Followup to last year's AI War postmortem (now discussing Bionic, TLF, etc).

[Aklyon]

Thanks again for posting these.  It's always interesting to read.  I'm very happy to hear you guys are doing better financially, and will be able to continue to make games in the foreseeable future.

I've got pretty much the same thing to say here. Quite the interesting read!

[x4000]

I am REALLY looking forward to the 4X you have coming out this year.  You seem unmatched in the procedural systems based gameplay content generation department of game design and having that level of expertise applied to my favorite genre.. let's just say I'm hopeful for fantastic results.

Thank you very much!  It's still a long ways off, but I'm pleased by how the (very) early prototypes are panning out thus far with the code and art and design, and I'm excited about the concepts that I have planned for the design.

Thank you for spending the time on responding to my previous questions. (I'd like to take credit for delaying the tLF expansion by a full work day.)

Haha.  No worries.

• Do you prefer to be addressed as x4000 or Chris (or something else) in the forums?

I honestly don't really care; x-4000 has been "my name" for well over half of my life in the digital world, so I'm as used to it as I am to Chris.  People who know me better tend to call me Chris, which I certainly don't mind.

• You had mentioned that that art assets in AVWW2 was much improved, but that people were complaining about the animation quality.  As a disclosure, I am sensitive to motion sickness. Many games with rapid camera movement or poorly done AA can drive me blinding headaches very quickly. I was able to play about 5 minutes of Quake before I had to stop and had a raging headache for hours.  I was playing AVWW2 last night for the first time (previously I had only played AVWW1), and was experiencing constant headache and nausea. I've been trying different graphic settings in the hopes of being able to play it, but if not, I'm afraid I'm not going to be able to finish it. So while it may seem fickle of your audience, there is definitely something that is disconcerting with the animation and movement in AVWW2. Curiously enough, AVWW1 does not give me the problem.

Oh, I know it's not a fickle audience, I didn't mean to imply that, though I realize now that I did.  The problem was a lower number of animation frames on the animations and a general issue with the fluidity there.  The monsters in Valley 2 were almost all animated by me, and I don't think any of them have the problems people complained about.  But the characters themselves were done by other artists and each frame was hand-drawn, and there can be problems with that when you have too few frames for too broad of motions.  It doesn't bother me personally, looking at it, but the difference is going from something like 24 frames to something like 6.

• AIWars and sale units.
• AIWars has the unique position of being popular as a base game and popular with expansions. I'd be curious to see how many base games have been sold compared to each of the expansions, and if they are bought separately or bundled. You can't get that information out of dollar sales, but only from unit sales. It might be informative about development and packaging approaches, or it might just be trivia.

On Steam, the base game by itself has sold $412k, and 51,259 units by itself.  The bundles far outpace this, and then individual sales of the expansions makes it so that barely anybody seems to have JUST the base game.  I would estimate that maybe 20% of the audience, at absolute maximum, have only the base game.

• Bionic Dudes is credited as having a weak name. Do you think it might be "rescued' by the release of a stand-alone expansion pack (which included the base game or could upgrade the base game) with a strong name? Something like: AIWars City Robot Defense League: A Bionic Dudes Expansion? Stardock has done this with Fallen Enchantress/War of Magic.

To be honest, I'm not too interested in trying to go back and rescue games that did not sell for whatever reason.  I mean, we try what we can in terms of getting the word out, marketing and so on, but that ball is in Erik's court.  And myself with interviews, and both of us with setting up promotions and whatnot.  But the game basically is what it is.  We thought about doing a more hardcore-puzzle-game visual style for Tidalis and making basically a Tidalis 2 that would look as hardcore as it is, but ultimately that is a lot of risk for not much potential gain.

To put it another way, there are easier (and vastly more fun) ways for us to make money.  And we have the advantage of not keeping all our eggs in fewer baskets.  Even if the Valley games never break even, which they likely never will, they still do sell regularly, and I won't say no to continued $50-$100k of gross sales per year from them.  Investing further in trying to "fix" them in terms of market appeal would be too high risk, and ultimately frustrating in general.  Versus if we focus on making new games, then we just basically keep on earning from the back catalog by virtue of the fact that they exist, they periodically go on discount, and new players discover them, etc.

