Raid Entry and Population

| Saturday, May 14, 2011
Thank you, Google Reader for saving this post. No thanks to Google Blogger from destroying it.

For ease of writing, assume that "skill" refers to some mix of time and skill.

That triangle is a rough representation of a theoretical population. It is a skill distribution, tipped vertically. In this form what it represents is the skill requirement that players will get stuck at. You might notice that it has a slight positive skew, that is because I am bad at drawing. This is supposed to be a bell curve, but it doesn't really matter.

The lines represent the entry barrier to raiding. Or more accurately, they represent a first boss kill, with the idea being, if players get past this, they are now active raiders with raid guilds, but they may fail to get past the second boss, or the last in the last raid at the highest end. This does not indicate how much content there is at each skill level, but it is easiest to visualize, and the most straightforward, if we view it as a chain of bosses of steadily increasing difficulty, which the players will then kill or fail to kill. To be clear, anything above a line is raiding. Anything below the line is not raiding. Incidentally this means that Line 4 shows a really, really hard non-raiding game, but that's not the subject of this post. Let's assume that anything below the line just doesn't exist (devs didn't make the content).

Line 1 means that the lowest raids are really easy. It also means that players will tend to cluster further up along the distribution. This makes it top-light and bottom-light and is a perfect setup for a "left behind" scenario. This is due to most players being able to get well past the entry content. In fact, it is so easy that only a small portion of the population is bad enough to not get past the entry content. New players may find that there are not enough player to play with.

Line 2 is a system in which there are a lot of players clustered near the bottom. If we pretend that the triangle means anything at all, there are 3-4 times as many players in the bottom content, which will reduce the prevalence of left behind syndrome. There are, however, more players excluded from the higher tiers. Take note of this: excluding more players from higher up content makes it easier for those who have the skill to advance.

Line 3 is an exclusive system in which most players are not getting past the entry tier.

Line 4 is a more extreme version of line 3. And it probably represents a game that barely anyone plays. Don't make a line 4 game.

Note that no matter what system you choose you will have a dwindling pyramid at the top. This means that top guilds will always have more recruitment difficulty than the rest since they are in content that most people cannot do. There are two ways to fix this. One is to get a bigger triangle: have a larger population. The other is to crush or chop off the top tier, causing the highest guilds to be doing content below their skill levels. This is inevitable in practice, since I doubt any dev team could create content of exactly the right skill level for that very very top of the pyramid (meaning almost, but not quite, impossible, since impossible isn't very hard to make). We see this in WoW, where top guilds aren't necessarily doing harder content, but the same content as players a few steps down, a lot faster.

This model has some limitations. Such as being an ugly drawing. But also that it models a perfectly linear progression system, which does not exist, since this assumes very exact rewards from previous bosses, such that killing boss A once means you have the gear for boss B, but maybe not the skill, and killing boss A again won't make boss B any easier. Despite that, I think this is a useful visual tool. If it wasn't so damn ugly.


Aracos said...

The more of these discussions on access and optimization I participate in, the more aware I become of the economic realities that drive these little virtual worlds of ours. Using Klep's example, a "Line 4" game is economically impractical. Nobody would play it, so therefore nobody would develop it. A "Line 3" game would be a sort of "niche" game that would attract a very specific following. The question is, would it generate enough of a following to keep the developer in business? At that point, a lot depends on your payment model and just how big your "niche" is.

I would consider WoW-Cata to be a "Line 2" game. Despite anecdotal evidence, if you look at sources that track progress, a lot of people are clearing a lot of content in Cata, just not as quickly as they did before. So overall accessibility is probably higher than most people would acknowledge, at least in the general sense.

WoW-Wrath would be a "Line 1" game to me. The early raids were a joke, outside of Malygos, and once past that content, no one bothered with it again. The only reason people kept running heroics was that Blizzard tied much higher level rewards to them than the content itself should have generated.

I don't think it is any coincidence that Wrath was the most successful expansion in terms of subscriptions and boxes sold. And I don't think it should be any surprise that Cata is less so. It's easy to "sell" to the bottom of the triangle.

Klepsacovic said...

Wrath of the Lich King doesn't quite fit into this model, since it was flattened. In terms of difficulty, I'd put it closer to line 2 (keep in mind they're somewhat arbitrary, meant to show extremes), but with content being constantly removed (effectively removed, due to the badge progression), there wasn't much of a high end of content.

What would have been interesting to see is how LK raiding went without the badge inflation, so that to get Ulduar gear or equivalent badge loot you'd need to do Ulduar. This would have also meant that ICC guilds wouldn't be getting early kills already wearing ICC-level gear, something which would be particularly noticeable in the 10-man track.

Anonymous said...

Is this just a random drawing? Or do you have some math to explain actual numbers at some point in your off center bell curve?

I could do what you have done with no graphic.

1 tried to kill a boss (100%)
2 killed a boss (75%)
3 killed 7 bosses (25%)
4 killed 12 bosses (01%)

That is my now proven theoretical graph for Frito eaters during raids.

I have inconclusively proven that 100% of people that have tried to kill a boss have eaten a Frito.

Klepsacovic said...

You're completely misinterpreting the drawing and post. It is based on the assumption that skill is normally distributed, which is not proven, but is not unreasonable either. Imagine a series of bosses and for any given level of skill, there is a boss which is just barely beatable and just barely unbeatable. If people raid until they hit their unbeatable boss, they will tend to cluster in a normal distribution at their corresponding boss. Now imagine that we substitute actual bosses, but rather than starting with the lowest difficulty, the first boss is somewhere higher. Those are the lines. From that we can consider the effects of low, moderate, and high-difficulty initial bosses.

Post a Comment

Comments in posts older than 21 days will be moderated to prevent spam. Comments in posts younger than 21 days will be checked for ID.

Powered by Blogger.