kshom.stack
New member
I joined the game on September 17th at the recommendation of a friend and have since played for a total of 250.7 hours, continuing to play regularly. However, my friends have quit the game due to dissatisfaction with the previous match-making system/MMR mechanism. After much thought, I've written down the following suggestions: A well-designed MMR algorithm or post-match evaluation system can significantly improve the current matching environment and overall gaming experience.
In competitive games, especially in MOBA games, MMR is designed to reflect a player's skill level with the aim of matching players of similar ability to create a fair competitive environment. However, it seems that the current matchmaking and rating system has deviated from this goal. I believe this is due to the evaluation system or MMR algorithm currently in use not being suitable for achieving the intended purpose.
I think that MMR should be adjusted based on performance in a given match, with wins and losses serving as a baseline for adjustment rather than the sole determining factor. This means that how much a player's MMR changes should depend more on their performance within the game, such as objectives achieved, teamwork, and other contributions, rather than just whether they won or lost.
Imagine a scenario in a match where both teams fight until the very last moment, each delivering outstanding performances. Why shouldn't the MMR of all 12 players increase? The winning team's MMR could shift positively by a larger margin, while the losing team's MMR could also increase, but to a lesser extent.
This approach acknowledges the high level of skill and effort from all players involved, rewarding them for an exceptionally well-played match regardless of the outcome. It encourages positive gameplay and recognizes that even in defeat, players can demonstrate remarkable skill and teamwork. Such a system could potentially foster a more positive community and competitive environment, as it values performance and contribution over just the binary result of win or lose
In a MOBA or pushing-line game type like Dota 2, which features rich gameplay data recording, the development and refinement of post-match evaluation systems or MMR algorithms are crucial for ensuring fair matchmaking and rewarding player performance accurately. During the testing phase, here are some suggestions that could be considered for improving these systems:
Firstly, player damage should be changed to "effective player damage," which means that during matches, only the damage distribution leading to the elimination of opponent players should be recorded
Secondly, add a metric for "damage taken," which records the amount of damage a player sustains from opponent players that does not result in their elimination
Additionally, "Total Souls" should be divided into two categories:
Competitive Souls: These include souls obtained through actions that are beneficial to the team, such as:
The acquisition of Competitive Souls contributes positively to the team effort, and a higher amount is desirable. In contrast, Non-Competitive Souls are neutral; players need to convert this resource into activities that benefit the team.
Therefore, the following new data indicators can be defined:
The final post-match performance MMR should consist of three components: the base MMR score from winning or losing, the Economy to Combat Effectiveness Conversion Rate multiplied by a coefficient, and Other Performance Metrics multiplied by a coefficient. Assuming a match win grants a base score of +5 MMR and a loss results in -5 MMR, I believe that after a match concludes, the MMR shift for the winning team should be between 0 and +15 points, while the losing team's MMR shift should range from -10 to +5 points.
Hero performance can also be rated using grades such as S, A, B, C, or a more sophisticated rating system could be designed based on in-game data to create a clearer and more reasonable MMR system. During the testing phase, it is crucial for players to understand how their performance compares to others playing the same hero, so they can provide better feedback on issues and balance concerns.
To summarize:
MMR Shift = Base Score (+5) + (Economy to Combat Effectiveness Conversion Rate * Coefficient) + (Other Performance Metrics * Coefficient)
This approach would address the current matchmaking issues by incentivizing players to continue performing their best even if the match-up of teammates and opponents is not ideal. Players would understand that by demonstrating superior performance, their MMR can still shift positively regardless of the match outcome. This encouragement could prevent both teams from giving up early in a game, fostering a more competitive and engaging environment.
By matching higher-performing players more quickly to higher ranks and ensuring that average-performing players find matches within their MMR range, the game can prevent players from being consistently outmatched and overwhelmed. This approach ensures that the enjoyment of the game is maximized, avoiding one-sided matches where one team dominates too easily.
Additionally, introducing a “Thrilling Match Score” could reward both teams when the total souls difference at the end of the match is very close or when both teams have successfully cleared each other's guardians. After such an engaging battle, players would receive this bonus score regardless of the match outcome, recognizing the high level of competition and effort from both sides.
I hope the above suggestions are helpful for you
In competitive games, especially in MOBA games, MMR is designed to reflect a player's skill level with the aim of matching players of similar ability to create a fair competitive environment. However, it seems that the current matchmaking and rating system has deviated from this goal. I believe this is due to the evaluation system or MMR algorithm currently in use not being suitable for achieving the intended purpose.
