sonderkrabbe's page

The biggest gripe that I hear from people trying out EVE is that (up until CCP began to adopt a more adversarial approach to their customers by trying to force features on them, wanted or not) the game seemed built completely around a core group of players that had been in the game for what seemed like forever, and all of the less tenured players really had no option other than becoming slaves to them.

Prior to the release of the Apocrypha expansion in 2009, EVE had 3 distinct areas of play:

Hi-sec: where random aggression against other players resulted in the automatic death of the attacker (after a variable length of time measured in seconds, depending on various factors)

Lo-sec: where aggression against other players was not an automatic death sentence (although you could be killed by NPC enforcers if you weren't fast/careful enough) but territorial control remained within the hands of the NPCs

0.0 (sometimes called "Null-sec"): no enforced rules of behavior other than the rules enforced by the players themselves, where territory was controlled by players

Hi-sec was the most populated area. Most characters in Hi-sec were engaged in commerce - resource extraction, crafting, logistics and arbitrage. Many characters in Hi-sec also were "mission runners", flying ships optimized to earn the most in-game currency possible per second by engaging in PvE.

Lo-sec was the least populated area. Pirates operate freely there and are always watching for opportunities to ambush and either destroy or ransom ships. There are some valuable resources in lo-sec and the highest value missions are often found in lo-sec, so non-pirate characters have some incentive to traverse it. The pirates complain endlessly to CCP that lo-sec is "broken" because there aren't enough viable targets for their piracy.

0.0 is the second most populated area. EVE is designed around a series of "jump gates" which limit travel between systems (you can't just fly off into space and eventually reach a new system). As a result, there are places where guarding one or two chokepoint systems can create a border dividing 0.0 into isolated kingdoms. Even so, these kingdoms are not truly static. You can watch a time-lapse animation of the ebb & flow of 0.0 Alliance territorial control here:

http://www.youtube.com/watch?v=5m4q2-gbBUE&feature=related

At the start of the animation in 2007 all the 0.0 systems are held by Alliances. Each Alliance has its own color. As the animation progresses, you can see slight changes at the edges of some of the Alliance territory. It doesn't look like much at this scale, but taking a single system from an entrenched Alliance is a work of thousands of people over days and often weeks of time.

If you wanted to play in 0.0 space, you needed to either join one of these Alliances, or play in the area controlled by the Alliance known as CVA. CVA was unique in that it allowed players to exist within its 0.0 space without requiring them to join CVA (this was called "Not Red Don't Shoot", or NRDS. "Red" referred to the color of a ship's icon in the game's HUD indicating a pirate or a character engaged in hostile acts. As long as you behaved yourself CVA would tolerate your presence within their territory.) Everyone else in 0.0 space had a policy of "Not Blue, Shoot It" or NBSI. "Blue" is the color of a like-aligned ship on the HUD. If you weren't in the Alliance, you were a target.

There are a couple of differences between 0.0 space and Hi-sec. First, there are ships and structures that can only exist in 0.0 space. If you want to build and fly the biggest, most powerful ships, you have to do it in 0.0. Second, there are some resources in 0.0 which are used to make the highest-end crafted objects in the game. If you don't want to buy those resources at the market price, you have to be in 0.0 to extract them. Those resources are the start of a logistics chain that produce the best ships and best ship "modules" (stuff you equip your ship with), so there's a feedback loop where 0.0 Alliances controlled the means of supply for the stuff needed to fight 0.0 Alliances.

Even in this seemingly static world came change. An organized group of players from the Something Awful community came to EVE, determined to break it (as they had other MMOs). The "Goons" created their own Alliance, Goonfleet, and started attacking other pre-existing Alliances. Their strategy was to win by zerging their better equipped, more talented, more experienced opponents - wearing them down through sheer weight of numbers.

This tactic worked, and Goonfleet was able to muscle its way into 0.0 space and become a major power. Their arrival forced many other Alliances into a series of power blocs, and a Great War ensued between the largest blocs.

In March of 2009, the Apocrypha expansion was added to the game. This added a new kind of territory - Wormhole space. Wormholes existed throughout the original map of EVE and spawned randomly each day. Finding a Wormhole allowed a pilot to go through the hole and into a system on the other side which might (or might not) have other wormholes leading to other systems. This topology was constantly changing and random. Living in Wormhole space was a kind of intermediate step between secure space and 0.0. You didn't get the high end ships or resources but you did get a sense of running a pocket kingdom and you could engage in a whole new logistics chain to produce the best "mid-range" ships.

