Funny story from testing tonight.

[x4000]

So I've been coding in some of the "bugs" that the various robots have.  Keith thought the ones up so far, and we're going to add more as we finish implementing more of the bots.  The bugs are things that are wrong with the robots that they do stupidly, and that later model revisions in the game fix (and often new different bugs crop up).  It makes for both a lot of hilarity as well as interesting strategic options.

So here's what just happened to me:

Me: Clicks on node on world map, starts mission.  Trundles on down hall and opens the door in the south.  Two DumBots are behind it, and a freaky activation noise plays as their lights turn on and they turn to face me.  (My wife remarks at this point "that was a scary noise.")

Me: Shoot first DumBot, who happens to be right next to me.

DumBot #2: Shoots DumBot #1 instead of me.  Their "bug" is that they always miss every other shot, hitting a tile adjacent to you (and whatever is in that tile) instead of you.  The next shot will hit me, though.  And it's lucky there was an enemy next to me, instead of something like a fuel tank.  And it's lucky that out of the four cardinal directions, they happened to pick the one where another enemy was standing.  Yay luck!

Me: Shoots DumBot #2.

DumBot #2: Shoots me for very little damage.

Me: Shoots DumBot #2 for the kill, and then trundles on down the hall some more.  It's midlength and suspiciously empty, so I make great time and am soon at another door. I open the door and the freaky noise plays again.  It's a room full of DumBots and BlunderBots.  Normally I would retreat down the hall I just came from, whistling and kiting and shooting them.  But I am here to test the BlunderBot's bug, so I step into the room to see what happens.

Opposing Forces: Oops!  Oops!  Miss!  Miss!  Miss!  Oops! (Explosions and death)

Me: Assault Exo down!

All the DumBots miss me, of course, and the ground on all four sides of me is riddled with holes (not really visually, just in my mind on that one; everything else is literal).  But BlunderBots don't quite have that bug.  They have a big honking rocket launcher, and they always fire a random direction on their first shot, and then they fire with accuracy with their second and final shot.

The problem?  Random direction means a random tile in their range, and that could be near you or far from you.  Two of the BlunderBots shot into the crowd of DumBots, but not before the DumBots could swiss-cheese the floor near me.  The third BlunderBot gives me an almost direct hit by chance, causing massive AOE damage to my hull that almost takes me out.  And the fourth BlunderBot barely wings me.

On to my Siege Exo to clean this mess up with his rocket launcher!

But in actuality that was where I stopped, because I had tested the BlunderBot and it was working just fine.

Keith and I keep having trouble with just wanting to play the game, which I think is a good sign.

[mrhanman]

The idea of bugs existing in the software of the enemy bots sounded pretty neat at first, but now that I have a better grasp of how it will work in game, it seems positively inspired!  One of my favorite parts of roguelikes has always been cautiously approaching new enemies to see what they "do".  I must admit I've never had an experience like this!

Sorry if you've mentioned this before, but how many unique enemies are you planning to include, not counting newer "patched" enemy revisions?  And, for that matter, how many revisions per enemy?

[x4000]

The idea of bugs existing in the software of the enemy bots sounded pretty neat at first, but now that I have a better grasp of how it will work in game, it seems positively inspired!  One of my favorite parts of roguelikes has always been cautiously approaching new enemies to see what they "do".  I must admit I've never had an experience like this!

It was Keith's idea on the bugs, so I can't take credit for that.  I thought that was really awesome also!

Sorry if you've mentioned this before, but how many unique enemies are you planning to include, not counting newer "patched" enemy revisions?  And, for that matter, how many revisions per enemy?

Bear in mind that you are almost always fighting crowds of enemies, not individuals.  So the mix of what you are fighting matters a lot also.  For a sense of perspective, I checked the Dredmor wiki and saw that they reported 58 enemies in that game, loads of which are remixes of one another.

In Bionic, at the moment, we have I believe 42 enemies in 12 "lineages."  Out of those 8 are "general lineages" that have three robots in a succession, but with unique properties on each.  For a period during the transition, the old and new robots from a lineage are both mixed into your levels, and then the newer robot takes over completely.

There is then a Sentry lineage that also has 3 bots, and that's something that works a bit different from the others for a variety of reasons.

There is then a Boss lineage, which has three different boss types at the moment.  These are only in assassination missions and the final battle for your HQ on the last day.

There is then the "Advanced Unique" lineage which has 9 various bots in there, all of which do oddball "special teams" sorts of things.  These are scattered around through most of the game, though not in high concentrations.


The individual robot types have levels that increase at random as the game goes on, incidentally, so they remain relevant (most of them, anyway).  By "random" what I mean is that every day there are a certain number of upgrade levels the AI gets (varies by difficulty, it's 3 on Normal and 5 on Hard).  These upgrade levels are then applied to all the bot types in random lineages, one after another.  So in one universe you might have DumBots that are the terror of the earth, and in another it's really the snipers line that is the problem.

