Author Topic: Want to read about the game's creator losing pitifully to his own AI?  (Read 3112 times)

Offline x4000

  • Chris McElligott Park, Arcen Founder and Lead Dev
  • Arcen Staff
  • Zenith Council Member Mark III
  • *****
  • Posts: 31,651
Yeah, it's wicked long of a write-up, so you have to be really into that sort of thing.  But, if you're not familiar with the AAR part of the forum, let this be an introduction!  There are some great other writeups already in there, with pictures and videos and everything in many cases, but there are a lot of styles of AAR.  If you're into that sort of thing, feel free to write up one (or more!) of your own.  It can be quite fun to read about other people's miseries or successes.  In the case of mine, I hope you like misery. ;)

http://arcengames.com/forums/index.php/topic,3221.msg20411.html#msg20411
Have ideas or bug reports for one of our games?  Mantis for Suggestions and Bug Reports. Thanks for helping to make our games better!

Offline RCIX

  • Core Member Mark II
  • *****
  • Posts: 2,808
  • Avatar credit goes to Spookypatrol on League forum
Re: Want to read about the game's creator losing pitifully to his own AI?
« Reply #1 on: January 16, 2010, 02:52:11 am »
Glad to hear your own ai can still pummel you ;)
i look forward to the read!
Avid League player and apparently back from the dead!

If we weren't going for your money, you wouldn't have gotten as much value for it!

Oh, wait... *causation loop detonates*

Offline x4000

  • Chris McElligott Park, Arcen Founder and Lead Dev
  • Arcen Staff
  • Zenith Council Member Mark III
  • *****
  • Posts: 31,651
Re: Want to read about the game's creator losing pitifully to his own AI?
« Reply #2 on: January 16, 2010, 02:57:51 am »
Oh yeah, it surprises even me routinely.  That's the beauty of an emergent AI, I guess.  Really, it wouldn't work any other way, since I was firstly creating it for myself to play with my dad and uncle and so forth.  If I was able to predict what the AI would do, and would never lose, it would be pointless as a game for me, even if others could get a few months out of it before figuring it all out.  So I'm glad I succeeded in the design goal of making the game not only interesting to others, but also interesting and challenging to myself.  That's harder than one might think.
Have ideas or bug reports for one of our games?  Mantis for Suggestions and Bug Reports. Thanks for helping to make our games better!

Offline Lancefighter

  • Core Member
  • *****
  • Posts: 2,440
Re: Want to read about the game's creator losing pitifully to his own AI?
« Reply #3 on: January 16, 2010, 03:00:37 am »
Good read, although atypical to post something like this in the general section

I indeed would agree that it seemed you tried to expand quickly - in my latest games against a special forces AI, I dont tend to go anywhere until 30-45 minutes in (I'm determined to beat a special forces one day <.< ) Then again, the timescale would probably step up a tad with more players, given the increased resource gain


Silly question then; Have you played against an alarmist often? It seems in your haste you didn't think properly of what alarmists would actually mean, especially with them in almost every other system
(note here, I dont even own the expansion, so I've never seen them :x )
Ideas? Suggestions? Concerns? Bugs to be squashed? Report them on the Mantis Bugtracker!

Author of the Dyson Project and the Spire Gambit

Offline x4000

  • Chris McElligott Park, Arcen Founder and Lead Dev
  • Arcen Staff
  • Zenith Council Member Mark III
  • *****
  • Posts: 31,651
Re: Want to read about the game's creator losing pitifully to his own AI?
« Reply #4 on: January 16, 2010, 03:15:15 am »
Good read, although atypical to post something like this in the general section

Yes, I'm trying to use it to drum up some interest in the AAR section, as it is the most-ignored part of the forums presently. :)  I think many people don't see it or know what it is.

I indeed would agree that it seemed you tried to expand quickly - in my latest games against a special forces AI, I dont tend to go anywhere until 30-45 minutes in (I'm determined to beat a special forces one day <.< ) Then again, the timescale would probably step up a tad with more players, given the increased resource gain

Yeah, more players does change the dynamics some -- you tend to have more ships, and more to lose as time passes.  And you need way more planets very quickly to sustain a comfortable economy.  But still, we rushed into it way too much.  We've beaten special forces captains a couple of times, but that was on various older versions of the game.  I think they have grown more formidable since last we faced them.

