sided win/loss (e.g. 10-18 or 18-10). Whether you have a positive or negative one is dependent on your form or luck. If you're interested in the specifics of this, it has been shown that players near the top (but not pro) will tend to have negatives ratios. This is because they will likely be matched with pros and will lose disproportionately. However, don't think too much about this as it may confuse you - just consider what is not in italics.
Long term
- In the long run, the player's win/loss will converge to 50% (unless high master or GM). At this point in time, the points will converge to MMR.
- And yes, on average, every game after the stabilisation will have on average 12 per win and 12 points for loss. Note that this average only occurs after the points converges to MMR. Before the convergence, this +12/-12 does not occur. The "extra points" generated is determined by how high the MMR originally is.
- As an aside, you will find many ppl saying that the matchmaking system will "force" a player to a 50% win/loss ratio. This is incorrect - the matchmaking system attempts to find the best possible opponent in the next game. What happens is that the win/loss will converge to 50% rather than being "forced" to 50%.
Also, to answer your questions:
1. You will not necessarily be paired against better opponents. Rather, it's not unusual to have a lop-sided win/loss in the short term. 10-18 is just as likely as 18-10, it really depends on your form and luck. However, eventually, the win/loss will converge to 50%.
2. I answered this above, but the average 12 points for win/loss only occurs when the points stabilises with MMR and not before. The "extra points" that you are referring to determined by how high the MMR is initially.
|
On December 26 2011 20:14 Azzur wrote:Show nested quote +On December 26 2011 19:59 Macpo wrote:On December 26 2011 19:27 Azzur wrote:On December 26 2011 18:57 Macpo wrote:I have a question, sorry if it has been answered before, but i couldn't find it in the OP. It's about the meaning of being in the higher or lower tier of a league. I basically have the impression it does not necessarily mean that you are good or bad; rather that your are getting better or worse (or to tell it differently, that MMR and ranking are two very different indicators, ranking is not a proxy of MMR). For instance, let's assume that a player A has a MMR at a value of X in master. he wins ten games and loses ten. he therefore has a number of points P (around the bonus pool he got) . Now, let's assume a player B in masters, who has a MMR of X+ Y. He is better than player B, proportionnally to Y. let's assume that he also wins ten games and loses ten. As he is paired with better players,there is no reason for him to win more games than a lower player, so this is very likely to happen. His number of points P' therefore is very close to P, and in the league, the rankings of players A and B are very similar ( For instance, they are both 40th or so in their league). So, player B could be much better than player A (actually the value of Y doesn't matter at all at first sight), but due to the fact that he plays against better players, they can be ranked at the same level. Rank indicates that you are getting better or worse, not that you are good or bad. The only situation in which I would see this not work is at the very top and bottom of player pool, as then you cannot be paired against equivalent ranked players, and therefore being good (in case of high master) or bad (in case of lower bronze) implies that you have a win/loss ratio tendancially different from 50% on the long term, and therefore in itself implies a position at the top or bottom of your league. Am I right? Another mystery to me, maybe linked to the first one: How can someone with 10 win 18 losses or 19 win 32 losses be ahead in rankings of someone who is 18 wins and 11 losses? I am curious (and this very fact kind of seems to invalidate the previous analyses: 2 identical win loss ratios don't imply necessarily identical rankings) Thanks I bolded the part where you have a mis-understanding of the ladder system. If player B has a high MMR, at the start of the season, since the number of points has not converged to MMR yet, the amount won/loss per game will be disproportionate compared to player A (i.e. player B will get more per win, and lose less per loss). As for the rank in their division, that is quite arbitrary - however, it is known that Master league is not tiered, so you can compare their points for who is better. Basically, win/loss is not a good measure of skill (unless the player is very high master or in GM). As an aside, you'll see many players who spout nonsense about their good win ratios, when someone who is far better than them (but not pro-level) can have negative win ratios. Also, if you're interested, it has been shown that these players who are just below pro-level will frequently start off with negative win/loss records. As for your question: Another mystery to me, maybe linked to the first one: How can someone with 10 win 18 losses or 19 win 32 losses be ahead in rankings of someone who is 18 wins and 11 losses? I am curious (and this very fact kind of seems to invalidate the previous analyses: 2 identical win loss ratios don't imply necessarily identical rankings) As explained above, someone with 10-18 could have played pro's and hence they received alot of points for each win and don't lose many for each loss. In the OP, the number of points won/loss for game is determined by a player's current points vs opponent MMR. Thanks for your answer Still there is a few things which are unclear to me^^. 