Howdy, Stranger!
It looks like you're new here. If you want to get involved, click one of these buttons!
Quick Links
A good grouping game must create good groups
Quizzical
Member LegendaryPosts: 25,499
Some people want an MMORPG with good group content. Rather fewer want a game where they spend half of their time trying to get a group so that they can do anything. But I would submit that the harder part of a good grouping game is not, here's the content you can do after you have your group. Rather, the harder question is, where does your group come from? That's an issue that I haven't yet seen a game entirely solve, and that's what I'd like to address.
The industry has moved considerably toward solo content as a way to avoid having to get a group. There's nothing wrong with some games being solo games. But there's also a place in the world for group-focused games. Some games end up as basically solo games up to the level cap, then abruptly become grouping games, which is a mess.
Some have tried to push grouping by making it so that you have to have groups to do the content that gives the best gear. Or perhaps instead, you'll level faster and get better gear while grouped. But that does exactly nothing to address the question of where groups come from. A combination of "you must group to play this game" and "you can't get a group" turns into "you can't play this game".
Some people propose making friends and then grouping with your friends. But that has never been a viable option outside of the relative handful of people willing to schedule their lives around a computer game. The problem is that you make friends, but then they're on at different times from you, or want to do different content from you. Often the game won't even let you group with them because of level differences, or at least it's strongly discouraged.
Some games have tried to assist with group formation by having an automatic grouping tool. People want a group for some content, so the game waits until there are enough people who want to group for the content, then throws them together in a group. While perhaps better than spending half an hour trying to get a group, this creates problems of its own.
One is that a lot of games implicitly assume some group composition that doesn't match the player base composition. For example, if a game assumes that 25% of group members are healers, but only 15% of the players are healers, you can get 60% of the players a group with a healer, while the other 40% can't get one as there aren't enough healers to go around. This could result in groups without a healer that have no real hope of completing content. More commonly, you end up with the players who aren't healers spending half of their "grouping" time waiting for a group to form. That's pretty much how it goes in FFXIV for players who are neither healers nor tanks, for example.
Another problem is that different players have wildly different playing styles, sometimes to the extent of being incompatible. Some players want to speed rush through the content as fast as they can, skip everything possible, get the loot, and move on. Others want to have a look around, see what there is to see, clear side stuff, and generally take their time. There's nothing intrinsically wrong with either playing style, but putting some of each in a group together is asking for trouble.
Yet another problem is not having enough players to do given content. Some content becomes unpopular for one reason or another, to the extent that it's very hard to get a group for it at all. If there are two dungeons at a given level, one of which is generally perceived to give twice as good of loot as the other, it can be very hard to get a group for the one that gives less loot. Now, double loot is bad game design, but a 5% difference can cause the same problem. Or even no systematic difference at all, but only a perception of such, as players get better at one dungeon from practice with it.
The reason this is a problem is that, if you can't get a group for half of the group content in a game, you only effectively have half as much content. Farming the same dungeon endlessly is boring, and you need as much content as possible. Adding more content that people ignore is scarcely better than not adding content at all. More commonly, games add new content with better loot than the old to get people to do it, which implicitly deprecates the old content and then you have no more content than before.
If a game is going to make group content into a major game mechanic, it needs to solve these problems. I would go so far as to say that solving the problem of where good groups come from is more important than the actual content that the groups do once formed. Plenty of games have done a decent job of the latter.
(continued on next post)
The industry has moved considerably toward solo content as a way to avoid having to get a group. There's nothing wrong with some games being solo games. But there's also a place in the world for group-focused games. Some games end up as basically solo games up to the level cap, then abruptly become grouping games, which is a mess.
Some have tried to push grouping by making it so that you have to have groups to do the content that gives the best gear. Or perhaps instead, you'll level faster and get better gear while grouped. But that does exactly nothing to address the question of where groups come from. A combination of "you must group to play this game" and "you can't get a group" turns into "you can't play this game".
Some people propose making friends and then grouping with your friends. But that has never been a viable option outside of the relative handful of people willing to schedule their lives around a computer game. The problem is that you make friends, but then they're on at different times from you, or want to do different content from you. Often the game won't even let you group with them because of level differences, or at least it's strongly discouraged.
