Unpredictable ARPG Development Cycles

[Moja]

Given what we've seen over the past year from developer estimated release times and actual release times for APRGs, I'm wondering: what is it about ARPGs that makes development times so difficult to predict? It's difficult to imagine technical hurdles throwing off predictions as far as they have.

Are there particular design challenges inherent in the genre that makes the prediction difficult? If so, what are they? Skills systems? Item systems? Balance? The sum of all parts?
As an aspiring game developer with a strong interest in this genre, I'm very interested to know which aspects of production make the development timelines so difficult to predict.

Or... does it have nothing to do with the genre, but instead with a design philosophy emphasizing polish and quality?

Very interested to read your responses

[Paws]

It's video games in general. Things I've personally run into on products that would be relevant to your area of interest, and did cause unexpected launch delays:

1. A localization patch completely rendering shields inoperable. Turns out it was an accented letter in a text file.

2. A networking issue corrected in a Korean MMORPG that manifested again in the NA version, despite the fact the games were 100% identical (barring language, naturally). Reasserting the original KR files did NOT resolve the bug.

3. Discovering a bug where stacking over 99 of a common item caused a duplication error where any item transferred between the user's bank and inventory would transform into the last potion used. This was an PVP oriented F2P title, so you can imagine how pissy someone would get when their +12 helmet of doom they dropped 50$ into buffing and enchanting turned into a 1000hp pot worth about a penny.

BONUS ROUND:
4. being told person A is your sole point of contact, and not to email anyone else on the matter. Then to discover two weeks later nobody's answering your emails because that person was fired the next day. I'm so glad I don't work with them any longer. Most irritating clients ever.

[amb2010]

I am no where near as educated as Paws is on this matter but I'll just throw out why I think ARPGs tend to get funky release dates: The amount of crap in em to test.

Most games don't have thousands of possible item combinations or skill combinations with a chance to create random quirks that only pop up under specific conditions. Torchlight's heirloom bug with +health items for example, hierlooming any other type of item doesn't cause any random issues but doing a +health item multiple times suddenly makes it implode on itself and go negative resulting in death on equip. It's also an issue that requires quite a bit of re-working in order to fix and even took players quite a bit of time to find as well.

That isn't to say other genre's don't have their work cut out for them, I just think it's a bit easier for them to focus and nail down issues resulting in faster polishing.

[Zidders]

It also depends on wether or not you have a publisher and wether or not they're the kind of publisher to force you to push a game out early or not.

[wolfmane]

Amb and Paws both give great points for a question that would take a lot of time to answer fully.

It's technical design in general.

Adjusted that. Engineering requires a commitment to technical discipline that most people outside the field don't understand. On top of that people fail to realize that Engineering isn't just about the thing you're working on but can also be about how you're working on it.

It doesn't matter if you're working with software or hardware; there are factors that can be difficult to foresee unless you have very experienced people paying attention to potential problems. Even then it's impossible to account for every situation that could ever arise. Paws has some good ones where 1-3 fall into what people typically expect to see as causes of delay. Number four is the kind that involves a human element that is very commonly overlooked or misunderstood. In fact, in my experience, while difficulties often arise from unforeseen technical issues they're relatively easy to deal with. They usually just take time to solve. It's often the human introduced errors that can really derail a project. Just imagine what happens to your schedule if your core project programmer elopes off to Vegas.

I will say that polish and quality are important values to shoot for in your design philosophy. I think more developers and publishers need to hold themselves to higher standards than they do today. Just remember that there's a flip side to that as well. The Russians were able to put objects into space and give us a run for our money based on a very simple design philosophy,"'Better' is the enemy of 'Good Enough'". That doesn't change the fact that putting an object into space is both inherently hard and dangerous. While it would be silly to directly compare making ARGPs to a manned space program, the loose analogy might hold some truth. ARPGs might be more technically difficult to balance, as Amb suggested.

The real point that isn't commonly understood is that it's not about preventing or stopping possible issues. That's a futile exercise. There can and always will be something that comes up. The point is to be able to deal with it in such a manner that you still meet your deadline (I won't even go into the discussion of the varied semantics of defining a deadline either). Trying to meet that criteria is an incredibly difficult undertaking as achieving good technical discipline is difficult in itself.

Keep in mind that since you're on the Runic forums that there was no real deadline. I'm sure you know this, I just mention it to bring up another issue, the problem of communication. Non-technical people come from a completely different background and process information in an entirely different manner. This holds very true for the many different shades and flavors of fans or people who are only interested in their own biased viewpoint.