We'll see how this volume of bots winds up feeling, but we're strongly considering adding some more before 1.0.  I think it already fares favorably compared to a lot of roguelikes, though.  The stat creep is there in the form of the levels, so the actual upgraded models are free to be distinct entities (and they all have unique art, none of them are recolors... well, except the MurderBot, but his creation was a happy accident).

[mrhanman]

MurderBot?  I like it already!

[Penumbra]

Are the bugs tied to the individual robot types, or are they randomized like potion colors and effects in a "standard" rogue-like?

[x4000]

Tied to robot types. This started out as a tactical hybrid, and it still maintains some of those trappings. That way when you look at a complex room of robots, you can actually attempt a strategy versus just getting creamed.

[nas1m]

So I've been coding in some of the "bugs" that the various robots have.  Keith thought the ones up so far, and we're going to add more as we finish implementing more of the bots.  The bugs are things that are wrong with the robots that they do stupidly, and that later model revisions in the game fix (and often new different bugs crop up).  It makes for both a lot of hilarity as well as interesting strategic options.

Pure genius. You guys (and gals) never cease to amaze me .

[Penumbra]

Ok, so I just had a dream where all the versions of my software from work that I have released attacked me. I had to use my knowledge of their existing bugs to exploit their behavior and defeat them. It was both the most exciting and nerdiest dream I have had. There are also only three people at work I could describe the details with.

I believe I have you guys to thank for this?

[x4000]

Hahahaha. Our work here is done.

[Cyborg]

Question on the bugs, is there any ability for the player to learn ahead of time or intellectually master the encounter, or are you always going to be caught the first time with a new enemy?

[x4000]

Question on the bugs, is there any ability for the player to learn ahead of time or intellectually master the encounter, or are you always going to be caught the first time with a new enemy?

The tooltips give some flavor-text-style hints as to what those are, but otherwise the player is caught blind and has to learn.  This is really kind of a mix of a roguelike and a tactics game, but it leans way more to roguelike.  In most roguelikes you wouldn't see any information about enemy behaviors (at least not in the ones I've played), versus here it gives you some data but not as much as something like AI War.

[mrhanman]

Assuming you make it to the end, approximately how long does a single game take?

[x4000]

Assuming you make it to the end, approximately how long does a single game take?

That is an excellent question that I don't really have a good answer for.  Given the alpha state of the game, we've never played a game of it all the way through (which shouldn't be surprising at all, but I guess you might find it to be at first blush).

I can say that most missions I've undertaken thus far have taken me about 5 minutes, but we've only implemented the more generic styles of map population thus far.  So some will take longer and some shorter.  Based on that, it would take about 4 hours to complete the missions portion of the game, which is probably on the low end.  If it was a range of 5 to 15 minutes with the average falling somewhere like 7 or 8 minutes, that would be more like 5.8 to 6.6 hours for the missions part, presuming you did not accelerate the game by doing any command centers of the AI (which most advanced players would).  If you did all five of those, then you'd be doing more like 30 missions, so 3.5 to 4 hours basically.

But that's just mission time.  There is also probably at least 1-3 minutes on average of looking at the city map and making decisions.  That's hard to gauge for us, as we spend less time contemplating things like this since we already know what all of it is.  So this would be an additional 0.5 to 2.5 hours, depending on how pessimistic you want to be there.

Then there's Exo customization.  I think this will vary enormously by how OCD people are.   I think that this could last anywhere from 2 to 20 minutes per mission, depending on the level of OCD-ness of the player.  So that's 16.6 hours on the high end, to 1 hour on the low end.

Adding this up, I see that the range of hours here is:
Missions: 3.5 to 6.6 hours predicted
City Map: 0.5 to 2.5 hours predicted
Exo Customization: 1 to 16.6 hours predicted

And since as developers Keith and I have an advantage, let's double those mission and city map estimates.  In total that would then give us a total time range of 8.5 to 34.8 hours to complete one campaign for a player who is not intimately familiar with the game, whereas a highly expert player could be doing it in more like 5 hours minimum, most likely.

Honestly the answer to AI War is kind of similar.  I tend to say 7-12 hours there, with closer to 12 being the average.  But lots of players routinely run 16-20 hour gamges, and a goodly number are up into the 20s and 30s for a game, and a few go north of 100 hours in a single freaking game (no joke).  And then there's the few that manage it in 3-5 hours somehow.

I actually think that the variance here will be less extreme than AI War, but it's still a lot more extreme than something that is generally quantified by hours (like a JRPG).

Clear as mud, right?

[mrhanman]

Clear as mud, right?

This actually a really good explanation that not only answers the question, but gives an overview of the game's various phases (which was my next question).  Thanks for that!

[x4000]

Great!  My pleasure then.