• Arcen & Modding
• In regards to opening up Arcen products for modding you responded with:
• "And opening it up to modding would either be just giving away the games for free, or selling them at a super low price."
• I'm not sure what you meant here. Possibly you meant that it would require a reasonably large manpower commitment and that then you'd either have to release it as a free upgrade to the game or sell the modability upgrade for a small fee, neither being very viable. Or I'm completely wrong, can you clarify?

I meant that the games are very hardcoded in most ways.  So either we would have to rewrite HUGE chunks of the engines with no real expectations of increased revenue from that, or we would have to open source the games, which is basically giving them away for free.

• I spend most of my time on a 14" laptop with a native resolution of 1440x 900 (standalone GPU). I'm frequently warned that changing to a different resolution may cause display problems and inferior performance. I'm guessing that this is an example of your cited improper resolution to monitor size issue (in regards to Shattered Haven and Skyward Collapse). What resolution would you say is optimal for a 14" laptop display?

Well, on a 22inch widescreen monitor, 1920x1280 is basically the expected resolution.  In terms of a 14" laptop, if it's widescreen, I guess that would translate to about 1280x720.  Although frankly, smaller than 15" screens are getting pretty small in general.  I can see where that would be a problem.  On my 15" MacBook Pro, I run in 1440x900 and consider that comfortable, but that's just me.

• You've repeated statements about how Arcen has suffered from Testing and QA resource limitations (e.g., Bionic Dudes naming and tLF late development gameplay weakness). Before the downsizing this was Josh's bailwick, but after, responsibilities were distributed across the remaining team. Efforts obviously had to suffer due to the limited manpower you have available and due to the overlap in duties (e.g., you have to design/program/test everything). Larger companies have an entire department devoted to inhouse QA. This department is continually testing products, running viability surveys, and evaluating test panel responses. In an ideal world with unlimited funding, such a department would be great. However, given Arcen's current status and your own plans for the company, that isn't an option. Do believe that QA has been a problem (trying not to put words into your mouth here)? If so, at some stage do you believe you will rehire for a QA position, or do you believe your new development plan (which effectively transfers manhours from design and programming to QA) will address this weakness?

Whoa, I don't think any of that has to do with QA at all.  Realizing that the name does not do well in the market is not a QA problem, that's a producer-level issue and/or a marketing issue.  We've had really excellent QA in general with our process for working with players, etc.  Our problem with the end of TLF was running out of money and having to rush it a bit, which was due to low initial finances coupled with a higher burn rate coupled with having more prototypes than I desired for the combat model (which ate up a lot of dev time).

None of that was Josh's fault, or a QA problem in general.  I would classify those as mostly project management issues, which is in my court.  Aka that I didn't plan well enough, didn't anticipate certain things, and also didn't have the resources behind me to deal with curveballs if they happened.  Having a larger staff exacerbates those problems, because each month costs more and thus you get fewer months.

The biggest problems, QA-wise, are making sure that:
1. The lead designer of the game gets sufficient time to actually playtest and make sure his/her vision is coming through in a coherent way.
2. The volunteer players get enough months with the game in order to find the bulk of the issues.
3. The programmer(s) have enough time to actually respond to the things found in #1 and #2.
4. And finally that #1 through #3 are not running right up to the last week, so that there is instead a "fallow" period right before the game comes out where things are stable and only small polish things and bugfixes are put in place.  Basically where the new AI War expansion is right now, and has been since the start of the month.

In other words, I don't feel like there are any QA problems in a direct sense, I think it's more a matter of project management and time schedules, which are kind of the same thing.  Along with allocation of my own time to make sure that I am heavily testing as much as possible.  Part of the success of AI War, I think, is that I playtested that thing to absolute death.  At this point in time, I probably have 4,000 hours of playtime into that game, far more than anyone else.  That really does matter.  With TLF, I had lots of time for testing things for a while, but then ran into the issue of not having so much time at the end.

