devtrackers.gg
about 1 year
ago -
GM_Goddess
-
Direct link
Hello Lumia Island survivors,
It's been ten days since the Eternal Return Ranked Season 1 began, and we're so grateful that so many of you have joined our revamped Ranked Matches to compete for Rank Points. As the Ranked System has undergone significant changes, there has been a lot of feedback on the new system. So we'd like to summarize some of the key feedback and improvement plans.
At the beginning of the season, there was a major disruption in matchmaking in the Platinum and higher tiers, which was a big reason why the soft reset process didn't go as smoothly as it should have. With a completely different scoring system and an emergency rebalancing of the scoring system due to excessive preseason inflation in the new scoring system, the formula for the soft reset process became very complex, and we made two major mistakes along the way.
First, the previous season's Solo and Duo Ranks were not properly reflected. We overlooked this, thinking that preseason would do most of the repositioning, but the top Solo/Duo players who didn't play any ranks in preseason ended up starting in a lower tier. Solos and Duos don't necessarily match the tiers of Squads, but they needed to be calibrated separately to some degree.
Second, the overall timing of the soft reset was too tight: the soft reset has always been tuned to have top players start at around the Platinum tier each season, and this time it was the same. However, while the old ELO system had a placement phase at the beginning of the season, the new Ranked System's Entry Cost system made it impossible for a placement phase to exist, and it took more time for the top tier players to find their place.
This was frustrating for top players, but it was also frustrating for other players who had to deal with these high tier players. We sincerely apologize for any inconvenience caused by this lack of preparation and management regarding placements.
Next season, we'll allow top players to start the new season in the Diamond tier and have their matches more separated from other players to avoid confusion early in the Ranked season. We're also considering limiting premade teams to the Diamond tier for the first week of the Ranked season only.
Early in this season, the issue of meeting top tier players' premade teams from previous seasons was largely related to soft resets, so we'll work on improving this measure for next season. However, we also need to take a longer and more systematic approach to tackle the advantages of premade teams in the overall Ranked System.
① Separate Matching for Premade Teams and Non-Premade Teams
Currently, Eternal Return's Ranked System does not allow premade teams starting at Mythril tier. This policy is to avoid making premade teams a must in order to achieve Immortal, as the advantage of premade teams tends to be greater at the top.
The ideal solution would be to separate "solo-only queues", where premade teams are not possible from "traditional ranks", but that would be a premature decision. Due to the much higher share of Normal Matches compared to Ranked Matches, there are still some gaps in matchmaking in Ranked Matches. Separating matchmaking will likely lead to even more significant deterioration in the matchmaking quality, especially on the solo-only side of the queue.
Prior to Official Launch during Early Access, when we had separate matchmaking queues for the three modes, we noticed a lot of issues, especially during off-peak hours, and this could be the case again. However, we're seeing an upward trend in ranked over time, so the situation may change in the next season or so, and we'll eventually move toward separating them.
While we don't want to separate the ranks outright, we do want to improve the matching of premade teams and non-premade teams to be as separate as possible during peak hours. We're looking to implement a tiered matching system where 3-player premade teams will only be matched against premade teams, and non-premade teams will only be matched against non-premade teams, but will be matched together if either queue time exceeds a certain time limit. This type of matching, by stages won't solve 100% of the issues with premade teams, but we expect it to be a significant improvement. However, this will require complex work across multiple servers, including the matching server, API server, and battle server, so we're aiming to apply this imrovement approximately on the 1.4 Update on September 28th.
② Bonus for Non-Premade Teams
How much of an advantage do premade teams have over non-premade teams is a tricky one because there are so many variables, but based on simply how many Rank Points they get, it's about a 5-point difference. 3-person premade teams earn an average of 34 points, while 1 and 2-person premade teams earn an average of 29 points.
If this new matching system works well, this gap could be reduced. However, even with this improvement, longer queues will likely mean that we'll need to offer a small entry discount to non-premade teams if they are matched together, which could help reduce the scoring gap.
