The Hard Problem: The New Concurrency

John Scott Tynes

New member
Dec 31, 1969
69
0
0
The Hard Problem: The New Concurrency

MMOGs brag about how many people they have all playing at once, but the real king of concurrency is something completely different.

Read Full Article
 

Warrior Irme

New member
May 30, 2008
562
0
0
I really do miss the good old days of running 40 man raids with 2-3 people constantly disconnecting and reconnecting in the middle of boss fights. I want an MMO to think in a greater number than this at some point. A Player vs enemy encounter that would require 60 people could be fun. Make it easy enough that you can grab a group of random players, but difficult enough to still have a good time with a wipe or two.
 

ckeymel

New member
Jun 24, 2008
234
0
0
Warrior Irme said:
I really do miss the good old days of running 40 man raids with 2-3 people constantly disconnecting and reconnecting in the middle of boss fights. I want an MMO to think in a greater number than this at some point. A Player vs enemy encounter that would require 60 people could be fun. Make it easy enough that you can grab a group of random players, but difficult enough to still have a good time with a wipe or two.
I think the success or failure of MAG will decide the fate of having that happen. It's so hard to organize that many players, let alone be effective with them that it's pretty much essential to have some sort of re-pore with those you are raiding with, so it would work with large raiding guilds, but the average person will end up probably getting overwhelmed. Not that that should be an excuse to not do it - but obviously it would appeal to certain people.
 

G-Mang

New member
May 11, 2009
92
0
0
I can't help but get this feeling that "proximal concurrency," as described in the article, isn't significantly more important than absolute concurrency, if significant at all.

If you're in a 1v1 poker tournament, you're only playing against a single winner of their bracket. The other players who lost to that player had a part in what led to that person's status, but that's a weak connection at best. I could make the same connection for every single person who interacted with an MMORPG character, as every world PvP, clan raid, and assisted leveling spree in which that character participated had in impact on that player's current status (same with every person who participated in the same auction as that character, or even just every person who wasted some of that player's leveling time). Requiring neither simultaneousness, spacial proximity, nor game mechanic impact to be proximally concurrent makes it seem worthless to me. It'd be like saying "if I get a high score in a single-player arcade game, I'm taking part in proximal concurrency with every single person who's ever used that arcade machine, because we're technically competing for the same goal: a high score." It's basically meaningless.


In my opinion, meaningful concurrency should only take into account players who have a direct impact on your present status based on how they interact(ed) with you. If the actual way someone interacted with you doesn't matter (for example, if their only impact on your route was winning or losing), then their contribution isn't presently relevant to your game decisions, and if you accept indirect impacts or impacts that affected previous status, then you basically need to call every particle in the universe a concurrent player.
 

Resin213

New member
Jan 22, 2009
61
0
0
This article made me think that proximal concurrency is a hugely variable term. If I buy a lotto ticket am I competeing against every other player that bought a ticket? I liked that he got into differences in location proximal concurrency vs. temporal proximal concurrency, but I think the depth of a game should be taken into an account as well, as we see in his disclusion of mafia wars. Perhaps some sort of quadratic plot of game depth vs. proximal concurrency, vs. mentions of overused internet/game memes will somehow give us new better vocabulary when discussing levels of concurrency.
 

Azmael Silverlance

Pirate Warlord!
Oct 20, 2009
756
0
0
I think we should just develop our technology so we can hold tousands of players at the same zone at the same time n start having some major PvP battles :D
Also id love sizes of maps to increase...imagine Orgrimmar with literally 10,000 buildings...truly an epic City. N ppl going around opening shops..buyibg houses.....*drool...
Tho i guess the internet connection should be somehow improved aswell.
 

tendo82

Uncanny Valley Cave Dweller
Nov 30, 2007
1,283
0
0
MAG is interesting because it plays with notions of both spatial and temporal proximal concurrency. Temporally, its proximal concurrency can definitely be said to max out at 256 players simultaneously working with or against each other. Spatially, I think the game the game works within a smaller scope since people are organized in squads of 8, which often results in 8 on 8 encounters.

This is when the game works best: when squads are fairly divided (geographically) and mobile. At moments when MAG achieves something like even 128 person (64 on 64) spatial concurrency, the game breaks down. You end up with a bottleneck and a lot of respawning. This has been my experience anyway.

So maybe another hard problem is what kinds of genres/accepted play mechanics can actually support massive spatial concurrency? It'll be interesting to see, as the community matures, whether or not MAG can eventually make a true D-Day-esque 128 on 128 person battle work, or if the game will only work as isolated 8 man squads on a large map pushing towards isolated goals, with only occasional cross over of more than two squads.
 