Some games have tried to assist with group formation by having an automatic grouping tool. People want a group for some content, so the game waits until there are enough people who want to group for the content, then throws them together in a group. While perhaps better than spending half an hour trying to get a group, this creates problems of its own.
One is that a lot of games implicitly assume some group composition that doesn't match the player base composition. For example, if a game assumes that 25% of group members are healers, but only 15% of the players are healers, you can get 60% of the players a group with a healer, while the other 40% can't get one as there aren't enough healers to go around. This could result in groups without a healer that have no real hope of completing content. More commonly, you end up with the players who aren't healers spending half of their "grouping" time waiting for a group to form. That's pretty much how it goes in FFXIV for players who are neither healers nor tanks, for example.
Another problem is that different players have wildly different playing styles, sometimes to the extent of being incompatible. Some players want to speed rush through the content as fast as they can, skip everything possible, get the loot, and move on. Others want to have a look around, see what there is to see, clear side stuff, and generally take their time. There's nothing intrinsically wrong with either playing style, but putting some of each in a group together is asking for trouble.
Yet another problem is not having enough players to do given content. Some content becomes unpopular for one reason or another, to the extent that it's very hard to get a group for it at all. If there are two dungeons at a given level, one of which is generally perceived to give twice as good of loot as the other, it can be very hard to get a group for the one that gives less loot. Now, double loot is bad game design, but a 5% difference can cause the same problem. Or even no systematic difference at all, but only a perception of such, as players get better at one dungeon from practice with it.
The reason this is a problem is that, if you can't get a group for half of the group content in a game, you only effectively have half as much content. Farming the same dungeon endlessly is boring, and you need as much content as possible. Adding more content that people ignore is scarcely better than not adding content at all. More commonly, games add new content with better loot than the old to get people to do it, which implicitly deprecates the old content and then you have no more content than before.
If a game is going to make group content into a major game mechanic, it needs to solve these problems. I would go so far as to say that solving the problem of where good groups come from is more important than the actual content that the groups do once formed. Plenty of games have done a decent job of the latter.
(continued on next post)
Comments
To start, we need to have an appropriate amount of group content for the game, and distribute players among it reasonably evenly. Level ranges do this to some extent, but if there are multiple options for a given level range, we need them to all be decently popular. This can be done by auto-balancing loot such that, if one dungeon is much more popular than another, we automatically decrease rewards for the former and increase them for the latter. While a designer should manually space and pace a dungeon and choose loot to try to make different dungeons comparably rewarding, this alone has no hope of getting everything exactly right. Putting some constants in to the loot tables to automatically scale loot up or down depending on the relative popularity of the dungeon over a period of time can fix this so that players are distributed roughly evenly among the dungeons.
Some games have gotten around the problem of necessary class compositions by making classes far more interchangeable. Rather than needing exactly one healer in a party of four, you grab four random players and go. Guild Wars 2 does this, for example. But if a game does want required class roles, then it needs to have alternate versions of dungeons for different required combinations of classes.
Let's take the above example of a party of four with exactly one healer, but the other three members interchangeable. If there aren't going to always be enough healers, then you create an alternate version of the dungeon that still has a party of four, but assumes no healer. Scale it so that the dungeon takes about as long for the healer version as for the non-healer version, and you get about as good of loot either way. You'll also need to have about the same probability of success. This will have to be done by auto-scaling some things in the dungeon based on empirical results of what happens when people actually play it.
What you can then do is to create however many healer and non-healer versions of the dungeon as necessary depending on who shows up in the queue. If you have 40 people in the queue when it's time to do matchmaking, 7 of which are healers, then you have 7 versions of the dungeon with a healer and 3 scaled to assume no healer. If only 4 are healers, then 4 healer versions and 6 non-healer versions. Everyone gets a group without having to wait unduly long.
Finally, there is the problem of matching playing styles. For this, I'd propose that the game keep some very extensive metrics on how people play. How much damage do you deal (scaled for level, etc.)? How much damage do you take? How much do you use purely defensive skills? How fast do you initiate the next battle after the previous ends? How much do you type in the chat? Track everything you can think of, and categorize players as best as you can. Some of the things you track will end up being very strongly correlated with others, but you can refine the model as time goes on. This sort of statistical approach will probably put a large fraction of players into several clusters for different playing styles.