Also in 2009, a series of decisions by the players of one of the largest Alliances broke the stalemate in the Great War, and the Goons were able to knock the long-time "best" Alliance in the game out of their territory. Things might have settled into a new stalemate, but the Goons themselves fell victim to their own internal problems, and they in turn lost control of that space as well.

Each time one of these tectonic changes happened, it had repercussions across 0.0. Agreements between Alliances are constantly shifting and these large-scale events affect them. In fact, CVA itself eventually felt the force of these changes as it was beaten and destroyed, removing the NRDS territory from 0.0

It is true that if you are a solo player or a small corporation in EVE you cannot access 0.0 space without making an agreement with one of the powers that controls that space. However, it is possible to develop into a sizable and strong Alliance (essentially a corporation of corporations, a guild-of-guilds) in Hi-sec and Lo-sec, and try to take 0.0 space by force. This is the goal of many people in EVE.

EVE is a game that rewards teamwork, long-range planning, strong communities, and trust.

Its is also a game that constantly dangles the lure of solo rewards, instant gratification, community drama and treason in front of people.

RyanD

Specifically, I refer to the comments about the 4500 player limit being a guaranteed game killer, about the game using poor software, about you getting booted from EVE for being incompetent, the one about a PVP MMO not being able to capture what Pathfinder is, and the part about the whole thing taking place in one specific area of Golarion.

Sure, I can respond.

The idea of making a game that starts small and grows over time isn't unique to Pathfinder Online. Sometimes it happens on purpose and sometimes it happens on accident.

However, the MMO field is particularly challenging these days because there are about a million to two million "MMO Hobbyists", who aren't really vested in any one game, but are very vested in playing the latest new releases when they launch on the day they launch. These folks just crush a lot of games and the result is a huge spike in activity, followed by a sharp drop as they "finish" the game and exit.

Take a look at this graph for data that reflects this syndrome:

http://users.telenet.be/mmodata/Charts/Subs-2.png

You can see the standout, can't you? It's EVE Online. A whole different growth pattern than its peers (at launch, or contemporary). If you exclude Second Life, it's also the only sandbox MMO on that chart.

If we went along with our plan to design the game to start small and then grow, and suddenly sprung that on folks at the last minute, there would be pandemonium and lots of feelings of betrayal from those who might feel we'd abused their trust.

By being up front about these plans now, we eliminate that problem. Everyone knows exactly what we're planning - no surprises.

As to it being "guaranteed to kill the game", well, if I believed that was even remotely likely to be the outcome we wouldn't be trying it.

(In fact, I fear the opposite scenario much more where too many people go into the sandbox too fast, and because there's no time for the economic, territorial or political system to develop there's nothing "to do" but kill one another, resulting in a dead-on-arrival scenario. CCP actually observed this directly when it launched EVE in China. The Chinese EVE server in many ways has never recovered from the fact that it launched with a "big bang" and delivered a sour experience for its early adopters.)

Obviously, we won't use poor software. That's just a silly concern. There are plenty of good options available, and Mark, our CTO is a very experienced software engineer who knows good code from bad code. The quality of the middleware options available today are extremely high - in some cases better than the games they were originally designed to support. Rather than "rolling our own", which is vastly more risky, we'll be going to market with software that has already been proven to work in games larger than the 3-year time horizon of Pathfinder Online we've discussed (120,000 or so paying players). Instead of finding out the hard way that the software can't deliver as expected (see: Conan, Darkfall, etc.), we'll remove that risk entirely.

I was very successful at CCP. The marketing team I led increased the subscribers of EVE Online by 50% (unheard of growth for a 5+ year old MMO). (Take a look at that chart again. I started in November of 2007 and left in November of 2010.) I hired the key leadership team of the marketing department, who remain in place and have been carrying the ball since my departure. I ensured the company was ready to plan for and launch the World of Darkness MMO and the DUST 514 FPS. I was fired from CCP because the CEO and I fundamentally disagree on the strategic direction of the company and he did not wish to continue to work with a head of sales & marketing who did not share his vision. They have announced 3 "triple A" titles, and my opinion is that they have the manpower and the resources to do two of them. Their recent 20% downsizing would seem to indicate that my analysis of the situation is more accurate than his, unfortunately.