This is simply an average of 5 points, but the difference varies by tier. The tier with the smallest difference between premade and non-premade teams is Gold, where a 1-person team and a 3-person team earn almost the same amount of points. The tier with the largest difference is Iron. We're only looking at 10 days of stats, and the metrics may change slightly, so we'll keep you posted on how we'll be making adjustments in the future.
③ Tier Difference Due to Matching by Highest Score for Premade Teams
While premade teams score slightly more points on average, there are times when they are at a disadvantage. This is when the tier difference between premade team members becomes significant.
Normal matches will attempt to match based on the average score of the premade team members, while Ranked matches will attempt to match based on the highest score. For example, if you create a premade team with 2 Gold and 2 Diamond tiers, you will be treated as a Diamond tier for matchmaking purposes and may also be matched with a Mythril tier. This is the most common case we've heard of people being matched to too high or too low tiers.
If you're matchmaking on the average of the premade teams, you might be at an advantage and get some sort of boosting. If you end up matching in Platinum, which is the middle of the two tiers in the previous example, as a Diamond player you'll be playing in an easier environment than usual, while the Gold players will have a harder time. Since the difference in skills between the tiers is more variable, this can be overall seen as an easier way to climb ranked for Diamond players.
As a result, we've placed a greater emphasis on completely eliminating the possibility of intentional boosting for the sake of ranked fairness. However, this policy does not address the issue of lower tier players on premade teams facing an overly difficult competitive environment, which is something we're really concerned about. So don't hesitate and let us know your thoughts, we look forward to hearing your feedback.
④ Team Communication
There are many reasons why premade teams have an advantage, but especially the means of communication within the team is a big one. There is even a consensus that premade teams that use Voice Chat are more efficient than non-premade teams that don't use it.
Voice Chat is also available as an in-game function in Eternal Return, but it's underutilized because it defaults to "Disabled". The recent trend in other team games is to have voice built in, but with PTT (Push to Talk) as the default setting. You have to press a hotkey to speak, but you can listen by default, and of course, you can change the setting to disable it if you want. We'll be bringing this feature to Eternal Return in the 1.4 Update on September 14th.
There's also a lot of room for improvement in the ping system using Alt and Ctrl. The ping system needs to be improved little by little over time, but a major improvement will be one of our most important update goals for Season 2 (October 26th), allowing our players to communicate even better than now. Voice is a big help, but a more fundamental improvement would be a ping system that goes beyond language barriers.
⑤ Feedback on the Amount of Rank Points Earned
Outside of the premade team, we also received a lot of feedback on the scoring system for the new Ranked System.
At the end of a Ranked Match, it's always nice to get more points, so we've gotten a lot of feedback saying that players would like to get more or less points for different situations. For example more points for first place or getting higher placement points even when getting few kills. More points for getting more kills and less points for being eliminated.
However, when it comes to how much Rank Points are earned, you need to account for the overall inflation. Last season, Squads averaged about +10 points per game. Since the Ranked scoring system is now more than doubled (2.5x), this season's score included roughly a plus sum of 25 points.
During the two weeks of the preseason, we averaged +45 points per game, but the tiers were rising so fast that the percentage of top tiers became so large that it was hard to actually call them "top tiers." If this continued for a full season, the value of ranked tiers would become less meaningful, defeating the purpose of the Ranked System.
After rebalancing, Eternal Return 1.0 Season 1 is currently averaging at +31 points per game. That's less than preseason, but still higher than last season, so we're still concerned about inflation. Therefore, in order to give more points, whether it's placement or kills points, we're going to have to cut them from somewhere else, and that's a much harder decision and one that requires a lot more thought.
It's also very important that the Ranked System in particular has long-term reliability, rather than just being tweaked from patch to patch. For example, if one week it's harder to get points and suddenly the next week it's easy to earn them, this could really undermine the credibility of the Ranked System, so we want to take a deep breath and analyze the feedback and stats a little bit more, rather than reacting immediately based on these 10 days of stats.
With a major revamp to the Ranked System for Official Launch, we're sure there will be many more iterations and improvements to come. We'd love to hear your feedback so we can continue to improve and make Eternal Return even more fun.
