The Massive Gap

As a follow-up to the post the other day about the largest online battle ever, and as an additional data point relating to Syncaine’s post yesterday, I present the following video. Go ahead, watch the whole thing, and pay attention. I’ll be right here.

The first thing I’d call to your attention is not the experiment that the video is about, but the numbers involved. Roughly 700 participants took place in this battle, which was (much) smaller than the epic struggle in LXQ2-T. Now I’ll point out what you didn’t see. Lag.

Seven hundred players, plus their drones and missiles (Drakes are Caldari, remember?) fighting it out in a single system, without lag. I’ve often defended WoW against Syncaine’s various assaults, but on this I agree with him – Blizzard’s failure to get even moderately-scaled open-world PvP working is nothing less than an embarrassment. But it’s not Blizzard who should be most embarrassed – it’s everybody other than CCP (and maybe Adventurine.)

It’s indicative of the larger problem in MMO development, which is the lack of will on the part of most developers to push technological boundaries. The battle for LXQ2-T was massive, both because of the number of players involved and the fact that they were all in a frefight. How massive is it for 100 players standing around, browsing the Auction House and dancing on mailboxes to crash Ironforge? How massive is it when a numerical advantage of a few dozen is enough to throw the factional balance of an entire Warhammer server?

The “massive gap” is widening between EVE and the rest of the MMO field (with a couple of weird outliers like Champions Online and Darkfall.) And it’s narrowing between the MMO field and conventional multiplayer games, even titles like shooters which are traditionally resource hogs. No wonder these types of games are calling themselves MMOs, nowadays – in their fields the technology is still advancing; they’re getting more massive all the time. Only in the MMOs that pioneered the idea are developers not even thinking about this anymore. Everybody assumes that the way WoW does this is perfect, without even asking themselves whether it might be desirable to actually live up to the massive name. Mythic decided 2500 players on a server with three-quarters of them blocked out of the open world by instancing was massive enough, but their game was dependent on being massive, so it goes down as perhaps the most epic fail in the history of MMOs thus far. Even bigger than Vanguard.

7 responses to “The Massive Gap

  1. Again, I’ll point you to (I think it was?) Scott Jennings who was among the group who created the MMO acronym. It simply meant a way to separate them from “normal” multiplayer games which at the time (and for the most part, still) supported up to 64 players, with most doing between 12-32 players. Is 100 players at the AH “massively multiplayer?” Apparently yes. Also, “massively multiplayer” does not mean “all within a few feet of each other at the same time.”

    EVE is kinda a special beast though, and I’ll take a few points off that people don’t consider. In EVE everything is done via a menu. You click the menu, the command gets sent to the server, and the server does it for you. In WoW, you physically control the character with WASD, spacebar, etc. Add emotes, spell effects, are you jumping, spinning, walking, running, etc. as well? That’s a ton of data for the servers to handle and we’re just not to a point in tech where it can be done easily and still transmit that data to players on often limited bandwidth. Not to mention PCs themselves simply cannot render that many players, polygons, particles, etc. independently without crashing yet. In EVE you just see dots, some ships, and some pew pew. Easier to render more of that.

  2. I didn’t say it was easy. I’m aware that there are technical challenges, some very difficult to surmount. What I’m railing against is the failure of practically everybody to even try.

    Aside from that, there’s also (in EVE, where you can and frequently do pilot manually in combat,) you have stuff like transversal velocity to calculate, distance down to the meter to measure, missiles and drones which react to target movement, and “animations” in the form of weapons firing and actually impacting on the target, smart bombs going off, explosions with debris, etc. That objects may be distant and therefore just dots isn’t any different from characters not visible because they’re in a different part of the zone. I’m not buying that the menu-driven UI is making up for this. In fact, I’m not buying that the menus are even a factor at all; in either case, the server and client are tracking position and movement independently.

    I don’t want to softpedal the challenges involved; CCP had to divert the high-powered Jita node to LXQ2-T to keep the server from going up like Bikini Atoll. It wouldn’t be correct or fair to say that this kind of thing could happen anywhere in New Eden, or at any time. But still – it’s an entire WoW server worth of players, or more, all in one place and shooting at each other.

  3. That video really is impressive as hell. I haven’t been in enough really big battles in other MMOs to know what the limits really are in most of the ones I play. I was in a very big city battle in CO at the end of the open beta. There were something like 60-100 players per shard all with utterly random costumes. The lack of lag there really impressed me. I have also been in battles that had two full raid groups on each side in the moors in LoTRO, that’s about 100 players altogether, and the client and server performance impressed me. However, I don’t think I’ve ever been in a battle that had anywhere near 700 participants. I expect most MMOs would melt under the strain.

  4. The 2500ish players per server limit isn’t necessarily technological as it is a design challenge. If you somehow scaled WoW’s server technology up such that you could have, taking a number at semi-random, ten or twenty thousand players all doing the things players do, some of the game systems would still die horribly when exposed to such heavy use. It requires a different sort of world and a different sort of idea of what constitutes (especially non-instanced) group content when you have that many players in one ‘world’.

    I’d argue that that’s part of the problem with WAR; they took their PvE model from WoW without really thinking about the implications of that model for the PvP experience.

  5. In my most recent post at Nomadic Gamer, I noted at the end that I was very favorably impressed by the improvements in lag I’d seen in just a few short months. The fleet for that operation was over 700 people, and while there was a couple of seconds of lag, I didn’t think it was bad at all.

    2 months before I was part of a fleet of about 230 and lag was so bad that over the course of an entire battle that lasted about an hour I managed to get my guns to fire a total of 12 times. That’s a TON of improvement in only 2 months. And they’re still working on it too, so it should only get better.

    Joke all you want that CCP stands for “Can’t Code Properly” — I think they’re doing an amazing job. I have VERY high hopes for Dust 514 and WOD as a result.

  6. This is very true, I’ve been around playing MMO”RPG”s for years (Eve-Online for five in fact), and I can say without a doubt that most development studios present an attitude of, “That’s good enough.” In contrast to that though, CCP keeps trying to push the edge of what is capable and possible for sever technology, and especially a single-server environment. Now granted there are some screw-ups and a few instances where someone forgot to feed the hamsters, but as Ardwulf points out, at least they are trying. Most don’t even bother.

  7. Pingback: No news is good news….right? « Sighfigh's Gaming Blog