General analysis and comparison of the 3 greatest hack’n slash games ever made (A.K.A ARPGS)

Lately I’ve been playing diablo 2 again and It’s the first time I had the maturity to actually understand it’s underlaying design. And it allowed me to trully see where path of exile and torchlight 2 stand next to it. So I’ll start in chronological order:

Diablo 2

I’ll only comment on the final iteration of the game, the lord of destruction expansion with the final content and balance patches. While diablo 1 was a ground breaking game that started the genre, it was diablo 2 that wrote that book. There is a great variety of classes, each with 3 distinct talent trees, a design choice that would also be incredibly influential, both inside and outside blizzard. That on itself allowed for an incredible amount of build diversity, but the itemization is where the longevity of the game lives.
Itemization was relatively complex. You had base stat bonuses, like strength, vitality. You had secondary bonuses that you would usually get from those stats, such as life and mana. You had advanced bonuses that could only be obtained from gear, like faster cast rate, faster movement speed. You had charges for using skills, bonuses to existing skills and even outright skills to be used, as if you had them on your talent tree. This kind of itemization expanded the possibilities even further. One of the most famous examples are bears builds, when you use a druid skill that morphs the character into a bear on a different class, making use of their own skills in addition to the transformation.In addition to your own build, you had a mercenary NPC that could equip a weapon, a helmet and an armor. So now you also can use this NPC to give it auras or effects on enemies.

While originally the game late game gear hinged on unique items and exceptional rolls on rare items, LOD brought runewords. Runewords are a combination of bonuses that you could put on items and had some strict requirements: items needed the exact amount of sockets for the runes that formed the runewords, items had to be plain and you had to put the runes in the correct order. A few factors compound this mechanic:
1: you couldn’t retrieve runes or any other socketable item once you put them on an item’s socket. You could clear the sockets, but that would destroy the runes in the process. Given the rarity of the most important runes, that gave it a sense of commitment.
2: magic and rare items can’t be converted to normal items, so getting a good base item can be difficult.
3: you can’t roll a random number of sockets on items that are of the superior variant.
4: once an item is given sockets, the amount of sockets can’t be changed.
5: depending on what you want to do with the item, you will also want the base to be or not be ethereal. Ethereal items can’t be repaired, but they are always stronger.

So the RNG could be a bit frustrating. You depended on few rolls, but some of these rolls could have very small odds. Specially if you wanted to min-max, some runewords also had ranges for some of their mod. So you finally dropped the base that had high stats, is the right kind of item, got the right amount of sockets, got the runes, then finally you depend on the runeword rolling good stats.
But this whole process matched the power you would usually get from the runeword, it was very proportional. And mind you, you could very well finish the game without any high-end gear. This is just for people who want to dedicate themselves to the game.
Runes can be rare, too. There are 30 runes in total, divided in 3 tiers of rarity. The tiers are not overt or well defined, but the rarity of each tier is exponetial to the previous one. Getting any of the last 10 runes is an accomplishment on itself. The fact you could use multiple runes of the same type to get one of the next one helped to alleviate the inconsistent RNG.
A last very important mechanic in diablo 2 were charms. Charms were items that gave bonuses while they were in your inventory. They were great to compensate for some small deficiency, like resistances or to push a character to it’s absolute limits, with bonuses to skills. But that also meant you had less room to loot or carry potions. In a game where looting is so, so important, that trade off felt as if you were just trading convenience for power. It was not a very meaningful or thoughtful choice.

Torchlight 2

Runic studios, formed by some members of the now defunct studio that developed diablo had released their own spin on the genre 8 years after the release of diablo 2. Torchlight was not just another diablo clone. It brought fundamental new ideas and a different style. Every item in your inventory used a single slot and you could send your pet npc to sell them while you kept playing.
Another important change was the game aesthetic. While the story was still pretty much the same as diablo 1 where there was a big hole on the ground with a big bad guy at the bottom of the hole, the visual style was a more cartoon and the overall atmosphere was of fantasy, not dark, gritty fantasy. It was NOT diablo. But just like diablo 1, torchlight 1 was very rough too and didn’t quite put the world on fire.

But on 2012, torchlight 2 was released. And it kind of evolved in the same direction diablo 2 evolved from diablo 1: a more open world with a more diverse set of scenery and some twists and turns in the story with more side-quests. It also increased the focus on loot, giving you 3 different inventories for 3 different types of items, gold was automatically picked up.
A fundamental difference of torchlight 2 was the difficulty setting. Unlike diablo, where you played the game each time for each difficulty setting one after the other, you started on the desired difficulty from the very beginning. Once you finished the game, you could start over with the base level of areas increased. But the way it worked, made the game a bit limited for hardcore players: you couldn’t change the difficulty setting or new game+ level. So if you wanted to farm something on a lower difficulty or player with someone on a different difficulty, you couldn’t.

