Mythic : Closing 43 Warhammer Online Servers

On the other hand, nothing in WoW really moves except for NPCs, players, and crafting nodes. You don’t really have to keep track of rocks and buildings if they’re always going to be in the same place. The amount of junk that gets left lying around in space in Eve is enormous.

I wouldn’t have thought the majority of the effort was in database tuning for these games. It is important for sure because if it is done wrong and you have high volumes of data e.g., early WoW, you have issues. For guys on this board who are in the biz: Is server size constrained most by database access speed, CPU or network?

Number of players that can be in the same INSTANCE (zone, whatever) of a specific game is heavily gated on things like CPU and memory. It’s also significantly affected by the number of interactions between players. This is for instance why zones that normally run fine in Eve get slow when combat starts: because things that were originally independent now start interacting with each other.

Number of players per active SHARD (also referred to as server, somewhat confusingly) is basically gated by database performance. As changes come in that modify player’s characters, the database can get backed up and start lagging. Also things like searching for auctions and related stuff can slow the database down. This is what caused WoW’s various performance problems with picking up items.

WoW and Eve have essentially opposite architectures. WoW has very high per-instance concurrency (largely because they don’t do collision or cheat detection server side), but somewhat low per-shard concurrency (hence the large number of shards). Eve is the inverse. Warhammer has DB performance about as good as WoW’s, but slower per-shard (They do zones)

I’ve posted a fair bit about DB issues with MMOs over at my largely-boring blog: http://doublebuffered.com/2006/10/30/why-sql-sucks-for-mmorpgs/ and http://doublebuffered.com/2008/02/26/gdc-08-sql-considered-harmful/ .

Sorry, you asked :)

No, I’m quite certain the vast majority of players are primarily interested in PvE. That’s not the minority. The minority (at least, so I assume) is people who don’t EVER want to do PvP. Who don’t even want it in the game for the reasons I point out above.

I remember from an article talking about the EQ servers that the vast majority of their CPU time on the server side is actually spent on physics, making sure the client is behaving properly, charting your predicted movement on the map, testing for obstructions, making sure you don’t fall through the ground (obligatory “real good job on that!”), etc. There’s lots of stuff the player could quickly come in contact with that the server has to validate.

Some MMOs have tried to save server time by relying on the client for these calculations. Thus were born speedhacks/wallhacks/etc. Hopefully this is not common nowadays. Still, if you only checked up on somebody 1/10th of the time, you might save a lot of processor time while still catching frequent scofflaws.

WoW is VERY vulnerable to all sorts of client-side movement hacks. There’s a reason they make you install Spyware to play the game.

I’m one of this minority.
I was fine with PVP shit up to introduction of arenas though I had to run AV to exalted for a ring on my rogue that I couldn’t get anywhere else (stat-wise) before AQ40 (rep reward).
However arena fucked everything up because people stopped raiding and went to get welfare pvp epix instead.
From there it went downhill and now my rogue in raids suffers because my rogue colleagues in pvp own…

Keep that shit away from raiding characters. Make a special PVP set of abilities only available in arenas at zoning in (like some pots already) or for heaven’s sake when zoning into a raid instance so people can do world pvp if they want. I don’t care just keep the shit seperated!

My pvp is Unreal Tournament not WoW.

I hated arenas too, but for a different reason. On a PvP server, there’s a different problem–the PvE raiding joins arenas and BGs in making world PvP non-viable for progression in terms of gear. Why even bother having servers flagged for PvP across the board and then give people not only opportunity but incentive to avoid world PvP by instancing, raiding, BGs, and arenas? As you simply cannot get (or couldn’t get, pre WotLK when I still played regularly) anything decent PvP-gear wise without arenas or BGs, anyone who wants to fight in the “real world,” with all of its vagaries, was simply doomed because those who ground out arena and BG rewards, or who raided heavily, dominated them.

It’s a tough problem to solve–in EQ2 PvP, raid gear was so powerful that whole guilds gave up on PvP activity to focus on PvE raiding. They tried to correct this somewhat with PvP gear, earned from token drops, but that was a mess too for a variety of reasons.

Personally I think the real solution is, on servers flagged as PvP servers, make all raid content contested. If you wanna get loot from raiding you’ll have to fight other players. The result, of course, will be nearly impossible raid progression, but that’s fine by me–if you want raid progression, the other umpteen “normal” servers will do fine.