Then what you do is, at the end of a dungeon run (successful or otherwise), ask players whether they'd like to group again with the other players in the group. This should be a simple questionnaire, perhaps with "yes", "no", and "indifferent", but separate for each player. This will probably lead to clear preferences of which playing styles particular players like to group with.
The goal here is not to reward popular players or punish unpopular ones. Rather, the goal is to find the playing styles that others like grouping with or want to avoid. If speed rushers like grouping with other speed rushers while role-players like grouping with other role-players, let both have what they want and everyone is happier.
When it comes time to create groups, you wait until there are enough people in the queue to create several groups. You choose some estimated probability that each player in the queue will like grouping with each other player, based on previous statistics of a player's own playing style and the playing style of others he has liked or disliked grouping with. Then you try to partition the players into groups to maximize the expected number of pairs of groupmates will like grouping with each other.
I don't know if there is an algorithm guaranteed to maximize the expected value. It strikes me as intuitively likely to be computationally intractable. But it's not hard to get a decent approximation by a finite improvement algorithm. Partition the players into groups at random, then try swapping pairs of players and see if it improves the score on how likely people are to like their groups. Swap players if it does and not if it doesn't. Repeat this until no possible swap improves the score, and that's your group partition.
This should, of course, allow players to form groups or partial groups on their own. "I want to group with player X" is a clearer statement of player intent than anything you can derive from statistics.
The advanced metrics approach isn't going to be perfect, of course. But it doesn't have to be in order to improve on picking people at random and throwing them into groups together. And there's a lot of room to improve on that.
Instead of make complex things , how about simple things with shiny skin and bikini armor ?
class roles would not really be required in this case, would they?
But it works rather the way our groups used to run. When there weren't enough guildies around, my partner and I would throw together pug groups, run some instances, and keep track of the folks we got along with (friends lists are freakin wonderful tools).
Very similar to what you're contemplating, less about "good tank" than it was about "likable person".
A mouseover "you rated Bonzo the Magnificent 8.3 over 5 instance runs" would probably help.
But give Bonzo a automatic neutral 5 if you don't want to bother rating, and hell, track every player you've ever group-encountered.
Most folks are gonna misuse it, of course. But at least it's their option to use (or ignore) the potential of the tool (like a Friends List).
If would be neat if it interacted with your 'seeking pugs' tool, "show me online players I've grouped with before, sorted by score."
The one thing working for groups in MMOs, is that players are ever increasingly enjoying group play in other genres. We have long reached past the point where players are more likely to have group play in a FPS than they are in a MMO.
MMO’s have become the online game where you don’t group. Every other genre is moving towards grouping from adventure to grouping platform games. It is like we woke up one morning in a parallel world.
For players who have hardly ever/never grouped in MMOs your experience of how good grouping can be in other genres should be enough evidence of how good it could be in MMOs.
Quizzical is showing us a way forward, to bringing back grouping without all the negative baggage that for some players it had.
Usually group make up is tank, healer, 3 dps and still it can take a few mins to get that tank. Tanking is hard work, healing/support less so and dps-ing is just easy, so most people play dps. I think tank should be rewarded somehow for doing the hard work so more people are inclined to play them.
Whenever I played FFIV ARR I always noticed when either the healer or tank was failing. The only indication that the DPS was poor was TTK and its length. Games need to place a higher burden on DPS to make it fair. This will lead to more people playing other classes I think as any game that places so much on a few roles is always targetting those roles for abuse and weariness from mechanics that always have them having to be alert and on the ball which leads to them leaving the game faster and reducing those roles numbers.
Any time a game allows certain classes to cruise through group content often even going afk while a few other classes toil through the content you have a recipe for grouping mechanics failure.
MMOs came into existence purely because of the social aspect of the game. People had zero problem gluing themselves to their keyboards for marathon gaming sessions with their friends. Hell, where did the idea of a LAN party come from?
You see, the issue has really come down to the fact that no one really cares to spend the time playing the game in the first place. Oh, they will dabble in it while they are bored... what you refer to as soloing for the most part, but they aren't really interested in playing the game as it was truly intended. If they wanted to be with friends, they'd do something else like watch the game on the big screen with a bunch of brews.
Honestly if you are dissecting the game and it's flaws, clearly the game isn't doing it for you beyond some sort of social interaction experiment. Back in the day you didn't do this, and yet, today, it has become the game.
