In the last part, I talked about the importance of loot as a motivator and game mechanic in action rpgs. The other half of the equation when it comes to character progression is leveling up. Improving characters through leveling has not changed all that much over the years. Probably because many designers copied Diablo 2's style, but that doesn't mean that it’s the de facto best way.
The challenge with the leveling up mechanic is how much should it affect the gameplay? Most action rpgs on level up, allow the player to improve their character's attributes and unlock/improve a skill. The attributes won't affect the gameplay but have an effect on what equipment is available. Skills are a big deal, as they affect the utility the player has.
One of the issues with designing skills is with the issue of scaling: where players will run through the game multiple times with stronger enemies. If a character has skills that do flat damage such as: "20-30 fire damage," those skills become noticeably weaker on repeat plays. In Diablo 2, each higher difficulty boosts the stats of all enemies which made set damage skills a waste.
To combat this, the most popular way is to implement skills that scale. Many action RPGs have skills that do: "X % of weapon DPS," where DPS stands for damage per second. Scaling allows skills to keep their viability and feeds back into loot as a motivator as now better equipment also equals more powerful skills.
Demon's Souls and Dark Souls also had scaling but it was done differently. In both titles, various weapons had an attribute that it would scale to. For example: magic wands with intelligence, or bows with dexterity. The respective attribute would also be graded on a scale of F to S if I remember right. The better the grade the more of a bonus that attribute would apply to damage. It's important to note that in both titles, there is a dropping off point of around 50 where the scaling will stop being as useful. This was probably done to prevent players from just power leveling through the game.
With that said, we can turn our attention to a few of the leveling formats used in action RPGs. Starting off with the most well known which is Diablo 2. Here, each character class has 3 completely unique linear skill trees. Each tree has the skills in order from top to bottom, or from lowest level to highest. While the final skill unlocks at level 30, players can continue leveling much further than that. Leveling up gives players 5 attribute points to distribute and one skill point. Skills can be improved multiple times with different boosts based on the skill.
The problem with Diablo 2's progression comes at how the skills are unlocked. Besides having a level requirement, each skill requires a point in a previous skill on the specific tree to use. Because of that, it led to a lot of skills that are more or less a stepping stone for a better skill. For example, the Necromancer class has two skills relating to confusing enemies. The first one will cause one enemy to attack other enemies. The second one makes one enemy the target of all nearby enemies. Now in terms of utility, the latter is miles above the former, but you still need to waste a point in the former to get it.
This issue is even worse for the damage causing skills. Why would anyone use the bone teeth skill (level one necro attack spell) once they get access to bone spirit (level 30)? Interesting enough, Blizzard tried to fix this issue with a later patch that added synergy bonuses. Basically, some skills would provide bonuses to more powerful skills giving the player a reason to pump them up. While it helps, this issue is still one of the few problems with Diablo 2.
Torchlight, which was talked about in part one, fared better in terms of progression. Like Diablo 2, each character had 3 skill trees and received attribute and skill points on level up. However unlike Diablo 2, there were no prerequisite skills, instead only the player's level was the factor. This meant that as a player, you would not need to take any skills that you didn't want to in order to progress through the game.
There were still some skills that were better than lower level skills, but there was more utility offered compared to Diablo 2. What also helped was that many skills were built around scaling with fewer exceptions. The only real knock I have with Torchlight's progression is that several skills are shared between the three classes, which do cut into some of the diversity.
Our last example for this post and my personal favorite progression system comes from Din's Curse. The game begins differently in terms of character development compared to other ARPGs. At the start you can choose from either a predefined class or create a hybrid one. The difference is that a pre-made class comes with 3 skill trees, while the hybrid lets you choose any two that you want. So if you ever wanted to be an archer necromancer, this was your chance.
Each skill tree had two different types of skills. The first are proficiencies, which determine what equipment your character can wear, along with any special bonuses. Second are the actual skills you can learn over the course of your game. Like previous ARPGs the skills are arranged in order from top to bottom going from least expensive to most. The big difference is that there are no level requirements for skills, only money and skill points which are earned at level up.
Without any level requirements, it gave the player complete freedom in defining their character. Allowing them to either get several cheap skills starting out, or save up for an expensive skill. By not having to set strict limits on acquiring skills, gave the designers the option of creating more utility skills to make characters personalized. Some players may not even get the most expensive skill on their tree and instead favor improving skills from each skill tree.
Providing meaningful choices in leveling up is an important part of any good ARPG. For the next part I'll be examining downtime in ARPGs and money sinks.
Josh Bycer
P.S As a strange coincidence I got into the Diablo 3 beta on my birthday. The skill system is interesting, but without being able to see everything there, I don't want to include it in this examination.
Don't forget Titan's Quest!
ReplyDeleteAs for D2, IMO if they had removed the "forced pre-req" skill invesmtents it would have been a marked improvement. There are a number of extremely tight skill builds out there that got punished because of it. 4-6 points doesn't seem like much but why force a player to waste them? The 1.10 synergies didn't fix the problem, I agree.
An upside of the D2 system was how organically it scaled into Hell difficulty.
Titan's Quest had a great system too, though a little different. No wasted points there, and it had passive skills that sometimes tied directly into actives in interesting way (e.g. adding piercing, some damage, and a low slow chance to your ice bolt). TQ wasn't without it's flaws. For both TQ and D2, I don't like having to wait until high level to get staple skills (true for some builds, of course). One benefit of TQ of course is that you get some of your staple skills quickly, and then get to watch them grow in interesting ways over time.
For TQ and D2, I loathed the ability score mechanics and how you just dumped points into them every level (and each increase is worse so very, very little). In TQ, I also didn't care for the need to invest in your "path tree". It's an interesting way to gate things I guess, but its a shared currency with skill points and I prefer to keep these sorts of things partitioned. They should have removed ability score increases and instead done something else with investing points into the path or something.
A big problem I have with Demon's/Dark souls is that you can't really know a lot of this stuff about ther system without looking it up. It's not obvious that there is a break point. There's also a discussion to be had about scale there. Allowing something to go to 99 even though it has long since stopped increasing significantly is weird. The upside is that it allows you to "bank" souls (as opposed to have, say, a 1-20 system but past 12 each point gets WAAAAAAY more expensive. In a game like hat it could be problematic).