I know, I know, it’s a pipe dream…

Actually Guild Wars is a very good example where both PvE and PvP work. What makes you say GW is a PvP only game? Most of the people I know still playing GW are playing PvE. mostly. The PvE is almost like a single player RPG in many ways.

This is not the game you’re looking for…

Players are going to flock to where the best rewards are. If two rewards are equal, they will go for the one that requires less time to achieve it. Has to be tough to design a game so that PvE and PvP have equal rewards and it takes equal effort to get them.

Total appears to be 63, when you add in the european servers: http://www.gamasutra.com/php-bin/news_index.php?story=22654

Quotes from Mark Jacobs in August 2008:

“…if you’ve seen a game consolidate servers, you know it’s in deep, deep trouble.”

“Look at us six months out. Look at us six weeks out. If we’re not adding servers, we’re not doing well.”

I stumble upon these quotes today.

I guess the earlier calling of WAR the next Daikatana, has more to do with the dev than the game itself.

MBJ makes tons of statements that he probably won’t able to fit it back to his big mouth if he were made to eat his words.

-official forum
-Adding server
-if(pop < 500k) WAR == fail.

Because the PvE and PvP are separate–you can’t be PvE’ing and run into enemies who then attempt to stop you from questing, etc. Where the PvP is segregated from the PvE, I consider that essentially a PvE game. Just my own predilections. Guild Wars’ PvE and PvP may well work fine in and of themselves (been a while since I played) but what I’m talking about is an integrated adventuring environment where you can’t separate the two.

WAR isn’t that, I know, but PvP is so much a focus of the game and PvE is so clearly designed to support that, that it matters less. And I do play on an Open RvR server, so there is some world action outside the RvR lakes at least.

This is true. But I’m not sure it matters to Mythic’s customer base. People who hate the game find evidence of hubris, incompetence or whatever else in Mark’s ramblings. People who like the game simply don’t care. Heck, I appreciate his willingness to say what’s on his mind, even if it sometimes looks pretty dumb in retrospect. Overcommunication is better than none.

By the way, Mark’s latest State of the Game message came out yesterday, outlining what is ahead:

http://herald.warhammeronline.com/warherald/NewsArticle.war?id=693

For me, the highlights are new keep maps (more ramps) and the ‘nemesis’ system. Heaps of problems remain, but Mythic does appear to be moving at a reasonable pace.

I would love to hear more detail about this since it is what many of us have deemed neccessary to really bring RvR to life. BOs that are relevant and worth fighting over, and not merely renown/influence generators.

Yeah, that’d be good. Most of the time, BOs don’t even get hit until the keep is taken, which I think is exactly opposite of what was intended. The theory seems to be that taking the BOs only reduces some of the NPC presence at the keep, which is trivial to a warband anyhow, but it alerts the defenders to the need to get to the keep. And as defended keeps are much, much harder to take, no one takes BOs until the keep is taken.

From the v1.2 (3/3/09) patch notes:

Zone Control Domination

To help place more emphasis on Keep claiming and defense, as well as to combat the “defense by not defending” strategy, we are introducing the Zone Control Domination mechanic. With this system, players who take and hold all Battlefield Objectives and claim Keeps in a contested zone (in Tier 4) or a shared RvR lake (in Tiers 2 and 3) have the opportunity to capture that zone, forgoing the standard mechanics of Zone Control. Each Battlefield Objective and Keep will be worth 1 Domination Point. To capture a zone via the Domination rules, your Realm must have all 6 Domination Points. In Tiers 2 and 3, capturing all six points in a shared RvR lake will give the capturing Realm control of the entire Tier. These are represented as small pip marks on the Zone Control Bar. Battlefield Objectives and Keeps have different requirements for earning their Domination Point, as follows:

  • To earn a Domination Point from a Battlefield Objective, it must be owned by your realm for 30 minutes.
  • To earn a Domination Point from a Keep, it must be claimed by a Guild and then held for 2 hours. If you lose control of a Battlefield Objective or Keep at any time, you lose the Domination Point.
  • Multiple Domination Points cannot be gained from a single source (for example, it isn’t possible to get 6 Domination Points by capturing Martyr’s Square 6 times).
  • If a zone that was taken over by Zone Domination is put into a contested state, all timers and Domination Points are reset.