Random Arena

[pxib]

I was going to put this in the Guild Wars forum over in the graveyard, but I thought it might have broader implications.

Casual Guild Wars is dead. I'm unbelievably lousy at FPS and RTS games, so this was the only game I'd found where I could just pop on for half an hour for and have some PvP without merely being a target and a rube.

No more.

The Random Arena, once home to fun experiments and crazy times is now composed entirely of two groups, people who play the twenty or thirty builds that work the best and the unwashed masses. This means that somewhere between eighty and ninety percent of random groups will die in the first round. Why? The 30 second queue times and dwindling player population.

A team who randomly gets enough capable players and builds to win will win their first round and move to a second round. A team who randomly gets a bunch of awkward scrubs will die and... go back to the queue. Rapidly all the worthwhile players are removed from the genepool and remain outside it for the five to ten fights required before they either meet a MORE capable team or get removed to the Team Arenas. Almost everybody stuck in the queue is going to be crappy. If they're lucky they'll get another group of losers first round and beat them, only to die in the second round when they meet a group of winners already beating their sixth group of losers.

So first question: Is there a better way to arrange fast, casual random team fights?

Worse yet, even though there are now more than a thousand skills, and probably millions of combinations thereof... there are only twenty or thirty builds that work well in a random team. Sure, if you have time to set up a group in TA with some synergy you can create many more unusual setups... but they will be crushed by the ten or so Team Arena group builds that actually kill everybody. That's been figured out. It's been posted and archived. The game is over and somebody won.

Second question: Must a competitive skill-based system necessarily devolve into classes?

[caladein]

Second question: Must a competitive skill-based system necessarily devolve into classes?

Yes, it is a simple consequence of the number of iterations your players are going through, fueled by the competitive environment.  I don't quite see it as a negative though (having participated in essentially the creation of a PvP meta-game, it was quite organic, and fun), assuming you're satisfied with the amount of variety in your meta-game.

There are a finite amount of things you can do with a health bar, a power bar, and two-dimensional movement.  I'll make the assumption that one person can't do it all, so you need to specialize.  Presto, you have proto-classes.  At that point, you're only looking at iteration cycles and encouraging variety by nerfing combinations that are too powerful.

So first question: Is there a better way to arrange fast, casual random team fights?

From a skills perspective, you will have players that simply play the percentages as to what is the most likely build that will synergize with what the unwashed masses usually do.  (The answer will almost always be a healer.)

[Stephen Zepp]

Second question: Must a competitive skill-based system necessarily devolve into classes?

Yes, it is a simple consequence of the number of iterations your players are going through, fueled by the competitive environment.  I don't quite see it as a negative though (having participated in essentially the creation of a PvP meta-game, it was quite organic, and fun), assuming you're satisfied with the amount of variety in your meta-game.

There are a finite amount of things you can do with a health bar, a power bar, and two-dimensional movement.  I'll make the assumption that one person can't do it all, so you need to specialize.  Presto, you have proto-classes.  At that point, you're only looking at iteration cycles and encouraging variety by nerfing combinations that are too powerful.

I disagree with your assumption that "one person can't do it all, so you need to specialize".

The game mechanic as implemented in most games of this nature is what forces specialization, but that is a design choice, not a "law of gaming". Planetside for example uses an underlying mechanic that does force some specialization, but not to the extent of GW, and there are game mechanics that can allow full flexibility (no specialization required or forced), that remove the actual or "meta" concept of classes and "ability sub-sets" that are inherent in the GW model.

[Samwise]

Second question: Must a competitive skill-based system necessarily devolve into classes?

Only if you define "classes" sufficiently loosely and your skill system is at all interesting.  I.e.:

1) There are clusters of skills that synergize well together.
2) All (or most) skills are useful enough to be worth having.
3) It is not possible to actually have all skills.

The above will tend to push people into getting certain skill clusters (because of 1).  Some people will get some clusters and some will get others (because of 2 and 3).  If you want to call people with those different clusters members of different "classes", then there you go.

Of course, you can "fix" the problem (if it really is one) by making sure each skill functions entirely independently of the others (i.e. your "skill tree" has a depth of 1), or by making most of the skills sufficiently useless that it doesn't matter whether you have them, or by letting each player have all the skills he wants.  Most people would consider that to be a pretty boring system, though.