Plenty of people still see these games as they once were and have been playing them as such for quite some time... the fact that you can't turn the hands of time back and reclaim your virginity doesn't mean the whole entire planet is in the same boat as you.
In first place why do you think MMORPG social was born ? because people with same hobby gather in one place .
That's why MMORPG dangerous as drug .
For those games, you can easily just treat other players as NPCs.
1. Role based game play aka The Trinity.
Roles are different functions that come together in order for a group to survive the wilderness and dungeons. In a group oriented game certain roles have to or are preferred are wanted in a group in order to progress. This creates forced grouping in which players are forced to wait for certain roles in the group. The solution to this is to completely reinvent the Trinity Gameplay.
The idea is to still have various viable roles for grouping but at the same time not at the cost for waiting. You've mentioned GW2 that tells you to pick up a random group and go. GW2 breaks the trinity by omitting a dedicated healer class. Which completely doesn't work because every class is a DPS class and game play becomes a zerg. Which is not the type of game play players want in a role oriented game.
The way to accomplish the reinvention of the Trinity is simply redefine the tank role. I tie in every thing together nicely in my other thread, "Reinventing the Trinity Game Play". Check it out here for more insights: http://forums.mmorpg.com/discussion/442172/reinventing-the-trinity-gameplay#latest
The solution of grouping based games are the reinvention of the trinity model as well as the core gameplay that is associated with the trinity. There also needs to be a more innovative combat mechanics that allow players to have some sort of minor to moderate utility to defend for themselves encase a healer class is not available to group. For instance, melee classes would have access to parrying and countering whereas caster and support classes could have access to healing or spell negation spells.
The point of my above thread is to have a group with any role in that composition. A group can survive without the trinity model in the wilderness or in dungeons. The only difference is that the challenge changes based on what roles are in or not active in your current group at that time. That way you're not spending hours looking for a tank or a healer. It simply means your play style has to adapt to something different. The whole dynamic of what we know of trinity game play has to change in order for this to work. The beauty of this reinvention of this core gameplay is that you don't have to necessarily have a tank or a healer to survive, simply the challenge changes.
2. The various types of players.
This is more of a variable and random as different types of players play the game differently. You've mentioned that there are players who want to rush through everything and players who want to take their time. Both of these types of players will be in every mmorpg. For me the best way to solve this issue is to create a niche game encouraging players to have a certain mentality when they play. I'll give you a perfect example.
In Everquest, things took time to be earned and obtained. Most of the time you couldn't simply steam roll in a dungeon because you had to carefully pull different NPC groups to not pull additional adds. Also, EQ had a harsh death penalty encouraging players to be on their toes and to be very careful. Thus creating an over all mentality that you have to slow down and take your time.
World of Warcraft encourages fast, instant gratification gameplay which gives the player the mentality that they have to finish everything fast. If content cannot be accomplished quickly then there is a lot of moaning and groaning. The penalty is very minute in WoW in which it doesn't create danger. I would say that there is a distinct player mentality from Everquest players to WoW players now.
If you design an mmorpg that encourages a slow to medium paced group content with penalty and danger, changes are that the player will take their time verses a player who will not.
It's very difficult to cater to a specified player who may or may not want to take their time in dungeon.
Can you explain how city of heroes was able to be a great game for both solo and grouping and many good pug experiences were in that game? Also, consider the number of variations on skill sets it was just amazing. It seems like the player base is the issue.
Epic Music: https://www.youtube.com/watch?v=vAigCvelkhQ&list=PLo9FRw1AkDuQLEz7Gvvaz3ideB2NpFtT1
https://archive.org/details/softwarelibrary_msdos?&sort=-downloads&page=1
Kyleran: "Now there's the real trick, learning to accept and enjoy a game for what it offers rather than pass on what might be a great playing experience because it lacks a few features you prefer."
John Henry Newman: "A man would do nothing if he waited until he could do it so well that no one could find fault."
FreddyNoNose: "A good game needs no defense; a bad game has no defense." "Easily digested content is just as easily forgotten."
LacedOpium: "So the question that begs to be asked is, if you are not interested in the game mechanics that define the MMORPG genre, then why are you playing an MMORPG?"