As always thank you for sticking with us!
It's been ten days since the Eternal Return Ranked Season 1 began, and we're so grateful that so many of you have joined our revamped Ranked Matches to compete for Rank Points. As the Ranked System has undergone significant changes, there has been a lot of feedback on the new system. So we'd like to summarize some of the key feedback and improvement plans.
At the beginning of the season, there was a major disruption in matchmaking in the Platinum and higher tiers, which was a big reason why the soft reset process didn't go as smoothly as it should have. With a completely different scoring system and an emergency rebalancing of the scoring system due to excessive preseason inflation in the new scoring system, the formula for the soft reset process became very complex, and we made two major mistakes along the way.
First, the previous season's Solo and Duo Ranks were not properly reflected. We overlooked this, thinking that preseason would do most of the repositioning, but the top Solo/Duo players who didn't play any ranks in preseason ended up starting in a lower tier. Solos and Duos don't necessarily match the tiers of Squads, but they needed to be calibrated separately to some degree.
Second, the overall timing of the soft reset was too tight: the soft reset has always been tuned to have top players start at around the Platinum tier each season, and this time it was the same. However, while the old ELO system had a placement phase at the beginning of the season, the new Ranked System's Entry Cost system made it impossible for a placement phase to exist, and it took more time for the top tier players to find their place.
This was frustrating for top players, but it was also frustrating for other players who had to deal with these high tier players. We sincerely apologize for any inconvenience caused by this lack of preparation and management regarding placements.
Next season, we'll allow top players to start the new season in the Diamond tier and have their matches more separated from other players to avoid confusion early in the Ranked season. We're also considering limiting premade teams to the Diamond tier for the first week of the Ranked season only.
Early in this season, the issue of meeting top tier players' premade teams from previous seasons was largely related to soft resets, so we'll work on improving this measure for next season. However, we also need to take a longer and more systematic approach to tackle the advantages of premade teams in the overall Ranked System.
① Separate Matching for Premade Teams and Non-Premade Teams
Currently, Eternal Return's Ranked System does not allow premade teams starting at Mythril tier. This policy is to avoid making premade teams a must in order to achieve Immortal, as the advantage of premade teams tends to be greater at the top.
The ideal solution would be to separate "solo-only queues", where premade teams are not possible from "traditional ranks", but that would be a premature decision. Due to the much higher share of Normal Matches compared to Ranked Matches, there are still some gaps in matchmaking in Ranked Matches. Separating matchmaking will likely lead to even more significant deterioration in the matchmaking quality, especially on the solo-only side of the queue.
Prior to Official Launch during Early Access, when we had separate matchmaking queues for the three modes, we noticed a lot of issues, especially during off-peak hours, and this could be the case again. However, we're seeing an upward trend in ranked over time, so the situation may change in the next season or so, and we'll eventually move toward separating them.
While we don't want to separate the ranks outright, we do want to improve the matching of premade teams and non-premade teams to be as separate as possible during peak hours. We're looking to implement a tiered matching system where 3-player premade teams will only be matched against premade teams, and non-premade teams will only be matched against non-premade teams, but will be matched together if either queue time exceeds a certain time limit. This type of matching, by stages won't solve 100% of the issues with premade teams, but we expect it to be a significant improvement. However, this will require complex work across multiple servers, including the matching server, API server, and battle server, so we're aiming to apply this imrovement approximately on the 1.4 Update on September 28th.
② Bonus for Non-Premade Teams
How much of an advantage do premade teams have over non-premade teams is a tricky one because there are so many variables, but based on simply how many Rank Points they get, it's about a 5-point difference. 3-person premade teams earn an average of 34 points, while 1 and 2-person premade teams earn an average of 29 points.
If this new matching system works well, this gap could be reduced. However, even with this improvement, longer queues will likely mean that we'll need to offer a small entry discount to non-premade teams if they are matched together, which could help reduce the scoring gap.
