Advertisement

How Torchlight 2 gets character progression so right

This is a weekly column from freelancer Rowan Kaiser, which focuses on "Western" role-playing games: their stories, their histories, their mechanics, their insanity, and their inanity.


How Torchlight II gets character progression oh so right

Here's the moment I realized I loved Torchlight 2: It's the screenshot right above this. This is an Engineer. He's actually my second engineer, and probably the sixth or seventh character I made – although most of the others had only been played for a few minutes. In this case, I created him because my previous main character, around level 40, was running into extreme, frustrating difficulty in the third act. He was also intended to be used primarily in multiplayer, built upon the Engineer's more supportive skill tree (Construction) for which hand cannons are the ideal weapon.

Then I bought him a helmet that looked like a mask, and I realized: I'd just made a Big Daddy from BioShock. And playing my Big Daddy was some of the most fun I'd ever had in an action RPG.

Aesthetically, it's good to look at. It's not just the excellent paper doll effects, it's also the way the Engineer carries the cannon and the recoil when firing. I actually found the original Torchlight's aesthetics off-putting, but tweaks in setting, tone, and graphics did just enough that my distaste turned to enjoyment. While that's a necessary component of what made me enjoy my 'Construction' Engineer, it's not the most important aspect.
%Gallery-166151%

What made me enjoy my 'Construction' Engineer was that I, well, constructed him. I looked at all the different skill trees, and picked the one that seemed to fit how I wanted to play. I was able to test and discard skills to some extent, but I had to choose what I wanted and stick with that choice.

In this case, it was simple: Everything on the Construction tree interacted well with everything else, and all of them made the character I wanted to become. Blast Cannon, a simple, powerful attack, became my primary skill, backed up by a range of different supplemental robots. The passive abilities also built defense. I was slow, ranged, difficult to kill, and had a range of helpers – so I didn't just look like a Big Daddy, I played like one too.


This was a far cry from my first main character, an Outlander. With that character, I danced across the skill tree, picking and choosing several different appetizers. While they all seemed to work decently well on their own, by the time I got to the third act, my character was a mess, trying to work with every stat and far too many useless skills. As much as I had enjoyed her aesthetically, I had, quite simply, built her wrong.

How Torchlight II gets character progression oh so right

In theory, I should have been upset. I'd put enough time and energy into this character that giving up on her should have seemed like a waste. But instead, I was happy. I'd learned Torchlight 2's systems. I could understand how the skills and stats and items interacted. And, with that knowledge, I could take a look at every class and see the potential for different characters in each one. The Engineer is the simplest: Each of its skill trees correspond to distinct weapons. On the other hand, the Embermage – with its different elemental trees – seems like it has dozens of viable combinations.

This isn't just exciting because there are different combinations. It's exciting because I can get those combinations gloriously, messily, foolishly wrong. And with that tension comes the idea that I could do it right. I admire the conceptual elegance of Diablo 3's mix-and-match skills, and I certainly had fun experimenting with different systems, but in the end, I didn't feel committed.

The ability to redefine my character almost instantly by switching her core attributes rendered her, well, something other than mine. Without commitment and risk in character building, it's hard to say there's much reward.


Back when I wrote about the growing preponderance of skills over attributes after I initially played Diablo 3, I noticed a few commenters suggesting that the game's skill system was a step back from the customization of Diablo 2. Although it's technically true that Diablo 2 was more customizable in terms of stats particularly, I find that Torchlight 2 does four things with its character progression that make its system superior to Diablo 2's, which suffered from a dependence on rigid, "cookie cutter" best character builds.

How Torchlight II gets character progression oh so right

Torchlight 2 doesn't use skill trees. You don't have to, say, learn two different fire spells before you can learn the Flame Pillar spell. Instead, you can learn it when you hit level 21. Thus having an ice, fire, and lightning set of spells is an entirely viable build – even encouraged. Second, increasing power in each of these skills is also level-based. You can't simply get to the point where you can use Blast Cannon and pump every following skill point into it. You can only do them at further level intervals. This encourages players to diversify their skills and engage in some long-term planning.

Third, Torchlight 2 allows for experimentation. A spent skill point in Diablo 2 was gone forever, but in Torchlight 2, you can reset the last three skill points you've unlocked. This is a good balance; there's enough permanence there to make the character progression matter, but enough freedom to experiment with a skill. And finally, in terms of statistics, Torchlight 2 is far, far better than Diablo 2 at both making all of the attributes relevant to each character class, and having a variety of different items with different attributes requirements encourages diversification and decision-making in that realm as well.

All these things combine to give Torchlight 2 one of the most gratifying experiences I've ever had building a character in an action RPG, and perhaps one of the best overall. It's a tremendous achievement, made all the more impressive by its relative simplicity.


Rowan Kaiser is a freelance writer currently living the Bay Area, who also writes for The A.V. Club, and has been published at Salon, Gamasutra, Kotaku, and more. He still occasionally finds Ultima VI Moongate maps and mantra notes when he visits his parents' house. Follow him on Twitter @rowankaiser.