So really a different question I'd pose is are ARPG's notoriously late or is it instead an issue with the mentality of a public viewpoint because of the nature of the community and their awareness of a projected release date?

Personally, if it weren't for the fact that they're busy pounding away on TL2, I'd love to hear some thoughts from the Runic veterans on the nature of making ARPGs, in general, compared to other types of games.

[Paws]

Oh man, if you want human examples I can give TONS of those! People is who I deal with 99% of the time as PR/CM, but here's a serious gem from one of my friends told me about just this week:
I won't give their name, gender, or position as people could piece together who I'm referring to, but I'll say this person is a a big muckety-muck of a NA-based company who both develops and publishes games; both the individual and company are widely known. Currently, he (for the sake of convenience we'll use male pronouns) is finalizing a game due out Soon(tm) but despite having a gold press of the game ready, the highest management has not told the marketing team it's done and ready for launch, so there's been ZERO movement on that front -- their teams are so isolated, they have no way of communicating with each other. Yes, they're literally about to launch a game nobody knows about.

[DarkTails]

Ask a mechanic or custom car builder the same thing. You never know what unforseen complications you're going to run into while working in a vehicle, sometimes you have to custom-make a part, sometimes you have to cut things, it's like that in many fields.

[Moja]

Personally, if it weren't for the fact that they're busy pounding away on TL2, I'd love to hear some thoughts from the Runic veterans on the nature of making ARPGs, in general, compared to other types of games.

This is the core thing I'm curious about. Is it something about the genre? Is it design philosophy? Is it the freedom granted by publishers (a point I had not thought about)? I'm sure each plays a part, but if there's something unique about ARPGs as a genre that would be really interesting.

From personal experience I've found "estimated time + 50%" is usually pretty accurate for reasonably well polished software projects, but I also don't have a veteran developer's grasp of scope and estimating development time yet. I'm aware that unforseen problems WILL arise, and delays WILL happen. That's inherent in all game development.

I know neither Runic nor Blizzard ever had a hard deadline, but both companies gave early (rough) estimates with a much larger margin of error than 50%, and both for the same reason: "more polish time". Max mentioned this in an interview, where he said people drastically underestimate the time frames for ARPG development/polish. I'm curious if it's particular systems or the amalgamation of the project that creates this disparity.

[xani]

Ask a mechanic or custom car builder the same thing. You never know what unforseen complications you're going to run into while working in a vehicle, sometimes you have to custom-make a part, sometimes you have to cut things, it's like that in many fields.

Its much more worse than that.
The more often you do something similar the better you are at predicting how much time it will take.
In case of things like "building a car", after 4th or 5th or 10th build you will know how much time will take doing ~80% of stuff (because some thing will just be same each iteration) and only like 20% of it is unpredictable (engine bit different, some new part doesnt exactly fit right etc.) You don't get new kind of engine every 2 years, and you dont redesign electric system every year.

In case of programming, especially games, base things change fairly often (new capabilites of graphic), unless youre making iterative game like another part of series, you have to do a lot of things from scratch (esp. if using new engine) some general knowledge of course can be reused (like skills in GFX design or general programming skills) but most of work is creating new stuff and that's very hard to predict.
If you do something for 5th or 10th time (like application that uses 90% of old code + few customisations tailored to client needs) it can be predicted with reasonable accuracy (tho things like "shit this will/did take x10 of allocated time because of some problems" will still happen), most games are not like that (and thank god for that), and even if "outside" of some games looks similiar it can be completely redesigned under the hood. And then there is stuff like "well this skill system we invented 3 months before it's crap, we need to make it right".

Whining about missig release date and "why it take so long" only makes fails like "Sword of the Stars II" get released way ahead of time (this game after 5 months after release is still buggy as hell and was practically unplayable on release).
Tbh I'd like game companies to say nothing about game right till 2 months before "real" launch ( i mean "testers said game is ready, now only distribution and eventually polishing/adding bonus stuff left")
When I see "hey we will make that awesomely cool game, but release will be at least 6 months from now" I'm just "why you are telling me, now I'll want this game for next 6 months".
I understand reasons (publishers want hype, game companies will get more money if publishers see that gamers are interested in game etc.) but still.

