Riggy wrote: Most of these are just speculation on how matchmaking is done. I recognize some of these as forum lore and some are new and as far as I know totally unconfirmed. However, I think that you'll find that any Matchmaking algorithm is going to be fundamentally flawed unless it takes into account persistent damage. I've long since argued that there are better ways to monetize the game as well as improve gameplay / matchmaking / play time if persistent damage was removed. For myself, I think any algorithm should factor in your "Best playable" team for any given event. In the current event, my best team only involves one 332 character, so why should I face off against someone with 3x 332 chars? Sure, I have 10+ max level characters, but since all but one are unusable they shouldn't count against me.
stephen43084 wrote: Riggy wrote: Most of these are just speculation on how matchmaking is done. I recognize some of these as forum lore and some are new and as far as I know totally unconfirmed. However, I think that you'll find that any Matchmaking algorithm is going to be fundamentally flawed unless it takes into account persistent damage. I've long since argued that there are better ways to monetize the game as well as improve gameplay / matchmaking / play time if persistent damage was removed. For myself, I think any algorithm should factor in your "Best playable" team for any given event. In the current event, my best team only involves one 332 character, so why should I face off against someone with 3x 332 chars? Sure, I have 10+ max level characters, but since all but one are unusable they shouldn't count against me. But if you calculate off best possible team, you are saying players are expected to only use that team composition. So I no longer get to choose my team. So instead of encouraging player roster diversity, this eliminates it to the max.
Riggy wrote: stephen43084 wrote: Riggy wrote: Most of these are just speculation on how matchmaking is done. I recognize some of these as forum lore and some are new and as far as I know totally unconfirface. However, I think that you'll find that any Matchmaking algorithm is going to be fundamentally flawed unless it takes into account persistent damage. I've long since argued that there are better ways to monetize the game as well as improve gameplay / matchmaking / play time if persistent damage was removed. For myself, I think any algorithm should factor in your "Best playable" team for any given event. In the current event, my best team only involves one 332 character, so why should I face off against someone with 3x 332 chars? Sure, I have 10+ max level characters, but since all but one are unusable they shouldn't count against me. But if you calculate off best possible team, you are saying players are expected to only use that team composition. So I no longer get to choose my team. So instead of encouraging player roster diversity, this eliminates it to the max. But that means MMR is impossible to calculate. If it factors in anything other than my best team, then my best team will be able to beat handily any opponent match-up. Ultimately, it sounds like you should have an event specific ELO ranking where everyone starts off at the break even score, with some amount of weighting based on the number of available characters, covers and levels for the given set of eligible characters for that event. But then, it still has to factor in the opponent's points, which is another issue altogether. So amend the previous sentence's suggestion to say they should offer an "easy", "normal", and "hard" node and give points based off the difficulty (and consequently cancel defensive point losses).
stephen43084 wrote: Riggy wrote: Most of these are just speculation on how matchmaking is done. I recognize some of these as forum lore and some are new and as far as I know totally unconfirface. However, I think that you'll find that any Matchmaking algorithm is going to be fundamentally flawed unless it takes into account persistent damage. I've long since argued that there are better ways to monetize the game as well as improve gameplay / matchmaking / play time if persistent damage was removed. For myself, I think any algorithm should factor in your "Best playable" team for any given event. In the current event, my best team only involves one 332 character, so why should I face off against someone with 3x 332 chars? Sure, I have 10+ max level characters, but since all but one are unusable they shouldn't count against me. But if you calculate off best possible team, you are saying players are expected to only use that team composition. So I no longer get to choose my team. So instead of encouraging player roster diversity, this eliminates it to the max.
Riggy wrote: Most of these are just speculation on how matchmaking is done. I recognize some of these as forum lore and some are new and as far as I know totally unconfirface. However, I think that you'll find that any Matchmaking algorithm is going to be fundamentally flawed unless it takes into account persistent damage. I've long since argued that there are better ways to monetize the game as well as improve gameplay / matchmaking / play time if persistent damage was removed. For myself, I think any algorithm should factor in your "Best playable" team for any given event. In the current event, my best team only involves one 332 character, so why should I face off against someone with 3x 332 chars? Sure, I have 10+ max level characters, but since all but one are unusable they shouldn't count against me.
stephen43084 wrote: Riggy wrote: stephen43084 wrote: Riggy wrote: Most of these are just speculation on how matchmaking is done. I recognize some of these as forum lore and some are new and as far as I know totally unconfirface. However, I think that you'll find that any Matchmaking algorithm is going to be fundamentally flawed unless it takes into account persistent damage. I've long since argued that there are better ways to monetize the game as well as improve gameplay / matchmaking / play time if persistent damage was removed. For myself, I think any algorithm should factor in your "Best playable" team for any given event. In the current event, my best team only involves one 332 character, so why should I face off against someone with 3x 332 chars? Sure, I have 10+ max level characters, but since all but one are unusable they shouldn't count against me. But if you calculate off best possible team, you are saying players are expected to only use that team composition. So I no longer get to choose my team. So instead of encouraging player roster diversity, this eliminates it to the max. But that means MMR is impossible to calculate. If it factors in anything other than my best team, then my best team will be able to beat handily any opponent match-up. Ultimately, it sounds like you should have an event specific ELO ranking where everyone starts off at the break even score, with some amount of weighting based on the number of available characters, covers and levels for the given set of eligible characters for that event. But then, it still has to factor in the opponent's points, which is another issue altogether. So amend the previous sentence's suggestion to say they should offer an "easy", "normal", and "hard" node and give points based off the difficulty (and consequently cancel defensive point losses). MMR is easy, despite the efforts to show complicated is better. MMR should reset for everyone each tournament. No need for tanking. Pick point levels for tiers. The higher the tier the harder opponents you face. This calls for the reward structure to be changed to provide a pathway for progress. If the goal for the bottom 100 is to be 1* rosters, make that reward 2* covers and I so. If the goal for the 200-300 is to be transitioning player at the minimum the rewards should include 1 3* cover iso and enough HP to buy a roster slot each X # of weeks with that consistent placement. And so on. The problem transitioning players face is there is no path to progress. There is only cross your fingers and pray. The status quo of punish vets with sharding, punish transitioners with a dearth of covers benefits no one. Im not saying the game should be a gimme palooza. However, there should be predetermined transition period for transitioning players, and that amount of time and gameplay should be reasonable and obtainable. The current system irks veterans and newer players. Instead of trying to outsmart everyone (including themselves), KISS and let players have fun.