1. I can understand that you received a lot of points for win and lose very little for losses when you are paired against (much) better players. Then the issue for me is : why would you be paired against better opponents for 50 games, if you lose about 66% of them? I am sure there is a reason, but I can't really see it at the moment. I guess it has to do with the sentence I bolded in your explanation, which I don't get ("since the number of points has not converged to MMR yet"). I don't see why number of points would converge to MMR (and be good proxy indicator). 2. I very much understand that win/loss ratio is not a good indicator, as two very different players can have very similar ratios, and reversely (ratio, especially in the beginning of a season, indicating more the variation of the level, than the level itself). But then I don't get why points would be a better indicator, as points are closely related to win/loss ratios in the long term, as you are tendancially paired with players of your MMR, and therefore have an average 12 points for win, and 12 points for losses (but this brings back to point 1). Basically you can make a few extra points here and there when getting better, because you play against (slightly) favored opponents; but being good in general doesn't imply that you get more points for your win, and therefore will not be reflected in your points. Obviously I am making some mistake here but I am just not sure where it is! thanks First, you must consider short term and long term. Short term- In the start of the season, the player is reset to 0 points but their MMR is leftover from the previous season. - Thus, because of this disparity (points vs MMR), a high MMR player will receive alot for a win, but lose little for a loss. For the technical explanation of this, MMR is used for matchmaking but the win/loss of points for a game is determined by points against opponent MMR. Since there is a disparity between points and MMR, in the short-term, a high MMR player will receive alot for a win and lose little for a loss. - Because of this, if the amount of bonus pool = 0, points is "only" a decent indicator. Until it stabilises, it will not be totally accurate. - In the short term, it is not unusual for a player to have a lop-sided win/loss (e.g. 10-18 or 18-10). Whether you have a positive or negative one is dependent on your form or luck. If you're interested in the specifics of this, it has been shown that players near the top (but not pro) will tend to have negatives ratios. This is because they will likely be matched with pros and will lose disproportionately. However, don't think too much about this as it may confuse you - just consider what is not in italics.Long term- In the long run, the player's win/loss will converge to 50% (unless high master or GM). At this point in time, the points will converge to MMR. - And yes, on average, every game after the stabilisation will have on average 12 per win and 12 points for loss. Note that this average only occurs after the points converges to MMR. Before the convergence, this +12/-12 does not occur. The "extra points" generated is determined by how high the MMR originally is. - As an aside, you will find many ppl saying that the matchmaking system will "force" a player to a 50% win/loss ratio. This is incorrect - the matchmaking system attempts to find the best possible opponent in the next game. What happens is that the win/loss will converge to 50% rather than being "forced" to 50%. Also, to answer your questions: 1. You will not necessarily be paired against better opponents. Rather, it's not unusual to have a lop-sided win/loss in the short term. 10-18 is just as likely as 18-10, it really depends on your form and luck. However, eventually, the win/loss will converge to 50%. 2. I answered this above, but the average 12 points for win/loss only occurs when the points stabilises with MMR and not before. The "extra points" that you are referring to determined by how high the MMR is initially.
Thanks a lot for detailed explanation! I guess the point where I was not clear is the fact that points for a win are not determined by the difference between your MMR and your opponent's ; but as you said "the win/loss of points for a game is determined by points against opponent MMR".