I've always thought that the way to make groups essential is to remove in-combat healing from the game. This requires a combat system that is deadly enough to discourage solo play, and simultaneously simple enough to allow most players to survive an encounter via distribution of damage. Instead of a keep-the-mob-attacking-me tank and a healer, the party would need to cycle the damage distribution between the group members. (Allowing everyone to heal makes solo play more viable). This would require a different mindset for 'agro' management, and tools for every class to both acquire and drop agro. A system like this would break the Trinity to hell and back, and many players wouldn't play simply because it lacked the Trinity. At the time I was trying to develop a game (02-03), I struggled with the niggling notion that too many people would want their Trinity-based game, and that this would crush any possibility of building any kind of customer base.
Games like EQ1 have tried to address some of the issues you raise with various mechanisms. The LDoN expansion introduced widespread instanced group-content, and was moderately successful. The problem with waiting for groups was temporarily offset by the novelty of the expansion -- everyone was there, and there were quite a few of most every class. Almost as soon as another expansion came along, these adventures were quickly abandoned. Monster missions took that concept one step further, by defining a dungeon and the characters available to that mission, but it distanced the player from their chosen class into a predetermined character who fit a specific role. They were very fun, but having your friendly enchanter delegated to a healer character this time, then that same enchanter having to tank the next time through wasn't ideal. Even less ideal was the fact that the MMs typically only provided for Tank, Healer and DPS roles, with no opportunity for Stealth or Crowd Control or other methodologies for approaching a problem.
Both LDoN and MMs tried to correct the problems of building groups, and the MMs attempted to address the level-difference issues that EQ1 had always had. Both addressed the differences between solo and group content, and controlled issues such as zerging or powerleveling. Each of these methods had some degree of success in bolstering the group game, but ultimately didn't really address the problems with grouping. (This problem has been around for some time).
In newer games, the communal encounters/quests took the approach that every player in the area could participate in the event. I enjoyed these a lot, if you were there, you were in a 'mega group' that was attacking a specific problem. This tackled the 'lack of group' issues, but the groups it made were sometimes pretty ineffective (0 tanks, 3 DPS, 0 healers). Many of these impromptu groups only succeeded due to the solo-friendly nature of those games (i.e., healing potions). If there was enough people interested in attacking the baddie, and they had enough healing potions, the players would win. Too often, one or more of the players in that impromptu group were not interested in bypassing that content, or doing something else.
Logic, my dear, merely enables one to be wrong with great authority.
Auto-group matchmaking
I also have wondered about enhancing match making so that it will have a tendency to match players you 'Liked' playing with more and less often match players you 'Disliked' playing with.
Except the flaw is that It'd be rare for me to happen to play with the same player more than once.(If the population was low it would, but then again if the game was poorly received it wouldn't have existed) So the effort to put the feature in would largely be wasted and even more so if it was in an MMORPG instead of a MOBA.
The obvious problem is, how do you statistically measure "likability", even to one particular person? Well, you keep statistics on everything that you can think of, and see what correlates well with your ground truth data of saying that you liked this person but not that person. Maybe your version of likable will be correlated with typing a lot in chat. Maybe it will be correlated with moving quickly to the next battle. Maybe it will be correlated with dealing high DPS. Any correlations could be positive or negative. But the preferences of different people will probably be correlated with different things in different ways. Very importantly, you don't need to know ahead of time what your preferences will be correlated with. And you don't need anywhere near a perfect correlation to do a lot better than random chance.
I do like your idea of showing you your previous ratings of people you've previously grouped with. The problem is that in games with large player bases, you tend not to repeatedly group with the same people very much unless you go out of your way to seek them out. Even so, "I liked grouping with person X" should trump "your statistical profile predicts a 73% chance that you'll like grouping with person X". And this approach could lead to "random" pugs placing with the same people that you like more often.
Someone who rates everyone else equal is really only hurting himself. That means that the system has no information on what he likes, and will instead only consider what others like, and stick him wherever is most convenient to accommodate others. It would likely tend to stick him with the people that others find most undesirable to group with, such as those who go AFK or otherwise vanish.
I agree with supporting various group configurations as well. Team play doesn't have to require the trilogy. Classes shouldn't be required.
There will always be bad PUGs, even in MP only games like MOBAs, but answer has always been to find a guild or group of like minded gamers if you want a consistently good grouping experience.