This is simply an average of 5 points, but the difference varies by tier. The tier with the smallest difference between premade and non-premade teams is Gold, where a 1-person team and a 3-person team earn almost the same amount of points. The tier with the largest difference is Iron. We're only looking at 10 days of stats, and the metrics may change slightly, so we'll keep you posted on how we'll be making adjustments in the future.
③ Tier Difference Due to Matching by Highest Score for Premade Teams
While premade teams score slightly more points on average, there are times when they are at a disadvantage. This is when the tier difference between premade team members becomes significant.
Normal matches will attempt to match based on the average score of the premade team members, while Ranked matches will attempt to match based on the highest score. For example, if you create a premade team with 2 Gold and 2 Diamond tiers, you will be treated as a Diamond tier for matchmaking purposes and may also be matched with a Mythril tier. This is the most common case we've heard of people being matched to too high or too low tiers.
If you're matchmaking on the average of the premade teams, you might be at an advantage and get some sort of boosting. If you end up matching in Platinum, which is the middle of the two tiers in the previous example, as a Diamond player you'll be playing in an easier environment than usual, while the Gold players will have a harder time. Since the difference in skills between the tiers is more variable, this can be overall seen as an easier way to climb ranked for Diamond players.
As a result, we've placed a greater emphasis on completely eliminating the possibility of intentional boosting for the sake of ranked fairness. However, this policy does not address the issue of lower tier players on premade teams facing an overly difficult competitive environment, which is something we're really concerned about. So don't hesitate and let us know your thoughts, we look forward to hearing your feedback.
④ Team Communication
There are many reasons why premade teams have an advantage, but especially the means of communication within the team is a big one. There is even a consensus that premade teams that use Voice Chat are more efficient than non-premade teams that don't use it.
Voice Chat is also available as an in-game function in Eternal Return, but it's underutilized because it defaults to "Disabled". The recent trend in other team games is to have voice built in, but with PTT (Push to Talk) as the default setting. You have to press a hotkey to speak, but you can listen by default, and of course, you can change the setting to disable it if you want. We'll be bringing this feature to Eternal Return in the 1.4 Update on September 14th.
There's also a lot of room for improvement in the ping system using Alt and Ctrl. The ping system needs to be improved little by little over time, but a major improvement will be one of our most important update goals for Season 2 (October 26th), allowing our players to communicate even better than now. Voice is a big help, but a more fundamental improvement would be a ping system that goes beyond language barriers.
⑤ Feedback on the Amount of Rank Points Earned
Outside of the premade team, we also received a lot of feedback on the scoring system for the new Ranked System.
At the end of a Ranked Match, it's always nice to get more points, so we've gotten a lot of feedback saying that players would like to get more or less points for different situations. For example more points for first place or getting higher placement points even when getting few kills. More points for getting more kills and less points for being eliminated.
However, when it comes to how much Rank Points are earned, you need to account for the overall inflation. Last season, Squads averaged about +10 points per game. Since the Ranked scoring system is now more than doubled (2.5x), this season's score included roughly a plus sum of 25 points.
During the two weeks of the preseason, we averaged +45 points per game, but the tiers were rising so fast that the percentage of top tiers became so large that it was hard to actually call them "top tiers." If this continued for a full season, the value of ranked tiers would become less meaningful, defeating the purpose of the Ranked System.
After rebalancing, Eternal Return 1.0 Season 1 is currently averaging at +31 points per game. That's less than preseason, but still higher than last season, so we're still concerned about inflation. Therefore, in order to give more points, whether it's placement or kills points, we're going to have to cut them from somewhere else, and that's a much harder decision and one that requires a lot more thought.
It's also very important that the Ranked System in particular has long-term reliability, rather than just being tweaked from patch to patch. For example, if one week it's harder to get points and suddenly the next week it's easy to earn them, this could really undermine the credibility of the Ranked System, so we want to take a deep breath and analyze the feedback and stats a little bit more, rather than reacting immediately based on these 10 days of stats.
With a major revamp to the Ranked System for Official Launch, we're sure there will be many more iterations and improvements to come. We'd love to hear your feedback so we can continue to improve and make Eternal Return even more fun.
As always thank you for sticking with us!