Geoffrey42

New member
Aug 22, 2006
862
0
0
On the topic of MAG: The way the game is designed, it seems like they could do 1024 players, and the experience for a single player in that environment will not change. They way they split things up, you're typically in an 8v8 to 32v32 engagement (in the 256 player battles, there are essentially 4 fronts being fought simultaneously, with some overlap at the borders). I like your argument that the level of concurrency that matters is how many concurrent players are actually interacting at the same time with you in regard to something, but I'd want to further narrow it to how many concurrent players are actually having an impact on what I'm doing. In the case of MAG, that seems to top out at 64 (myself, my 31 proximate teammates, and my 32 proximate enemies). MAG provides a sense of scale, and those other 192 players doing well or badly can make me win or lose, no matter how well I play, but the actual experience of playing is pretty much just with 64 people.

With regard to the article: For some reason, I think 1vs100 is a better example to support your idea of alternative approaches to large concurrency, because I'm directly competing with those other thousands of people, for the same goal, at the same time. But if you want to count the unique players across the entire week during extended play, you could make that number bigger (under my understanding of the Zynga Poker example you put forth) because every game played factors into your overall likelihood of being in the 100, or to be the 1. We're all simultaneously clawing at the leaderboard, and the One can score some funny-money if they do well. To come at it from another way, I could get (theoretically, not actually) 6 billion people involved in a Chess tournament. They would all be playing alongside each other, and at the end, there would be a winner. But is that really worth calling 6 Billion Player Concurrency? Does it alter or enhance the game of chess? It's just playing a bunch of chess games between different pairs of people. The existence of a leaderboard, or a weekly progression, doesn't seem to me a key to useful/worthwhile concurrency. WoW and MAG and EVE all have persistent overarching contexts (PvP leaderboards, territorial concerns), but you've already dismissed those aspects of their interactions as counting toward concurrency. Why does Zynga Poker's asynchronous level of interaction get counted? Because it has prizes?

Sorry for rambling, but one last thought: I feel like for concurrency to matter, it has to have an impact on the way I play a game. In basketball, or Halo, player behavior in a 1v1 seems very distinct from 2v2 and 5v5. But, past a certain point the game either ceases to function (a 16v16 basketball game would look like little-league soccer, I imagine), or it begins to act as a larger morass of smaller games playing out with small-game dynamics, vaguely interacting at their fringes (like MAG). At that point, does increasing concurrency matter? Is what I've described as "worthwhile" concurrency inherently limited by something akin to [a href=http://en.wikipedia.org/wiki/Dunbar%27s_number]Dunbar's number[/a]?
 

That One Six

New member
Dec 14, 2008
677
0
0
I believe that Runescape's "There are currently X people playing." counter has reached 200,000 in the past.
 

Jared

The British Paladin
Jul 14, 2009
5,630
0
0
That One Six said:
I believe that Runescape's "There are currently X people playing." counter has reached 200,000 in the past.
I remember seeing that.

And I do agree, its not the number that play a game at all that make it good!
 

Fearzone

Boyz! Boyz! Boyz!
Dec 3, 2008
1,241
0
0
Awhile ago, in Arathi Basin, while I was capturing a flag, a thought crossed my mind that the mechanic of strategic points there was very similar to Dawn of War, except each one of us was an independently acting RTS unit. Instead of resources for expanding our army we were getting points toward a final winning score, but it was close enough.

So, I had a thought the other day. What if you make a whole MMO like that? Except instead of a Dawn of War skirmish, make it huge, more along the scope of Sins of a Solar Empire, or any turn-based strategy with RTS elements, where across all zones people might be playing for points or resources or whatever toward and ultimate victory. It could be a PvP horde vs. alliance or such, or it could be PvE with the server spending resources to tech and spawn mobs to take over the map. Everyone across the realm would have to cooperate to win, either that or they could just go about the usual MMO stuff but avoiding enemy occupied territories.
 

Spritzey

New member
May 18, 2009
47
0
0
Fearzone said:
Awhile ago, in Arathi Basin, while I was capturing a flag, a thought crossed my mind that the mechanic of strategic points there was very similar to Dawn of War, except each one of us was an independently acting RTS unit. Instead of resources for expanding our army we were getting points toward a final winning score, but it was close enough.