[Zidders]

Thought this might provide a bit of insight. http://gamification.org/wiki/Game_Design

[MrHill]

Oh man, if you want human examples I can give TONS of those! People is who I deal with 99% of the time as PR/CM, but here's a serious gem from one of my friends told me about just this week:
I won't give their name, gender, or position as people could piece together who I'm referring to, but I'll say this person is a a big muckety-muck of a NA-based company who both develops and publishes games; both the individual and company are widely known. Currently, he (for the sake of convenience we'll use male pronouns) is finalizing a game due out Soon(tm) but despite having a gold press of the game ready, the highest management has not told the marketing team it's done and ready for launch, so there's been ZERO movement on that front -- their teams are so isolated, they have no way of communicating with each other. Yes, they're literally about to launch a game nobody knows about.

You are talking about Blizzard Entertainment and Diablo 3, right? Sorry to mention that but first I LOLed then I raged. ... but now I'm calm again...

I just wonder why we can't discuss this in a civilized manner. Without encryption. Anyways, I'm wondering whom you are accusing of being a big muckety-muck? If it's the other guy who got fired. I don't know the exact circumstances of his kick, so I do just assume he got fire for a good reason. However, if you are talking about Mike Morhaime or Jay Wilson, I have to protest big time. But since I'm not 100% sure of your encrypted post, I won't write an entire story of why you are wrong about that, sir. In fact I won't write anything at all until you may want to explain yourself in a clearer way.

But maybe I'm wrong. It seems you know more about game development business than I do, Paws.

[adoomgod]

She isn't talking about D3...

[MrHill]

She isn't talking about D3...

How do you know?

Paws, used "Soon(TM)". That belongs to Blizzard (as far as I know).
She said the game was as good as finished. That was the case with D3, too. Until then they started to cut out things like hell...

OK, maybe it is not D3. Just wondering what's with the encryption. I mean we are here not at Battle.Net where one could get flamed and trolled the hell out.

Here, people can speak freely! If put in the right words. But you really have to forgive my inquisitive nature. So what is it!

[Syl]

I was a game dev and I usually sucked when I had to estimate the time it will take to work on a feature ^^;
With tha being said, I think the main difficulty here is randomization all over the place, balance and IA.
Some games take already 1 to 2 years to make without those features. Balance is always a concern, but it's even more a problem with randomization (and that's what make hack'n slash awesome).
And they have to adjust those features to make the game fun to play.

The devs already did the randomization engine for TL1 (maybe they'll only have to tweak it a bit), they'll still have to work on skills and AI.

Graphists will have their share of assets to produce (I saw the video with armors, I think they did a really good job ). With 4 classes and 2 genders each, they'll have to produce a lot more.
And new acts... they'll have to produce the tiles aswell.

And finally the game designers will have to make sure everything is balanced, take part of AI tweaking (if there's some kind of scripting system ingame). Our game designers always took part in the tasks above to some degree (based on their skills). Level design, AI, skills...

The sound engineer always has plenty of time due to others tasks taking more time ^^;

I was fairly impressed with the quality of TL1 for a game being developped in 1 year.
I guess the amount of features in TL2 will be outstanding ^^

[Ingmartin]

She isn't talking about D3...

How do you know?

Paws, used "Soon(TM)". That belongs to Blizzard (as far as I know).
She said the game was as good as finished. That was the case with D3, too. Until then they started to cut out things like hell...

OK, maybe it is not D3. Just wondering what's with the encryption. I mean we are here not at Battle.Net where one could get flamed and trolled the hell out.

Here, people can speak freely! If put in the right words. But you really have to forgive my inquisitive nature. So what is it!

For all we know, Paws is talking about Torchlight 2, and the closed beta has been going on for the last month or so with an NDA attached. Encrypted language isn't meant to be cracked, and Paws knows how to be Mysteriosa McMysterious!

[gold163]

Paws can't be talking about TL2. Runic doesn't seem like the kind of company to have the marketing team so separate from the goings-on of the rest of the the company, and there's no way TL2 has gone gold yet.

Diablo 3 has a launch date. I'm pretty sure it's gold. That being said, it's none of our business what game they're referring to; it's the situation the example portrays that matters.

[UncleJim]

She isn't talking about D3...

How do you know?

Yes, they're literally about to launch a game nobody knows about.

That last part. Every one knows about D3.

[MrHill]

She isn't talking about D3...

How do you know?

Yes, they're literally about to launch a game nobody knows about.

That last part. Every one knows about D3.

Hmm, such a simple sentence and I overlooked it, what a shame. I have to read more books.

[Paws]

No, it isn't Diablo 3, I used the term "soon" as the way Blizzard uses it, as a pretty understood term in a community of ARPG fans. Sorry for the confusion ^^