Travis with some Torchlight II Updates

July 17th, 2012

So, hey there. It's been requested that I bring back the 'Not a Blog' features that I used to do for Mythos, in days of yore.

This won't be exactly like that. I'm not going to spend a lot of time talking about potential features, and so on - because we're way beyond that point in development.

Instead, I'm going to try to give you a general idea of what we're spending time on, why we do it, and why it matters. I may have to rename it tl;dr though.

Many folks are anxious to see the game released - many have preordered, and want a date! We've been deliberately vague on this point. My hope is that if I give you an idea of what we're doing, what it takes to do it, and the scale of it, it will make it apparent why that is.

So, anyway. Immediately after the Beta we spent a good bit of time doing the last big changes that we were comfortable doing. We didn't know how long these would take when presales launched, because, well, we hadn't done the Beta yet. Most of these changes were in response to feedback we received during the Beta and our own observations. The biggest of these had to do with our skill system.

We've talked about the skill changes in previous posts, but to recap, the gist is that there isn't a 'tree' per se. Skills are made available for investment by level (just as they were in the tree), but their unlock level is no longer determined by vertical positioning in the tree, which gives us more flexibility to place them where we want them. In addition, you receive specific bonuses to each skill (apart from standard investment bonuses) at every 5 point mark, or tier, which are hopefully a good enticement to continue investing in a skill.

I'll take the Engineer's Machinegun bot as a quick example (subject to change, of course). At 5 points, his shots begin ricocheting. At 10, they pierce, and at 15 they are tripled. So, a level 15 machinegun bot skill has triple, piercing, ricocheting shots. Way cooler than level 1. Of course, damage, and other properties are improved with every point of investment, but there's a bigger jump every 5. I should stress that we didn't REMOVE cool stuff from skills and ball it all up to make the tiers neater. We have tried to make sure that the skills are designed to be cool as is, and just get a notable 'extra' at the tiers.

The other primary change was to move all passives into their own section so that we could unlock them a lot earlier.

In addition to the skill system changes, we had lots of other alterations we wanted to make after the beta. That is, after all, what the beta is for - to find things you want to fix or change, before it is too late. So we did that.

I don't want to list ALL the changes here, because frankly, I don't remember them all off the top of my head - but some of the bigger ticket items were to do with respecs and how we handle them, how we join games from friend lists, detection of low-ping players, and a bunch of other stuff. We made skills queueable, added Function key binding for skills, added inventory sorting, and so on. Justin spent a lot of time working on additional connectivity improvements and workarounds for funky routers. We also decided we wanted to put more work into our uniques to make them much more special, and so on and so forth.

We had some other items that weren't really finalized yet pre-beta that we wanted to get wrapped up too (item recipes for item combination had some changes to how they function, for instance. We wanted more flexibility in how we could have items cast skills to aid in the creation of niftier uniques).

These changes all took time, of course, and we didn't get back to polishing the rest of the game until we were largely through with them.

There are one or two of these items that are fairly simple that remain to be done (we want to add a checkbox to new game creation in the lobby so you can force a 'reroll', which will reseed your entire game and repopulate it. Handy). We also want to finish setting up our relay server (a relay server bounces messages between players that are having trouble connecting to one another), and we want to handle some specific game-fracture cases where two clients in a larger pool lose connection to one another, but not everyone else (the game has to automatically split them out into their own isolated games).

We finished our primary goals and moved back into polish. There are two main parts to this - skills, and Acts. Skill polish involves making all the tier bonuses interesting, and finishing up the skills that we didn't show in the Beta, and just generally making sure that everything feels good and is reasonably balanced. A pass has already been done to give everything tier bonuses, but some of them are not as nice as we would like, so we are punching them up.

Act polish is the most time consuming. Marsh and I are now focused on polishing Act 2 and Act 3 simultaneously.

So, what does that mean? How 'done' are these acts that we are polishing?

All of the acts exist. (There are 4, but the 4th is quite small in comparison to the previous 3) All of the levels are there, stitched together to form a game. There are events, secret rooms, towns, merchants, the works. They all spawn appropriate monsters, boss events exist, quests, etc. The loot is there. The wardrobes. The NPCs. The cinematics. The 'game' exists.

Polish means taking that and making sure that the individual monsters are fun to fight, that their groupings work well together, that we expose balance issues late in the game (be they from monster leveling, or skill leveling, or what have you) and correct them. We want the individual monsters to each have a 'thing' that makes them unique, and interesting to fight in the context of other monsters around them.