There are certain things that only the lead designer notices about their own design, and there are certain things that only anyone BUT the lead designer of the game notice, and you have to have proper testing from both camps to really do a game justice.  Aside from AI War and Tidalis, and to some extent Bionic and TLF and Shattered Haven, all our games have suffered from not enough lead designer testing.  Note that I was not the lead designer on Tidalis, so despite the fact that I tested on it plenty, it wasn't my testing that was crucial in that case.

• Is it arKEN or arSEN?
• I've noticed that you have problems with formating lists too. Any thoughts to why why a seemingly straightforward process is so unwieldy on this forum software?

ArKEN.  And the list formatting gets odd because I copy and paste and can't be bothered to fix it because it looks good enough.  You make it very very difficult to quickly quote you, and I'm trying to respond to you as quickly as possible while being informative, anyway.  I'm a very fast typist, but there's a limit to my output.

Thanks again for posting these.  It's always interesting to read.  I'm very happy to hear you guys are doing better financially, and will be able to continue to make games in the foreseeable future.

I've got pretty much the same thing to say here. Quite the interesting read!

Thank you to you both!

[Histidine]

Great writeup, as always 

One of the things that struck me as a beta tester (i.e. in the outer rings of the development cycle) was the fact that it seemed so surely trapped in development hell there. Sudden feature creep (seemingly started by my comment about weapon switching, heh) with under two months left to release, switching the combat model twice midway through... finally having to release prematurely due to the company's untenable financial situation.

And yet it turned out better than any Arcen game to date. All that hard work (especially at the last minute!) definitely went to make a game that still shined despite the lack of polish, but I have to admit, I was worried for a while there, heh.

[x4000]

Thanks.

And yep, that was neither our worst "development hell" experience, sadly.  But yeah, both Keith and I were burning through hours like crazy.  It was 7 days a week for me for several months there.  That was true at the end of Valley 1 and Valley 2, too.  That's one of the big things that I'm trying to get away from, among other things.

Part of the problem with why TLF got stuck in development hell is that it's essentially two types of game mashed up, and doing that either takes a lot of correct guesses on the first try (in terms of design), or a lot of overtime, or both.  With Spectral Empire, we're focused more on one specific genre (like we did with Bionic or AI War or Tidalis), but in a super intensive way, and with inspirations from other genres (like AI War in particular).  I'm excited about that, and I think it will lead to a less hellish development cycle at the end.  Knock on wood.

We've also very much overbudgeted the time that we think it will take to make it, by design.  We're aiming for next April, and that's probably when we will actually release 1.0 of it, but I expect that we'll be basically feature complete before the end of this year.  We'll see.  Leaving us with time for polish and whatnot in Q1 of next year, while we work on Airship Eternal with the part of our time that doesn't need to go to Spectral.

If things go to plan, then that's a lot more relaxed of a schedule, and one that winds up with much more polish at the end, too.

[Pepisolo]

Good read, thanks!

Oh, I wonder how many more copies Bionic Dues would have sold had it actually been called Bionic Dudes...

https://dl.dropboxusercontent.com/u/30011618/Bionic%20Dudes.mp4 (old testing vid from mantis that might not work via browser)

and been a bit more Cowabunga-y!

[x4000]

Haha, yeah, if it had played up the humor and been called Dudes I think it would have sold better for sure.

[mrhanman]

I think I always called the exos "dudes" anyway.

Any one remember Bad Dudes?

[x4000]

I think we all remember Bad Dudes.   Bionic Dues was actually intended as kind of an evocation of both that and Bionic Commando.

[Misery]

Hm, this was an interesting read.  It's neat to see just where the company has gone with all of these over time, and some insights on the processes behind them.  Very cool, as you guys always have my respect, and I'm glad to have been able to take part in the various betas/alphas to whatever extent.


....also I agree with what you wrote about Skyward's expansion.  I really enjoyed the heck outta that game, but that expansion.... I dont think I'll forget what the testing period for THAT was like anytime soon.   There wasnt anyone around, even here on these forums, where there's usually much interest in everything.  It was very surprising at the time, and probably made things a bit of a pain for everyone involved.   And then it came out and didn't seem to do much.   *I* thought it was pretty good, but nobody listens to me of course....


Very interesting post overall though.   It'd be nice if more developers were this open about things like this, but not many of them are, and most of those that are are all indie devs... ya sure dont see it from the big guys much.  Except when it's time for them to brag about profits, that is.