I think that MMR should be adjusted based on performance in a given match, with wins and losses serving as a baseline for adjustment rather than the sole determining factor. This means that how much a player's MMR changes should depend more on their performance within the game, such as objectives achieved, teamwork, and other contributions, rather than just whether they won or lost.
Imagine a scenario in a match where both teams fight until the very last moment, each delivering outstanding performances. Why shouldn't the MMR of all 12 players increase? The winning team's MMR could shift positively by a larger margin, while the losing team's MMR could also increase, but to a lesser extent.
This approach acknowledges the high level of skill and effort from all players involved, rewarding them for an exceptionally well-played match regardless of the outcome. It encourages positive gameplay and recognizes that even in defeat, players can demonstrate remarkable skill and teamwork. Such a system could potentially foster a more positive community and competitive environment, as it values performance and contribution over just the binary result of win or lose
In a MOBA or pushing-line game type like Dota 2, which features rich gameplay data recording, the development and refinement of post-match evaluation systems or MMR algorithms are crucial for ensuring fair matchmaking and rewarding player performance accurately. During the testing phase, here are some suggestions that could be considered for improving these systems:
Firstly, player damage should be changed to "effective player damage," which means that during matches, only the damage distribution leading to the elimination of opponent players should be recorded
Secondly, add a metric for "damage taken," which records the amount of damage a player sustains from opponent players that does not result in their elimination
Additionally, "Total Souls" should be divided into two categories:
Competitive Souls: These include souls obtained through actions that are beneficial to the team, such as:
- Last-hitting soul orbs,
- Souls gained from kills and assists,
- Souls from destroying guardians,
- Souls obtained via Urn,
- Souls from killing bosses.
- Killing troopers,
- Killing neutral NPCs, i.e., jungle camps.
The acquisition of Competitive Souls contributes positively to the team effort, and a higher amount is desirable. In contrast, Non-Competitive Souls are neutral; players need to convert this resource into activities that benefit the team.
Therefore, the following new data indicators can be defined:
- Percentage of Non-Competitive Souls Economy = Individual Non-Competitive Economy / Team Total Non-Competitive Economy
- Percentage of Effective Player Damage = Individual Effective Player Damage / Team Total Effective Player Damage
- Percentage of Objective Damage = Individual Objective Damage / Team Total Objective Damage
- Percentage of Damage Taken = Individual Damage Taken / Team Total Damage Taken
- Economy to Combat Effectiveness Conversion Rate = (Percentage of Effective Player Damage + Percentage of Objective Damage + Percentage of Damage Taken) / Percentage of Non-Competitive Souls Economy
- Hero Healing Amount
- Hero Control Time and Number of Controlled Heroes
- Number of Enemies Hit by Hero Skills
The final post-match performance MMR should consist of three components: the base MMR score from winning or losing, the Economy to Combat Effectiveness Conversion Rate multiplied by a coefficient, and Other Performance Metrics multiplied by a coefficient. Assuming a match win grants a base score of +5 MMR and a loss results in -5 MMR, I believe that after a match concludes, the MMR shift for the winning team should be between 0 and +15 points, while the losing team's MMR shift should range from -10 to +5 points.
Hero performance can also be rated using grades such as S, A, B, C, or a more sophisticated rating system could be designed based on in-game data to create a clearer and more reasonable MMR system. During the testing phase, it is crucial for players to understand how their performance compares to others playing the same hero, so they can provide better feedback on issues and balance concerns.
To summarize:
MMR Shift = Base Score (+5) + (Economy to Combat Effectiveness Conversion Rate * Coefficient) + (Other Performance Metrics * Coefficient)
This approach would address the current matchmaking issues by incentivizing players to continue performing their best even if the match-up of teammates and opponents is not ideal. Players would understand that by demonstrating superior performance, their MMR can still shift positively regardless of the match outcome. This encouragement could prevent both teams from giving up early in a game, fostering a more competitive and engaging environment.
By matching higher-performing players more quickly to higher ranks and ensuring that average-performing players find matches within their MMR range, the game can prevent players from being consistently outmatched and overwhelmed. This approach ensures that the enjoyment of the game is maximized, avoiding one-sided matches where one team dominates too easily.
Additionally, introducing a “Thrilling Match Score” could reward both teams when the total souls difference at the end of the match is very close or when both teams have successfully cleared each other's guardians. After such an engaging battle, players would receive this bonus score regardless of the match outcome, recognizing the high level of competition and effort from both sides.
I hope the above suggestions are helpful for you