So, I had a thought the other day. What if you make a whole MMO like that? Except instead of a Dawn of War skirmish, make it huge, more along the scope of Sins of a Solar Empire, or any turn-based strategy with RTS elements, where across all zones people might be playing for points or resources or whatever toward and ultimate victory. It could be a PvP horde vs. alliance or such, or it could be PvE with the server spending resources to tech and spawn mobs to take over the map. Everyone across the realm would have to cooperate to win, either that or they could just go about the usual MMO stuff but avoiding enemy occupied territories.
Good point, and I would love to see that. However at such a scale your contribution, in my mind, would be too little. So what's the motivation in playing?
 

Fearzone

Boyz! Boyz! Boyz!
Dec 3, 2008
1,241
0
0
^^The usual... experience points... maybe some area or realm-wide bonus for a captured area... reputation points... plus some added bonus like honor points toward gear if you win. Not unlike battlegrounds.
 

Fenixius

New member
Feb 5, 2007
449
0
0
As always, cool article. I love The Hard Problem. And, as always, I disagree with a whole bunch of stuff! So, let's get to it.

I would definitely argue that Zynga Poker does NOT constitute important proximal concurrency. Even if my chances of winning a prize are affected by 200,000 people, the way I play the game does not. The only thing which can change the way I play are the 7 other people at the table. I think there were 8 people in a game of poker, anyway. But the point remains - they don't affect my game experience just like everyone in eBay does not affect my auctioneering/bargain hunting experience. Only those people -interacting- with me do. Not just "affecting".

This is one thing that seriously bugs me about most Massively Multiplayer Online Games. Especially the RPG ones. There's so little proximal concurrency, and it's usually more of a pain for me to go find people to hang out with rather than just go solo. It's why I prefer team-based games like Team Fortress to a team-if-you-like system like Champions or World of Warcraft.

In something like TF2, what I do now is directly impacted by my teammates. I'll have to heal to help with this assault. Or go Spy to take out that rogue sniper behind us, because all our Heavies are dying. Or whatever. World of Warcraft's 5-/10-/25-/40-man instances are better than most of the game, because you're directly working with a team, but I find that until you hit 80, those are pretty uncommon occurances. I know Champions didn't even try to make use of them at lower levels.

As you start increasing proximal concurrency, the impact each player can possibly have is lessened. Spritzey just pointed it out, talking about Fearzone's game idea: What's the point in playing if you don't make a difference? If you spend all day trying to get this hard-to-capture point, and finally get it, but overnight a thousand enemy players carve a bloody red hole through your bright blue nation? There has to be a game which you can have fun in with that huge overarching goal only sitting behind it, not just a game where you work on a goal. Players would rather feel important than insignificant, and so game design has followed. Excepting technical limitations, I think that's the main reason games tend to work more on the basis of low proximal concurrency, with potential non-immediate bonusses if high server concurrency happens.

With regards to "temporal" proximal concurrency, I'm not really clear on how it's important at all. I mean, I could just go and play Space Invaders or something, and post a score to the leaderboard. Are you trying to tell me that my competition with every other player playing the game is important? Because I can't see how it is, at all. Their actions don't affect me, except possibly to egg me on to get a higher score and waste more of my life competing against people who don't have one to waste.

Finally, what's the point of increasing proximal concurrency? How does it make the game any better to play? Let's take MAG as an example, since lots of people like it. Even if we're only talking 64p games. Does having 32 enemies really make the game more interesting than having 16, or 12? You're more likely to be sniped from some obscure location. You're more likely to have your kills stolen by your own, larger team. There will be more spamming of ranged attacks. I'm mostly speaking from my experiences with Team Fortress 2, here, which goes from very small 3v3 games up to about 16v16. And those last, the 32p servers, are just unbearable for me. So I don't get it - what's the appeal, here? I understand the MMORPG thing, where you can hang out with heaps of people, you can play cooperatively with people, but player versus player with huuuuuuge teams is just no fun at all.

So, tl;dr:
-Temporal concurrency doesn't mean much.
-Players like to feel important rather than insignificant, so increasing concurrency (in any way) isn't necessarily good.
-Most so-called "Massively" Multiplayer Online RPG's tend to not do much with immediate proximal concurrency.
-Why are we trying to increase proximal concurrency again? I hate kill-stealers and ninja-looters!