[x4000]

I think it's honestly more instructive to look at the things that did NOT go well, compared to those that did go well, in the main.  I mean, both are useful, but the former cases are usually where you can spot things that are actually meaningful more easily.  It's why I focused more on those things in this particular post.

[Sounds]

...

....also I agree with what you wrote about Skyward's expansion.  I really enjoyed the heck outta that game, but that expansion.... I dont think I'll forget what the testing period for THAT was like anytime soon.   There wasnt anyone around, even here on these forums, where there's usually much interest in everything.  It was very surprising at the time, and probably made things a bit of a pain for everyone involved.   And then it came out and didn't seem to do much.   *I* thought it was pretty good, but nobody listens to me of course....

I wasn't part of the beta, but I do recall looking at a few of the screen shots and reading through the forums that it kind of unsold it for me.

However a later post of Chris' perked my interest, so I took a gamble. After playing it on the very first sitting for a a few hours  it was one of those games that just clicked for me. If there had been further expansions I would have been first in line to buy them. Sadly I was in the minority camp on this title.

Still whilst I don't play it as much as I would like to (game time is very limited), I class Skyward as my favourite Arcen title. So I'll just tip my hat and give another big thank you to Arcen for taking a chance on an odd idea and releasing it.

[x4000]

Thank you very much!  Skyward definitely is not my favorite out of our roster, but it was incredibly fun to work on for several reasons.  It satisfied several longtime goals of mine in game design.  Making a Carcassone-like game where the pieces come to life was partially satisfied. 

Making a game where the player is their own worst enemy and they have to intentionally basically play a game of brinkmanship with themselves was another.  That's something that AAAaaaaAAa does with "kisses," and it's been something that is my own personal goal whenever I am DMing for players in an RPG or tabletop dungeon crawler (which is my preferred role at the table, as probably surprises no one).  This was a big big thing for me in game design, and I hope to revisit the concept again sometime in the future in a very different way.  But for Skyward, it really was something that I felt came together surprisingly well.  My very favorite press quote for this game, and among my favorite press quotes for all our games is "It's like playing Chess against yourself, and you're always a total asshole."  Haha.  Mission accomplished!

This was also the first game that Arcen ever did that was historically-based rather than pure fantasy or sci-fi.  So the types of research required for this game were completely different compared to things like space games (where we look at data about various planets and stars and whatnot, and various theories about future tech, etc).  Josh Knapp did the lion's share of the research here by FAR for both the base game and the expansion, and was super good at it.  He'd then present me with various options and we'd talk about them together and look at how we could tie in game mechanics to the actual historical mythological stuff.  That sort of research really drove the game design, rather than the other way around, which was very unique for us and a fun way to do things.

Incidentally, the reason that the Japanese have a metric ton of mythological creatures, but no god tokens?  The nature of their mythology.  There just aren't a lot of magical artifacts that Josh could find info about, unlike the incredible multitudes in Norse and Greek mythologies.  He was finding some, sure, but they just were thin on the ground and kind of boring, because of the way the Japanese mythologies were written.  But he had so incredibly many creatures and demigods and similar that it was just vastly more than the other cultures.  So we had the idea to make the Japanese work differently, simply because their mythology itself is different, and that was such a good decision for the game design itself.

The process was really different for making Skyward compared to our other games, in short, and I really love it for that.  It's not among my favorite tiles to play anymore because after you have a certain number of dozen hours into the game I think the concept gets kind of played out.  But that's part of where the price being so low came from, is that we knew that not only was the title quirky, but you wouldn't get 100 hours out of it, it would be more like 20.  In terms of the average player, not the extremist fans of a title, obviously.  Part of me does wonder how much money it would have made if it had been priced higher, because of course it is 1/3 the price of Valley, and 1/4 the price that TLF and AI War both were (and TLF still is).  So in terms of units sold, Skyward is actually vastly more popular than its indicated by its purely financial performance.  I think that a lot of that was people taking a happy chance on something that was small an odd as opposed to making a larger investment in something pricier and larger, though.

Anyhow, rambling, but I love each of our titles for really different reasons.