For any given time to level to 80, a linear curve is actually going to be worse than a linear one. Therefor I think that it would have been better to opt for less levels, and to use a exponential (or similar) experience curve.
To explain a little better what scenario I am talking about, lets discuss the constant and variables.
* Power curve. In this scenario we assume a constant power curve. That means that, for example, a level 80 character will be 10 times more powerful (damage, health, ...) than a lvl 1 character. I assume the game company designs a power curve, and then sticks with it during these tests.
* Time to max level. We assume that the company decides how long it should take until the players has reached maximum level, and then sticks with this during all the scenario's outlined below.
* (edited) We also assume developer time is limited. That means there is only so much time they can work on adding new zones, and that the time spend on developing the game is the same for all scenario's outlined below.
* Time spent in each zone. Is this a constant or a variable? It depends on the design of the game: How large is each zone withing a level bracket? How are the level brackets distributed? This is variable because it depends in part on the leveling curve.
* Leveling curve. This is what this article is about, our main variable.
pic1gw.jpg 29.14KB 271 downloads
For each given level and time spend M, there is a curve that illustrated the time needed for each increase in level.
Curve 2 shows us a linear model similar to GW2. In the beginning of the game you level faster, but then it becomes a line with inclination.
Curve 1 shows us a more exponential curve for the same level and time.
Curve 3 shows us an exponential curve for a lower level but the same time.
* shows us average zones per level and average level brackets.
* Leveling speed and zone size remains relatively constant.
* Would need more or lager zones than curve 1 at higher level because leveling speed is slower.
* Leveling speed in the early game would be fast, and continue to be fast for longer than curve 1 and 3. This would mean that the zones for these levels would be relatively smaller and fewer in between than for curve1.
* Early game leveling speed is fast but similar to curve 1, so there is no need to create more low level zones.
* Leveling speed and level spread is lower, so there will be more zones of the same level compared to curve 1, or these zones would be larger.
An exponential curve with less levels would allow for more zones of the same or similar level. If you believe this is a good thing (more zones for high level players to use), then an exponential leveling curve is better than a linear one.
For any given amount of time to reach maximum level and an exponential leveling curve, the less levels there are the larger the zones would have to be, or there would have to be more zones for any give level bracket. This again gives the players more choice at higher level.
At the same time, there is no real disadvantage to slow leveling, considering that the time to reach maximum level is constant.
tl;dr A more exponential leveling curve allows max level players to enjoy more zones.
Just because the time to reach max level in GW2 is shorten than in WoW does not give anet a good excuse to use a leveling curve that is worse.
Attempt to summarize some of the arguments from players who posted in this thread.
[there is] a large variety of dungeons available to max level that are challenging and engaging.
Also, the speed and pace of leveling in GW2 felt perfect to me, whereas in other games, the later levels felt absolutely grueling.
you're going to lose granularity on the ability to gate people off. Much easier to keep people out of places 10 levels above them than 5 levels above them.
Edited by marvalis, 29 November 2012 - 08:37 AM.