Messages

31

AI War Strategy Discussion / Re: Your "must-buy" techs

« on: September 11, 2011, 06:05:53 PM »

I usually go for Mk2 econ station pretty quickly, set those up and use logistics for most other bases, seems to give me the econ I need... but then again, I usually play on tiny (10-25 star) maps.

Also I LOVE Mk3 engies... I get those as soon as it's feasible, lump the 15 together and hotkey 'em, and keep them constantly zapping around to wherever they're needed... they're OP in an awesome/fun way, as opposed to the game-breaky/exploity way.

32

AI War Strategy Discussion / Your "must-buy" techs

« on: September 11, 2011, 01:50:59 AM »

Knowledge is one thing I think has been done especially well in AI Wars. In every game I play, knowledge is such an effectively limited resource and the benefits so dramatic, that choosing tech paths is a new decision every time. There's no one obvious I-win-button path that you always want to take.

That being said, there are a few technologies that are almost mandatory for standard games. I'm wondering what techs people will pretty much auto-buy by default.

For me, I (virtually) always pick up the warp sensor, and usually most of the other 250-cost stuff. Scout II's for 500 are also automatic almost always. Mk2 of the "big three" ships usually get purchased eventually, but especially bombers, which I think is almost mandatory (although I've experimented with going down the bomber starship route instead). The one other thing I think I never go without is a level or three of FF tech (and with the HFFs, there's potentially over twice as much knowledge suck as before!)

Oh yeah, and grav turrets. Love those puppies.

There's plenty of other stuff I like to pick up when I can, but those are the ones that get bought in 90% of my games.

33

AI War / Re: Martyr Balance Feedback

« on: September 11, 2011, 01:39:24 AM »

Attritioners are on my list of things I want to make more useful (preferably in a more interesting way than only buffing the damage though that could happen too).

Problem seems to be that attritioners won't be able to dish out significant damage to be worthwhile in most cases. I realize this would probably be extra work to code and might even involve impractical runtime overhead, but just to throw it out there: have attritioner's damage rate increase over time against any given ship while it remains in system. Hanging around for more than a couple of minutes in an attritioned system would start to get really painful.

34

AI War / Re: Guide to Finding Coop Players

« on: September 11, 2011, 01:03:31 AM »

Wait, so, to be half a mathematician, all I have to do is stay up for thirty hours?

(I'm an English major.)

...well played, sir.

35

Game Development / Re: What video game would give you the most nostalgic feeling if you played it?

« on: September 09, 2011, 04:31:57 PM »

I love 'em all. Every game listed here (with the exception of Diddy Kong Racing, which I don't know) is awesome, and I still play each of 'em within cycles of about two years. For all-timers, though... Chrono Trigger or MOO/MOO2.

Edit: If anyone's into that kind of thing,

http://soundcloud.com/mindloss/sets

contains me improvising on Chrono theme on piano (track 3) and my electro-dnb remix of Schala's Theme (track 27). (The rest of the tracks are original stuff, but it's all 7-10 years old.)

36

Game Development / Re: Relevant anecdote

« on: September 09, 2011, 04:23:42 PM »

Yeah, now that you mention it, 2D is pretty popular in the indie scene. Great stuff, too; Braid comes to mind. Gemini Rue was also cool. Seems like a better choice for a whole host of reasons.

I've always maintained that the SNES/NES were "superior platforms" in many ways to what we have now (although the N64 holds a lesser place in my heart too). I'm not an idiot, so I know that's probably largely due to childhood nostalgia, but I think there is a very strong case to be made that limiting a medium almost necessarily increases the quality of works in that medium. The prime example of this is NES music with its 2.5 synth channels or whatever; this literally forced composers to design top-notch melodies since there was no "aural room" to include anything else. Sure, there was still shitty music in NES games, and plenty of it, but the tracks by the good composers on the good games are absolutely stellar. (Could you hum the Super Mario Brothers theme, or Tetris, or Zelda? Now how about Half-Life 2 or Crysis?)

Same applies to gameplay. To create a fun game with limited graphics, processing and storage, you had to be creative. Of course, this theory doesn't really hold if you take it to the extreme... the best Atari 2600 games couldn't stand up to the NES (again, probably childhood bias ), but godDAMN it's unfrickinbelievable what they managed with 4K cartridges. All in all, I kind of feel like they hit a sweet spot with the SNES, where the music could be rich but did not have total freedom, and the graphics could be very slick but only to a point... Donkey Kong Country 1-3 looked (pretty much STILL look) incredible, StarFox not so much.