I'm polishing Act 2 right now (the Desert Act - we let people play it at last year's PAX - a lot has happened to it since then). I'm going to guess that I'm about 3/5 of the way complete with my polish pass on that Act. Marsh is deep into his as well on Act 3. It is difficult to predict how long exactly it will take to complete an Act's polish - a monsters' ability might take as little as 20 minutes to make, or as much as a day in rare, difficult cases. Sometimes they just feel 'wrong' and you have to bang on them until they feel right, or get creative about replacing them with something more appropriate.

Here's an idea of what polishing an Act entails. Act 2 has two distinct halves - each one centered around its own large outdoor region.

The first half of the act has about 28 unique monster types in it. (the second half has its OWN unique monsters - although a few from the first half reappear). Each unit has an ability or two (well, they do now), and they are grouped between the 5 spur areas off of the first outdoor region. There are 2 boss battles and 2 'superchampion' battles (these have scripted elements, special skills and areas, and a unique look, but are not quite on the same level as bosses).

Polishing for me meant going through all 28 of those monsters (several of which were newly created where we felt they needed to be), setting up their new abilities to complement one another, tuning health, damage, attack rate, behaviors, spawn density, in many cases altering or adding to their sound sets, new and cooler death particles and explosion particles, weapon alterations, and so on and so forth. Often as we are polishing an act we find a 'gap' - it feels like something new should be there. So we fill it.

Here's an example of how that happens. In the Vault areas (we also showed these at last year's PAX) there are lots of different monsters (Ezrohir) in an area with a decidedly electrical theme. But they didn't really sell that as monsters. They mostly had melee attacks and some ranged stuff, and a few exploding bomb minons. We also had some machine gun and electrical traps in the area, but they were stationary. It felt like we needed more of an electrical vibe. I added a new Ezrohir unit with a shock staff who can shoot bolts of electricity from a distance. We took our machinegun turrets and made mobile versions, which can fire independently of a moving base and follow you. We also added mobile turrets which serve only to spawn 'ezrocopters' - little metallic flying devices that can shock you. Every Ezrohir got a new skill (some of them more than one). Some of them are pretty neat and unlike anything seen in other acts. That area has much more flavor now, the battles are more interesting, and it has a stronger elemental theme which makes it easier to plan for in harder difficulties.

The boss battle already existed, of course. We also let people play it at the previous PAX. But its intro was a little boring (the boss didn't enter in an interesting way), and he didn't have enough skills and they weren't dynamic enough. Now he has 5 new ones, a cool intro, and various elements of the battle have changed to make it more interesting. Polished.

I'm in the second half of the Act now, doing the same sorts of things. Making boss battles more interesting, fleshing out some more monster skills where they seem anemic, strengthening monster groupings, making sure that events have the appropriate kick when they happen. Marsh is in Act 3 doing the same sorts of stuff.

It's progressing well. I'm really happy with how Act 2 is feeling, and I think it will feel very different and fresh compared to the previous Act (which is what we want).

Everyone else in the company is doing the same sorts of things. Mostly, they have the time to add more love because Marsh and I are still finalizing polish on the monsters, bosses, and skills. This lets our Level Design team, for instance, squeeze in some extra rooms, or more neat random events.

Our background team can polish up props that were repurposed and make them unique, or make previously static background elements smashable. They can also add new support props for us to make our boss battles cooler.

Our animators stay busy - because a lot of Monster polish is changing up or adding abilities to them, which goes through our animation team.

Our character artists also stay busy - making better champion variant skins, or helping us fill those 'gaps' by putting together a new weapon, a mesh variant, or what have you.

Erich has been steadily making uniques cooler, and making more special drops for bosses. Those skill-on-item changes I mentioned earlier were done for that reason. You can find uniques that shatter shields, perform skills on strike (or swing), or at other triggering events. This allows for much more interesting uniques that are well and truly unique - instead of just more powerful in terms of a given arrangement of passive properties.

Anyway, I'll try to continue to post on a more-or-less weekly basis until this ships to give you something to chew on, so that you won't have the impression that we're all twiddling our thumbs behind our desks - we really aren't. Nor are we feverishly polishing the doorknobs while the house is yet to be built. We tie things up and we move on, and we do it over and over again until we get to the end.

Thanks for hangin' in there with us. We love ya!

Please feel free to respond to this post on our forums.

Beat the heat this Summer by adventuring in dungeons of ember! Our award-winning debut, Torchlight, is on sale now at Steam for 40% off - just $8.99. Sale begins today and lasts through July 22nd.

« Next Entry « » Previous Entry »