Silly question then; Have you played against an alarmist often? It seems in your haste you didn't think properly of what alarmists would actually mean, especially with them in almost every other system
(note here, I dont even own the expansion, so I've never seen them :x )

I have not played against alarmists much, and we as a group never have.  We've encountered alarm posts several times, but that's very different from having alarms everywhere.  From my own experience in solo with the alarm posts (and my initial design of them as a unit) I of course knew what to expect -- what I didn't expect was how they combined with the speed boosters to create super-speedy-giant-waves of ships.  Had it not been for the speed boosters, we would have been perfectly fine with the strategy we had.  Alternatively, had it not been for the alarm posts, we would have been perfectly fine with just the speed boosters.  Put them together, and I didn't anticipate the result.

EDIT: And then, after seeing that result, the second time I still did not expect the severity of that result.  When I lost my home command station, it was not clear just how badly I had been overwhelmed, since the cleanup went quickly.  So my forces in the western planet seemed more likely to hold.  But by the time I saw what was coming from those alarm posts, it was far too late to do anything about it, anyway.  That's a case of where if I had been more observant earlier in the game, I could have made a better logical prediction and would have known to not use a general brute force tactic against the alarm.  But so it goes when you're somewhat tired and still recovering from a protracted period of over-intense development. :)
« Last Edit: January 16, 2010, 03:18:22 am by x4000 »
Have ideas or bug reports for one of our games?  Mantis for Suggestions and Bug Reports. Thanks for helping to make our games better!

Offline quickstix

  • Sr. Member
  • ****
  • Posts: 297
  • Buy Now
Re: Want to read about the game's creator losing pitifully to his own AI?
« Reply #5 on: January 16, 2010, 03:30:31 am »
Fantastic read. The AI always manages to find a way into my nicely built network every now and then, and it's nothing but chaos when that happens and a huge relief if I come out of it alive. It's also great there isn't just 'the only way to win all the pros use', but there are so many ways to play the game, all dependent on one's playstyle.

I'd love to write some AARs, but I always get too involved in my games and forget half of the stuff I do. Any tips for the keen writer here?

Offline Lancefighter

  • Core Member
  • *****
  • Posts: 2,440
Re: Want to read about the game's creator losing pitifully to his own AI?
« Reply #6 on: January 16, 2010, 03:36:21 am »
I suppose that does bring up a point - Was it intentional that the units used by the AIs be mostly intertwined? For instance, my current game (well, short range, less than an hour so far) has a spec forces and a stealth master; Yet it seems that both of them use cloaked units extensively in their waves (schizophrenic waves, by the way)
Ideas? Suggestions? Concerns? Bugs to be squashed? Report them on the Mantis Bugtracker!

Author of the Dyson Project and the Spire Gambit

Offline RCIX

  • Core Member Mark II
  • *****
  • Posts: 2,808
  • Avatar credit goes to Spookypatrol on League forum
Re: Want to read about the game's creator losing pitifully to his own AI?
« Reply #7 on: January 16, 2010, 03:38:10 am »
Fantastic read. The AI always manages to find a way into my nicely built network every now and then, and it's nothing but chaos when that happens and a huge relief if I come out of it alive. It's also great there isn't just 'the only way to win all the pros use', but there are so many ways to play the game, all dependent on one's playstyle.

I'd love to write some AARs, but I always get too involved in my games and forget half of the stuff I do. Any tips for the keen writer here?
Me too :) i tried one time but then i quickly abandoned the game i was narrating... I promise i'll stick to a game soon honest! :D

Though can that be moved out of the Meeting ground subforum? it seems like more of a general AI war thing...
Avid League player and apparently back from the dead!

If we weren't going for your money, you wouldn't have gotten as much value for it!

Oh, wait... *causation loop detonates*

Offline HellishFiend

  • Hero Member Mark II
  • *****
  • Posts: 758
Re: Want to read about the game's creator losing pitifully to his own AI?
« Reply #8 on: January 16, 2010, 03:55:54 am »
Me too :) i tried one time but then i quickly abandoned the game i was narrating... I promise i'll stick to a game soon honest! :D

Yeah, you really should. 900 posts and not even one completed game! I've had three completed games (two wins, one loss, in 2 player co-op), and two abandoned games (both solo). I think so much more satisfaction comes from the game by finishing it, win or lose, as opposed to just playing it. If youre contemplating throwing a future game, you should just make a rash attempt at pushing for victory since you have nothing to lose anyway.
Time to roll out another ball of death.

Offline Harry

  • Jr. Member Mark III
  • **
  • Posts: 88
Re: Want to read about the game's creator losing pitifully to his own AI?
« Reply #9 on: January 16, 2010, 09:56:57 am »
X4000, thanks for posting that!  Gotta love it.