Pathfinder Online isn't trying to 'capture' what Pathfinder 'is'. The brand is too large to be contained within the game we're building - or I would argue any digital product that could be built given today's technology. Pathfinder is many things to many people. Golarion is many things to many people. We are going to be some things to some people. Each individual player will have to decide on the merits of what we deliver if we've given them the Pathfinder they expect or not.

My personal opinion is that I have a pretty good idea of what the core drivers of the Pathfinder brand are, and they're not to-hit mechanics or the classification of combat actions. Those mechanics predate Pathfinder anyway. Pathfinder to me is a certain way of looking at fantasy roleplaying, a way that is heir to a 30+ year tradition. We'll get that tradition into the game in ways that surprise and delight our players. But you'll also see all sorts of new kinds of ways to play "Pathfinder" that are very hard to do on the tabletop - merchants and crafters and diplomats and soldiers and other kinds of characters who aren't a good fit for the "tabletop adventure game".

We're going to talk a lot about the geographical focus of Pathfinder Online in an upcoming blog, and as I've said before we are going to apply the rule of "underpromising and overdelivering" but here's one thing I can say that's just obviously true:

In a world with teleportation magic, it is not unreasonable to imagine excursions from one area of the world to another, to experience content variable in scope from small to vast. Over the years to come, I'm sure we'll all be surprised with the many kinds of places that Pathfinder Online characters may venture into.

Lol. You get a 120k active playerbase (again using the 20% rule means about 24k online most nights) in a fantasy game that is on one "server" that is majority non-instanced content without horrendous lag in populated areas or daylong queues and I'll eat my hat on youtube.

I'll take that bet.

BTW, just curious, are you aware of what the (single shard) Champions Online game achieved in its PCU?

Or the (single shard) Star Trek Online?

How about Darkfall?

Ultima Online?

Second Life?

Do you understand the difference between "PCU", and "number of active players in the same virtual space"?

Do you understand the difference between server-lag, client-lag, and network-lag? And how those relate to the number of active players that can inhabit the same virtual space?

EVE can support about 1,000 active players in the same virtual space without major performance problems on the server or network (client lag is the biggest problem at this scale, primarily due to video card issues). It can get up to about 2,000 active players in the same virtual space without crashing the server, but performance degrades badly on the server.

EVE has 7,500 star systems. Each system can be hosted on its own node (called a Sol server), although in practice most systems are hosted on shared Sol Servers because they do not generate enough traffic volume to require dedicated hardware.

Within each system, virtual spaces called "grids" by the players are created for each active player, and are shared by active players within a certain range (which is dynamically resized). It is possible to have those 1,000 active players scattered all over the system, or concentrated in one grid - the load on the server is effectively the same, although the network load goes up as a function of N^2, where N = number of active objects being tracked by the server. So 1,000 active players in 1,000 grids would be less taxing on the system than 1,000 active players on one grid.

When a player is in a station, that player is not considered an "active player" in the system's virtual space - so the total population of a system could be much, much higher than 1,000 without causing performance problems.

EVE's system would be infinitely scalable to the point (at least) where each grid could be run on its own hardware except for one problem. The game's core logic is written in Stackless Python, which is one of the reasons EVE was able to grow to the size it was, but now has become a limiting factor. Stackless Python cannot be run across multiple cores. It has a programmatic object (called the GIL) that doesn't enable it to spawn multi-core processes. Long before I left CCP there was work underway to address this issue, but until the problem is solved, there likely are hardware limits to the growth that EVE can maintain within a single virtual space. (The interested can watch this talk: http://blip.tv/carlfk/mindblowing-python-gil-2243379) (And you can see some of how CCP is dealing with this here: http://www.eveonline.com/devblog.asp?a=blog&bid=925)

The idea that a modern "massively" multiplayer online game should have a PCU in the 3,500-5,000 range (I'm looking at you, WoW & WoW clones) is ridiculous. WoW was designed in the early part of the 2000s. It's been nearly 10 years since its network topology and database systems were devised, 10 years which have seen the rise of all sorts of better solutions for transaction processors (look into CouchDB, for one example). But since Theme Park MMOs don't gain much value from having large simultaneous server populations or large virtual space population, most of the theme park segment hasn't bothered to try and come up with a better solution. In fact, they're actively developed to avoid large concentrations of players. Can you imagine what some of the shared spaces in WoW would be like if 1,000 characters were all in the same virtual space? They feel excessively crowded with just a few dozen!

Anyway "tad10", you should probably start picking out that hat now, and we'll be sure to link the video on our community site after your meal.

RyanD