Itemization was more limited than diablo 2. Socketed items were very simple and did nothing but a plain bonus. You had primary, secondary and advanced bonuses, but nothing that completely changed the game, like skills. The stats were extremely imbalanced too. Armor made absolutely no difference due to the way it reduced damage, block was extremely overpowered. A huge issue was how mods were rolled, some mods would never appear above a certain item level, but you can’t go back once you finished a playthrough. Enemies were too much of bullet sponges in the higher difficulties, making the game a bit tedious at times, specially on bosses.
Melee builds were at an extreme disavantage in higher difficulties because of the relation between damage dealt and health pool of enemies. But aside from that, it was a very entertaining and casual friendly game with the difficulty level choices, skill trees with no requirements but character level and every piece of gear could be equiped after the character reached a certain level without meeting the stat requirements.
But being casual also means being shallow. Crafting was extremely limited, your best shot was looting a ton of items and finding incrementally better items. By the way, diablo 3 had come out a few months prior and was shit. *insert error 37 or real money auction house joke here*

Path of Exile

So, while blizzard was too busy nailing hookers and doing coke, a small indie studio in New Zealand named Grinding Gear Games started their own game in the genre: Path of Exile. Starting alpha in 2010, it had big ambitions and knew exactly what it wanted to be. First, the talent tree. Instead of a few trees, it had a singular gigantic web of talents that had to be traversed from a starting point. Any class had access to any node, but started at a different point in the tree. Second, how sockets worked. You had one type of item to put in sockets: gems. Gems always were related to skills and could have one of 3 colors that had to match the color of the socket. They could be an active skill, a passive aura or add an effect to another skill. You could change any gem on any socket at any time. For a gem to affect another it had to be on a socket linked to the other socket.

But most important was: how to obtain items that had the socket configuration you wanted. While diablo 2 had a few rolls that all had to align, path of exile had hundreds of rolls that could be repeated until the desired effect was achieved. If you didn’t get the number of sockets you wanted, you could just try again, same for links or colors. But here’s the catch: the odds of getting greater amounts of sockets or links was exponentially smaller. What this does is to make the RNG more consistent. Having to roll 4 sockets on an item that had 1/6 chances is more frustrating than rolling 4 sockets with a 1/60 where you can keep rolling on the same item and there is a certain abundance of chances to roll. You can even turn items back into white items and then into magic or rare items. Taking an item and getting what you want out of is at the very core of this game.
So not only you didn’t had that sense of going back to square one by losing the base item, but you also had a second important place to keep rolling RNG other than just killing enemies. Coupled with the new socket color and links systems, it added more diversity to the RNG part of the game. Because as far as we know, back in 2001, diablo 2 developers might not even have realized this facet of the genre, everything was too new and the book on game design was still far from finished. I think the several drastic balance changes made to diablo 2 are an evidence of that, after all they were at the bleeding edge of the medium at the time.

Back to more direct mechanics, the support gem system brought an infinite amount of possibilities to builds, together with the talent system. Any class could be anything, it would be only at much later changes that class exclusive mechanics and talents would be implemented. But there was an issue by allowing that kind of complexity: balance. Over the years, balance spiraled completely out of control, with some builds or skills completely dominating the meta. So while there was an amazing offer of possibilities, the disparity was too great to make most of them viable. The game started to be balanced around it’s imbalance and the last time I played, you either killed everything in less than a second or died in less than a second in the late late game.
Due to the nature of the game, it also started to suffer from feature creep. Every new season a major mechanic is added. After some time, many of theses were made obsolete and new players are often overwhelmed by this complexity. But at least these mechanics would often include new ways to craft items that would make it more interactive and add thought and choices to it.

Lastly, I’d like to comment on Path of Exile art style. It was everything players wanted from diablo 3. It was moody, gritty, sharp. It is to this day considered the spiritual successor of diablo 2.

So this is all I have to write about these games. All three are great, each with their own strengths: diablo 2 has the balance, torchlight 2 has the pace and path of exile has the crafting and weaknesses: diablo 2 with the RNG, torchlight 2 with the shallowness and path of exile with it’s balance, but I’d recommend anyone to try all three at some point.