Just to make it clear, does this mean that if you have only a few cumulated points and are playing against high MMR, you make a lot of points for win and little for losses, even if you are yourself to a similar high MMR; but if you have many points and are playing against high MMR, you make a more average number of points? If I am correct, that means that high MMR players are kind of guaranteed to make a lot of points in the beginning of the season (which sounds completely logical and fair).
Thanks again for the answers
|
On December 26 2011 20:40 Macpo wrote:Show nested quote +On December 26 2011 20:14 Azzur wrote:On December 26 2011 19:59 Macpo wrote:On December 26 2011 19:27 Azzur wrote:On December 26 2011 18:57 Macpo wrote:I have a question, sorry if it has been answered before, but i couldn't find it in the OP. It's about the meaning of being in the higher or lower tier of a league. I basically have the impression it does not necessarily mean that you are good or bad; rather that your are getting better or worse (or to tell it differently, that MMR and ranking are two very different indicators, ranking is not a proxy of MMR). For instance, let's assume that a player A has a MMR at a value of X in master. he wins ten games and loses ten. he therefore has a number of points P (around the bonus pool he got) . Now, let's assume a player B in masters, who has a MMR of X+ Y. He is better than player B, proportionnally to Y. let's assume that he also wins ten games and loses ten. As he is paired with better players,there is no reason for him to win more games than a lower player, so this is very likely to happen. His number of points P' therefore is very close to P, and in the league, the rankings of players A and B are very similar ( For instance, they are both 40th or so in their league). So, player B could be much better than player A (actually the value of Y doesn't matter at all at first sight), but due to the fact that he plays against better players, they can be ranked at the same level. Rank indicates that you are getting better or worse, not that you are good or bad. The only situation in which I would see this not work is at the very top and bottom of player pool, as then you cannot be paired against equivalent ranked players, and therefore being good (in case of high master) or bad (in case of lower bronze) implies that you have a win/loss ratio tendancially different from 50% on the long term, and therefore in itself implies a position at the top or bottom of your league. Am I right? Another mystery to me, maybe linked to the first one: How can someone with 10 win 18 losses or 19 win 32 losses be ahead in rankings of someone who is 18 wins and 11 losses? I am curious (and this very fact kind of seems to invalidate the previous analyses: 2 identical win loss ratios don't imply necessarily identical rankings) Thanks I bolded the part where you have a mis-understanding of the ladder system. If player B has a high MMR, at the start of the season, since the number of points has not converged to MMR yet, the amount won/loss per game will be disproportionate compared to player A (i.e. player B will get more per win, and lose less per loss). As for the rank in their division, that is quite arbitrary - however, it is known that Master league is not tiered, so you can compare their points for who is better. Basically, win/loss is not a good measure of skill (unless the player is very high master or in GM). As an aside, you'll see many players who spout nonsense about their good win ratios, when someone who is far better than them (but not pro-level) can have negative win ratios. Also, if you're interested, it has been shown that these players who are just below pro-level will frequently start off with negative win/loss records. As for your question: Another mystery to me, maybe linked to the first one: How can someone with 10 win 18 losses or 19 win 32 losses be ahead in rankings of someone who is 18 wins and 11 losses? I am curious (and this very fact kind of seems to invalidate the previous analyses: 2 identical win loss ratios don't imply necessarily identical rankings) As explained above, someone with 10-18 could have played pro's and hence they received alot of points for each win and don't lose many for each loss. In the OP, the number of points won/loss for game is determined by a player's current points vs opponent MMR. Thanks for your answer Still there is a few things which are unclear to me^^. 1. I can understand that you received a lot of points for win and lose very little for losses when you are paired against (much) better players. Then the issue for me is : why would you be paired against better opponents for 50 games, if you lose about 66% of them? I am sure there is a reason, but I can't really see it at the moment. I guess it has to do with the sentence I bolded in your explanation, which I don't get ("since the number of points has not converged to MMR yet"). I don't see why number of points would converge to MMR (and be good proxy indicator). 