stephen43084 wrote: MMR is easy, despite the efforts to show complicated is better. MMR should reset for everyone each tournament. No need for tanking. Pick point levels for tiers. The higher the tier the harder opponents you face. This calls for the reward structure to be changed to provide a pathway for progress. If the goal for the bottom 100 is to be 1* rosters, make that reward 2* covers and I so. If the goal for the 200-300 is to be transitioning player at the minimum the rewards should include 1 3* cover iso and enough HP to buy a roster slot each X # of weeks with that consistent placement. And so on. The problem transitioning players face is there is no path to progress. There is only cross your fingers and pray. The status quo of punish vets with sharding, punish transitioners with a dearth of covers benefits no one. Im not saying the game should be a gimme palooza. However, there should be predetermined transition period for transitioning players, and that amount of time and gameplay should be reasonable and obtainable. The current system irks veterans and newer players. Instead of trying to outsmart everyone (including themselves), KISS and let players have fun.
daibar wrote: stephen43084 wrote: MMR is easy, despite the efforts to show complicated is better. MMR should reset for everyone each tournament. No need for tanking. Pick point levels for tiers. The higher the tier the harder opponents you face. This calls for the reward structure to be changed to provide a pathway for progress. If the goal for the bottom 100 is to be 1* rosters, make that reward 2* covers and I so. If the goal for the 200-300 is to be transitioning player at the minimum the rewards should include 1 3* cover iso and enough HP to buy a roster slot each X # of weeks with that consistent placement. And so on. The problem transitioning players face is there is no path to progress. There is only cross your fingers and pray. The status quo of punish vets with sharding, punish transitioners with a dearth of covers benefits no one. Im not saying the game should be a gimme palooza. However, there should be predetermined transition period for transitioning players, and that amount of time and gameplay should be reasonable and obtainable. The current system irks veterans and newer players. Instead of trying to outsmart everyone (including themselves), KISS and let players have fun. The current MMR makes it more fun for casual players and new players by actually giving them a chance to rank. This is probably the majority of players, so it makes sense that they will cater to this crowd. Few new players are going to spend money on boosts and shields to go from rank 500 to rank 400, but some will to get into top 50 or higher. Creating a minor league of players encourages growth, keeps the new blood flowing. If you're going a purely f2p route, why are you so surprised that the game makes it difficult? Praying is useless without effort. Most of the bottom 50%-60% don't even really show up in PVP; why should they automatically get a 2* cover? If you're going the f2p route and still expect to win, you should be prepared to work harder and possibly give up sleep in order to compete with other people who optimize their strategies, AND/OR spend money on the game. The game almost tells you: if you're ranking below top 100 and want to move forward, then either you need to build your 2* roster more, you have to make a great effort to get into the top 10%, you need to shield, or settle for the hp and few tokens you get.
Grosnours wrote: With simple tools like average, median and standard deviation you can already imagine quite sophisticated ways of evaluating a player's roster. On top of that there's also a ton of data about previous behavior available to fine tune those calculations. But the MMR is a tool for a purpose, a way to enact a certain policy. So the question should be: Should the players with the most diverse roster be advantaged ? Or the ones with the most ISO and HP invested ? Or the beginners ? Or the ones which dealt the most damage in the last matches ? Or the ones that received the less damages in the last matches ? Or any other category ? And that's ultimately for the devs to decide, even though we can certainly voice our opinion on the subject.
wirius wrote: Welcome to RPGs. The higher your level, the harder you need to work to get higher.
Bowgentle wrote: wirius wrote: Welcome to RPGs. The higher your level, the harder you need to work to get higher. Usually in RPGs the harder the battle, the bigger the rewards. Here I'm supposed to fight all maxed teams all the time for the same rewards that Joe Blow, who started a month ago, gets? If building my roster only means harder fights for the same rewards, why build my roster? We all weren't born with good rosters. We worked for them. Yes, we had an easier time to get covers back then. The fact remains that we have been here for a year and put in the time and money to get good rosters. We continually get penalised for it. Sure, pair me against maxed Sentry/Daken/Hood every time. But then give me 1.000 ISO per victory or a meaningful progression reward below 1100.
benben77 wrote: the problem is the gap between veteran and new comer Path to catch up SHOULD be easy.
wirius wrote: The problem I see with your list is this. "I want MMR that lets me fight people weaker than myself." Welcome to RPGs. The higher your level, the harder you need to work to get higher.
MarvelMan wrote: benben77 wrote: the problem is the gap between veteran and new comer Path to catch up SHOULD be easy. No, the path should NOT be "easy." It should be straightforward and doable, but not easy. And there should be rewards for having traveled it.