Okay, done with my little rant, and it occurred to me I may be preaching to the choir.

37

Game Development / Re: Chess (or other board game) AI. How to implement difficulty levels?

« on: September 09, 2011, 03:58:26 PM »

I realize this is far too late to be useful to you, but I'll just leave this here anyway because it's an interesting question and I thought about it for a few minutes so now I feel compelled to do a mini core dump.

I would say the answers above are undoubtedly best in terms of dev-time-to-measured-success ratio: limiting search depth or including a randomization factor. However, if you wanted to get fancy and be able to adjust the granularity of the difficulty really well, I would think you could modify your minimax a bit and make it a "lesstomore" algorithm, to coin an ugly term. I realize this is somewhat incompatible with relying on a persistent minimax database, but if you're able to support the calculation for it at runtime, you could tweak a minimax to purposely choose, say, the second best move in every situation, or the moves that over the course of the game most closely approach some percentage of what the optimum eval score (as determined by your actual minimax database) would be for that point, or artificially inflate the value of branches where the human player has a bunch of bad moves and one good move (that one seems especially good for (human) training purposes), etc... there are plenty of similar options to play with there. I'm sure if this is a viable approach, people have already formalized it, perfected it and written about it to a great extent, i.e. it'll be easily googleable.

38

Game Development / Relevant anecdote

« on: September 09, 2011, 03:36:22 PM »

I read an old entry on Chris's blog or FAQ or something about how Arcen Games will only make 2D games, and it reminded me of the last time I went to a store to buy a console game. I walked into a GameStop in... what, maybe 2004? Early-to-middle-age for PS2, whenever that was. I didn't have any particular game in mind, and so I was just staring at the wall of games. A clerk walked over and asked if he could help me find something. I pointed at the wall and asked "Which of these games aren't 3D?" He said "None of them," and I walked out.

Really, though, it's sad to see how with so few exceptions that's all that the console market (PC to an extent, but especially console) is pumping out. Was, anyway. Like I said, I haven't been up on that side of things in a while.

39

AI War Strategy Discussion / Re: AI War "Tip of the Day" -- What do you wish you'd known?

« on: September 09, 2011, 03:27:05 PM »

Figure out about how many turrets you think you'll need. Multiply this number by 5-10.

40

AI War / Re: Okay, who's *actually* used the doomsday device?

« on: September 09, 2011, 03:25:01 PM »

Yep, I almost pointed out the parallel to software design... but figured someone would jump on it for me.   Wasn't expecting a visual aid though!

41

AI War / Re: Okay, who's *actually* used the doomsday device?

« on: September 09, 2011, 03:07:19 PM »

With 6 players, for instance, once you get down to a single-wormhole chokepoint and have all players with a full cap of Riot mkIIs with tazers could create a permanent-paralysis zone to stop the guns of all non-para-immune parts of a wave, cpa, etc.

In my couple of games, the level of coordination was a little more along the lines of "don't let the devourer eat my stuff while i'm in the john." I mean, not quite, but yeah.

In Starcraft 1 I had a MP group that used to do the multiple-players control one team mode.

I'm jealous. I absolutely loved that mode because of the huge increase in potential organization (not to mention controlling all three races), and nobody ever ever played it. (Well, almost never.) Good thing there's AI Wars

mindloss, do you play with lots of other things switched on? Hybrids, Golems Hard and suchlike? Or is it just basic options on 7/7?

I think one of the unusual things about this game also is one's definition of "finding it hard" shouldn't necessarily equate with "no winning". Losing is half the fun!

Well, to be fair to myself, I should say that after I beat my first game (the easy one), I went straight to a 7/7 with a bunch of options on while having no idea what they did. Died a few times, and went back to playing easier levels and slowly learning all the options one or two at a time. I generally avoid loading in most of my games if I lose in a 'reasonable' way (including a couple games that were 10-15 hours in). And now that I think about it, I haven't really played a proper 7/7 since I started (the one I won was that 10-planet snake shield-bugged ship-stalled-bug cluster-f). I'm doing a Fallen Spire on 60-or-something 7/7 now and it's going fine.