2. I very much understand that win/loss ratio is not a good indicator, as two very different players can have very similar ratios, and reversely (ratio, especially in the beginning of a season, indicating more the variation of the level, than the level itself). But then I don't get why points would be a better indicator, as points are closely related to win/loss ratios in the long term, as you are tendancially paired with players of your MMR, and therefore have an average 12 points for win, and 12 points for losses (but this brings back to point 1). Basically you can make a few extra points here and there when getting better, because you play against (slightly) favored opponents; but being good in general doesn't imply that you get more points for your win, and therefore will not be reflected in your points. Obviously I am making some mistake here but I am just not sure where it is! thanks First, you must consider short term and long term. Short term- In the start of the season, the player is reset to 0 points but their MMR is leftover from the previous season. - Thus, because of this disparity (points vs MMR), a high MMR player will receive alot for a win, but lose little for a loss. For the technical explanation of this, MMR is used for matchmaking but the win/loss of points for a game is determined by points against opponent MMR. Since there is a disparity between points and MMR, in the short-term, a high MMR player will receive alot for a win and lose little for a loss. - Because of this, if the amount of bonus pool = 0, points is "only" a decent indicator. Until it stabilises, it will not be totally accurate. - In the short term, it is not unusual for a player to have a lop-sided win/loss (e.g. 10-18 or 18-10). Whether you have a positive or negative one is dependent on your form or luck. If you're interested in the specifics of this, it has been shown that players near the top (but not pro) will tend to have negatives ratios. This is because they will likely be matched with pros and will lose disproportionately. However, don't think too much about this as it may confuse you - just consider what is not in italics.Long term- In the long run, the player's win/loss will converge to 50% (unless high master or GM). At this point in time, the points will converge to MMR. - And yes, on average, every game after the stabilisation will have on average 12 per win and 12 points for loss. Note that this average only occurs after the points converges to MMR. Before the convergence, this +12/-12 does not occur. The "extra points" generated is determined by how high the MMR originally is. - As an aside, you will find many ppl saying that the matchmaking system will "force" a player to a 50% win/loss ratio. This is incorrect - the matchmaking system attempts to find the best possible opponent in the next game. What happens is that the win/loss will converge to 50% rather than being "forced" to 50%. Also, to answer your questions: 1. You will not necessarily be paired against better opponents. Rather, it's not unusual to have a lop-sided win/loss in the short term. 10-18 is just as likely as 18-10, it really depends on your form and luck. However, eventually, the win/loss will converge to 50%. 2. I answered this above, but the average 12 points for win/loss only occurs when the points stabilises with MMR and not before. The "extra points" that you are referring to determined by how high the MMR is initially. Thanks a lot for detailed explanation! I guess the point where I was not clear is the fact that points for a win are not determined by the difference between your MMR and your opponent's ; but as you said "the win/loss of points for a game is determined by points against opponent MMR". Just to make it clear, does this mean that if you have only a few cumulated points and are playing against high MMR, you make a lot of points for win and little for losses, even if you are yourself to a similar high MMR; but if you have many points and are playing against high MMR, you make a more average number of points? If I am correct, that means that high MMR players are kind of guaranteed to make a lot of points in the beginning of the season (which sounds completely logical and fair). Thanks again for the answers
Yep to all your questions.
|
Amazing analysis. How does it feel to be so amazing that you have your own color on the blizzard forums?
Also, does anyone know if Grandmaster starts as soon as the next day starts tommorow, or just sometime during.
|
Norway10161 Posts
I was wondering why I keep being stuck in Silver. I only play 2v2s or 3v3s, but our MMR sees us facing high plat / low diamond most matches, yet we never seem to advance to gold. And we have a positive win ratio.
How are advancements determined?