Thanks for the discussion, everyone, and to John for writing the article.
 

copycatalyst

New member
Nov 10, 2009
216
0
0
It seems that if you include the poker game because it has weekly competition over the same prizes under the heading "MMOG," you would also have to include any game with a shared leaderboard. This would dilute the term until it had no meaning. If you take out the "Online" from "MMOG," you would also have to go back to the physical arcade, where a single machine (such as Space Invaders, as posted above) might be played by a few hundred people, competing for the glory of putting three letters at the top of a high score list. But asynchronous competition is not the same as [/I]multiplayer.[/I]
 

econael

New member
Apr 15, 2009
37
0
0
I agree with most of the commentors.

I've had more interaction with other players in a dozen LoL matches than in several months of EVE. 'nuff said
 

Wolfrug

New member
Feb 11, 2009
57
0
0
Fenixius said:
As you start increasing proximal concurrency, the impact each player can possibly have is lessened. Spritzey just pointed it out, talking about Fearzone's game idea: What's the point in playing if you don't make a difference? If you spend all day trying to get this hard-to-capture point, and finally get it, but overnight a thousand enemy players carve a bloody red hole through your bright blue nation? There has to be a game which you can have fun in with that huge overarching goal only sitting behind it, not just a game where you work on a goal. Players would rather feel important than insignificant, and so game design has followed. Excepting technical limitations, I think that's the main reason games tend to work more on the basis of low proximal concurrency, with potential non-immediate bonusses if high server concurrency happens.
Not all players' fun lies in feeling like the most important cog in the wheel, you know. I personally think the best game designs are the ones in which the player is rather made to feel like a smaller, but well-working part of a larger whole. I remember a mod for the original HL way back when called something like Firearms (?), which was basically class-based team combat à la TF, except there was a system of "lives" included - whenever one of one's own got offed, a "life" was removed from the team counter, until they were all gone (and your team lost). There were other objectives too, capture and hold, capture the flag bla bla, but the lives were always there - no lives, no win (I think you could also replenish these lives by completing objectives and stuff like that, but my memory is fuzzy).

Point of the matter is, I played a medic. A medic could heal, but a medic could also go up to dead-but-not-really corpses of friendlies on the battlefield and 'evacuate' them, which would return the lost life to the pool. This was not an important role, really, most were just in it for the shooting and the fragging, but I really enjoyed playing support. I felt that I was making a small difference in the big whole. I wasn't a half bad bunny jumper either, so I could've easily made a much bigger difference for the team if I'd just run at them guns blazing. But I didn't want to do that, since then -no-one- would've played medic.

There's a whole bunch of people around who like playing support, who like helping out behind the scenes while the 'heroes' rush in to cap flags or shoot people or whatever it is they're doing. That's why most class-based combat games HAVE classes like Engineer or Medic or whatever, to allow for us tinkerers and repairers and healers and so on. I personally get a lot more enjoyment from supporting my immediate ally, group, army, nation, whatever than I get from leading or feeling like a hero in said group. So yeah, if someone comes in and takes over the blue zone I had just conquered the next day - bring it on!

tl;dr: People don't always like being the hero, people also like being in support roles - and really, -everyone- should support everyone else rather than there being some kind of weird divide. The end.
 

AC Medina

New member
Oct 12, 2009
238
0
0
I read the byline and immediately thought "Assuming control" ...which, I suppose, just means I've been playing way too much Mass Effect 2.
 

Phishfood

New member
Jul 21, 2009
743
0
0
Skimmed other posts so I can type before I forget, but I'd argue that EVE does have some of the best conccurency, since everyone that plays IS bidding on the same items, competing for the same asteroids to mine etc.

"Hulkagedon" managed to double the price of hulks for all players.

<edit, wrong button :p>

The point being that in EVE all 250,000 or whatever it is subscribers DO interact in a roundabout way, whereas the 2M people playing mafia wars probably never see each other.

I would love to see an FPS with 100 on 100 battles in a nice open map (think battlefield vietnam) but such a thing would probably never sell since it would probably require relatively poor graphics by today's standards thus putting off all the "I get 10 billion FPS with my quad SLI tower that uses 3 1,500W power supplies" tards.

The solution to managing that many players is quite simple, battlefield did it, eve does it and the real army does it. Chain of command. Someone in command who orders platoons/wings around and then someone in each wing/platoon that orders squads around and someone in each squad who orders players around.
 

veloper

New member
Jan 20, 2009
4,597
0
0
I don't care for this proximal concurrency.
Optimal is 6 guys on your team and 6 on the other team. Or just 6 PCs for PvE.

When you get to 2 digit teams, your character becomes too insignificant and gameplay becomes a grind.

Server concurrency matters alot because the MMO depends on it.