[cmlancas]

Why don't the divisions of random PVP get tiered? If a group of people is rolling over everyone, send them to tier 2, and so on and so forth.

[Stephen Zepp]

Of course, you can "fix" the problem (if it really is one) by making sure each skill functions entirely independently of the others (i.e. your "skill tree" has a depth of 1), or by making most of the skills sufficiently useless that it doesn't matter whether you have them, or by letting each player have all the skills he wants.  Most people would consider that to be a pretty boring system, though.

I used to agree with the point I highlighted in blue above, but I've changed my tune in some ways.

I remember a long and drawn out discussion I had with a wiz in Shadowbane regarding balance, and he suggested that perfect balance was letting all players use any combination of skills they wanted, with no restriction. Effectively, removing the concept of classes, and especially "flavor of the month" templates by removing the need to have to even pick a class or template.

I disagreed strongly at the time, mostly because I felt the need for the complex strategies involved in large scale fights based upon different "combat functions" filled by different classes--i.e. damage dealers, damage takers, healers, support, effect, etc. I couldn't see how a game could remain fun and interesting (to me, with the assumption that the complexity and interactions between roles is what was fun) when you didn't "force" roles based on class/template decisions made prior to the fights.

I've evolved from that stance in some ways--I still see the need for roles, but I think those roles will evolve naturally, and don't have to be forced on players via a game mechanic...instead, players should be able to "auto-select" roles based on what skills they actually use in a given situation, instead of limiting the skills they have available in a given situation.

[Samwise]

Of course, you can "fix" the problem (if it really is one) by making sure each skill functions entirely independently of the others (i.e. your "skill tree" has a depth of 1), or by making most of the skills sufficiently useless that it doesn't matter whether you have them, or by letting each player have all the skills he wants.

letting all players use any combination of skills they wanted, with no restriction.

What I said and what you said aren't the same thing.  Having all the skills isn't the same thing as being able to pick some subset with complete freedom.  If certain subsets of skills make more sense or are more useful in certain situations, you will have "classes".  At least as the OP seems to mean it.

My take is that if you define "classes" so loosely that they're completely player-defined, there are a thousand of them, and they're all fun to play, I don't think there is really a problem.

[Stephen Zepp]

Of course, you can "fix" the problem (if it really is one) by making sure each skill functions entirely independently of the others (i.e. your "skill tree" has a depth of 1), or by making most of the skills sufficiently useless that it doesn't matter whether you have them, or by letting each player have all the skills he wants.

letting all players use any combination of skills they wanted, with no restriction.

What I said and what you said aren't the same thing.  Having all the skills isn't the same thing as being able to pick some subset with complete freedom.  If certain subsets of skills make more sense or are more useful in certain situations, you will have "classes".  At least as the OP seems to mean it.

My take is that if you define "classes" so loosely that they're completely player-defined, there are a thousand of them, and they're all fun to play, I don't think there is really a problem.

Ok, but are you describing a stage or phase of character development where the player "picks a subset of skills" and then goes out and does stuff, with the implication that they are "locked" into that subset for any duration of play time until they go through the stage of "picking a new subset"?

What I'm positing here is removing that defined, atomic stage of character development where they have to "pick a subset" completely. If they find themselves in a situation where they are saying to themselves "damn, I wish I had that heal spell loaded", then I agree, in some meta-definition level you have classes.

If they can say "ouch, that hurt, gonna heal a sec" (insert ANY possible use of any skill in any scenario), then the 3rd assumption you made is a good one, but I differ in thinking that "most people would consider that a pretty boring system".

[Samwise]

If they can say "ouch, that hurt, gonna heal a sec" (insert ANY possible use of any skill in any scenario), then the 3rd assumption you made is a good one, but I differ in thinking that "most people would consider that a pretty boring system".

The game itself isn't necessarily boring, but your "skill system" doesn't seem that exciting if all players have access to all skills at any point.  Usually a skill system is meant to break down "this player can do X but not Y, and this player can do Y but not X".  If all players can do X, it doesn't seem worthwhile to call X a "skill". 

Under that definition I could call chess a skill-based game, where the ability to move each piece is a "skill".  I submit that the "skill system" in that game is boring (although the manner in which those "skills" are used is not).