Thanks!
|
do you have to be placed in a league in order for your MMR to carry over to the next season or does just 1 game carry it over?
for example:
if I wanted my 1v1 MMR reset and I played a 3v3 random match but did not get placed in a 3v3 league, would my 1v1 MMR carry over into the next season still? Or do I have to actually get placed in a league?
|
United States12267 Posts
On December 30 2011 14:43 darkgcn14 wrote:
do you have to be placed in a league in order for your MMR to carry over to the next season or does just 1 game carry it over?
for example:
if I wanted my 1v1 MMR reset and I played a 3v3 random match but did not get placed in a 3v3 league, would my 1v1 MMR carry over into the next season still? Or do I have to actually get placed in a league?
1v1 and random teams are linked, so it would still carry over to the next season.
|
United States12267 Posts
On December 30 2011 09:27 ToKoreaWithLove wrote:
I was wondering why I keep being stuck in Silver. I only play 2v2s or 3v3s, but our MMR sees us facing high plat / low diamond most matches, yet we never seem to advance to gold. And we have a positive win ratio.
How are advancements determined?
Thanks!
The concept of moving average, as explained in the guide. Either that moving average will exceed the top end of Gold completely and you'll move into the highest tier of Gold, or it won't and you'll eventually settle into one of the Gold tiers.
|
Is there a known fix for the bug where my placement matches aren't counting?
|
United States12267 Posts
On January 02 2012 06:21 LordOfDabu wrote:
Is there a known fix for the bug where my placement matches aren't counting?
Should be fixed as of this morning. Let me know if it isn't.
|
Yep, played a game and got placed. Thanks.
|
|
|
On January 05 2012 12:21 kellymilkies wrote:
This thread is God Tier.
So it doesn't exist?
|
I never get players according to my skill. I am ranked 80 gold and consistently play high platinum players. It's so frustrating that I don't even want to play anymore because I always lose each game. I am like a kitten thrown into the lion's den.
|
On January 07 2012 19:09 DirkLance wrote:
I never get players according to my skill. I am ranked 80 gold and consistently play high platinum players. It's so frustrating that I don't even want to play anymore because I always lose each game. I am like a kitten thrown into the lion's den.
If you keep losing, the system will begin to match you up against lower and lower opponents. It just takes time. Be a little patient.
If you are really upset and don't want to wait, you can try throwing games, but not everybody feels that this is acceptable and I am by no means promoting it, just putting it out there.
|
I have an interesting question for us lower end noobs:
1.) I was Gold for a long time. Towards the end of Season 2 and during Season 3 and Season 4 I hit a slump and lost a butt-ton of games. Then in Season 4 I have been matched with Bronze players and Silver Players that were slightly favoured against. (I understand it was their MMR vs. my points that made that happen.) However the gist of this post is:
Season 5 hits and I got demoted to Silver and I only played against Slightly favoured and Favoured opponents and they ended up in Platinum and Gold leagues. until I finally (after about 9/10 wins) was even-matched against bronze players, then as time went on I played against even-matched Silver, then even-matched Gold, then some even-matched Platinum until I got promoted to Gold League (Division Marauder Rho, if it matters and I dunno the Tier.)
After this happened, I once again only played vs. Slightly Favoured and Favoured player and they were in Gold Leagues and a couple in Platinum with one being in Silver(top8). Now I only play against Even-matched players in Gold and Platinum.
Does this mean that I'm climbing extremely fast? If so, why did I face the favoured opponents when I just enter a league and after around 8-10 games it starts to stabilize (even with wins and losses)?
Funny thing is, I started at a much lower win ratio in Silver. My win ratio is about ~70% now. (can't tell how many games (not more than 110, though))
2.) Also, I'm just getting clarification: If I earn 14 points for a win and my opponent loses 10 points for that loss, does that mean we're of same Tier?
|
|
|
|
|
|
|
|
EPT