So yeah, I'd probably do okay now. I'm finding there are maybe three really important things that kill you often which you have to learn, and then a bunch of small random things that will kill you but are relatively infrequent (e.g. EMP Guardians).

The important things, IMHO:

1) If you poke unsuccessfully at the AI, it will poke back. Hard. (This goes 10x for homeworlds.)

Corrollary: if you poke successfully at the AI, it will still usually poke back.

2) The AI likes to hit you while your attention is on hitting it, so don't leave your guard down and keep an eye on the alerts.

3) Identify your primary defensive chokepoint(s). Figure out about how many turrets you think you'll need. Multiply this number by 5-10.


I've been doing better since keeping those in mind.

42

AI War / Re: RIP Community Wiki

« on: September 09, 2011, 02:43:48 PM »

That was literally the best 'silver lining' I've heard someone come up with in years. Way to turn negative to positive extremely effectively. 

43

AI War / Re: Okay, who's *actually* used the doomsday device?

« on: September 09, 2011, 12:59:15 PM »

Funny you should say that, because in the couple of multiplayer games I tried, it seemed way easier (with 2 or 3 people) than solo. Of course, I've got to admit this all fits with the hypothesis that I'm not terribly good at this game.

44

AI War / Re: RIP Community Wiki

« on: September 09, 2011, 12:23:40 PM »

Yeah, I just wish the search was better on the forums or something... half of the time it's hard to find an answer to a specific question so I just ask it again, and I'm sure it ends up wasting collective cycles. Alas.

45

AI War / Re: Okay, who's *actually* used the doomsday device?

« on: September 09, 2011, 12:20:46 PM »

now I play consitently minimum 8/8 in all games

I look upon your works, ye mighty, and despair.

When you said 7/7 was where the fun started, why play any lower?

I've been meaning to say something about this... I know it's partially a matter of preference, but I started at something like 3/4 and worked my way up from there, and had plenty of fun even in that first game. Also, maybe I'm an outlier, but I think the advice given wherever it is about "start on 5 or 6 if you're new to things, 7/7 if you're familiar with RTS" is more or less insane. I am an average-to-good game player, have racked up thousands of games of Starcraft, MOO2 etc, and found it challenging to beat my 3/4 the first time... and after hundreds of hours on this game, am still having a hard time on 7/7. Maybe I just happen to suck at this particular game, but if many people are reading and taking that advice, that might account for a lot of discouraged players (and 4.7% completion rate).

I know that personally I have maybe 20 games on my steam list that I bought on sale but haven't had time to play yet -- but that I intend to play someday.

Ditto. This was one of them, grabbed during that sale a couple months back on a whim. Finally I needed a change from Civ5 and Frozen Synapse and said "what the hell, let's check out AI wars" (I'd loaded it once or twice before and only made it about ten minutes before getting overwhelmed and giving up), and this time made it through the tutorials. By the time I finished my first game, I was hooked.

I've completed a few games, although I've noticed lately that I tend to get half way through and get distracted. Once I've taken all the nearby planets that are remotely useful and built up some fair defenses and done any gate raiding, I tend to hit a wall. I think I just don't like using transports to hop planets. Splitting up my empire just isn't that much fun. So as soon as I get to the point where I'm looking at the next useful planet being 3 or 4 jumps out, I lose all will to continue. Having to manage multiple fleets just seems too much like work, I suppose. Maybe this is a sign that I should just play on smaller galaxies. Or just learn to live with massive AIP inflation from always taking useless planets in a row whenever I want to get anywhere.

This is also very much my reaction. I've been completing and enjoying the 10 to 25 star maps much more -- in fact, the biggest I've actually finished might only have been a 25. I'll grant you it's a different game and I'm sure you do lose something that way, but I know that feeling of bogged-down-ness, and it's no fun stalling out halfway through a big campaign. [edit: this is of course a recurring issue in the 4X genre, and AI Wars handles it a bit better than most games.] I have the same problem not taking planets... I try, I really do... I always end up neutering it first, usually knowledge raid it after a while, and eventually get sick and tired of dealing with the constantly spawning little pissant AI fleets and just decide it's worth the 15 AIP to own the thing and grab more resources. I just haven't been able to adapt to the concept of jumping over planets yet. But then, I always prefer Pangaea maps in Civ too.