Count me as another person who started but didn't finish an AAR in the AAR forum.  I even called it a "MAR" -- a middle-of-action report.

Then I re-started and now I have an exciting ongoing game.  If I finish this one, I will post it as a complete AAR.

Offline Harry

  • Jr. Member Mark III
  • **
  • Posts: 88
Re: Want to read about the game's creator losing pitifully to his own AI?
« Reply #10 on: January 16, 2010, 10:08:23 am »
By the way, when I first started posting here, I didn't notice the AAR forum.  It's in a kinda small font at the top of the board.  If I were czar, I'd consider merging it with "ongoing games" and making it the same "status" as the other 3 or 4 main forums -- strategy, development discussion, etc.  In the wargame boards I frequent (e.g., Matrix Games), the "AAR" section is home both to completed games and ongoing ones.

Also, an active AAR forum can be a good marketing and teaching tool.  Teaching because newbs learn from others.  Marketing because a well-written, exciting description can help sell the game.  "General" forums like this one inevitably have bug reports and other stuff that can turn off newbs, whereas AAR forums tend to show off the game at its best.

In any case, I really enjoyed reading about your demise.  It made me feel better.  :)

Offline RCIX

  • Core Member Mark II
  • *****
  • Posts: 2,808
  • Avatar credit goes to Spookypatrol on League forum
Re: Want to read about the game's creator losing pitifully to his own AI?
« Reply #11 on: January 16, 2010, 05:10:52 pm »
Me too :) i tried one time but then i quickly abandoned the game i was narrating... I promise i'll stick to a game soon honest! :D

Yeah, you really should. 900 posts and not even one completed game! I've had three completed games (two wins, one loss, in 2 player co-op), and two abandoned games (both solo). I think so much more satisfaction comes from the game by finishing it, win or lose, as opposed to just playing it. If youre contemplating throwing a future game, you should just make a rash attempt at pushing for victory since you have nothing to lose anyway.

One of the big problems with that is -- well like in my current game, i've hit a roadblock of sorts and if fail at it too long then i just feel like chucking the game. I think i'll go ahead and just nuke the planet!
Avid League player and apparently back from the dead!

If we weren't going for your money, you wouldn't have gotten as much value for it!

Oh, wait... *causation loop detonates*

Offline x4000

  • Chris McElligott Park, Arcen Founder and Lead Dev
  • Arcen Staff
  • Zenith Council Member Mark III
  • *****
  • Posts: 31,651
Re: Want to read about the game's creator losing pitifully to his own AI?
« Reply #12 on: January 16, 2010, 08:35:53 pm »
I really like the idea of making the AARs section more prominent -- so I have done so.  Hopefully it will see some more love over time. :)

As for the accuracy of reporting on a game after a lengthy section... well, I think in all cases, details get minorly flubbed and switched around unless the player was taking notes.  There may be some minor inaccuracies in my report as to when I attacked the western planet exactly, versus when my planet got destroyed.  Because it is hard to remember, and I hadn't been playing with an eye toward remembering it for later.  But, the spirit of what was going on, and the general events and analysis, are easier to remember -- if you can't remember the details perfectly, just fill in the blanks as best you can. 

Throughout my AAR, I used a lot of vague language for timing and similar, which I think makes it fairly apparent that the game wasn't recounted from a photographic memory of it; I think people don't mind that at all, and it's sort of expected with AARs of any length and detail.  Plus, there are a variety of different levels of depth you can go into.  This game was short and full of notable events, so I went into great depth. In other AARs I have done, I just briefly covered the high points of 10 or more hours of play.  Sometimes it is more interesting just to describe the forest and the grand strategy involved.  Sometimes the trees are the focus.  I think it's great either way. :)
Have ideas or bug reports for one of our games?  Mantis for Suggestions and Bug Reports. Thanks for helping to make our games better!

Offline Echo35

  • Master Member Mark II
  • *****
  • Posts: 1,703
  • More turrets! MORE TURRETS!
Re: Want to read about the game's creator losing pitifully to his own AI?
« Reply #13 on: January 17, 2010, 01:23:31 am »
Oh yeah, it surprises even me routinely.  That's the beauty of an emergent AI, I guess.  Really, it wouldn't work any other way, since I was firstly creating it for myself to play with my dad and uncle and so forth.  If I was able to predict what the AI would do, and would never lose, it would be pointless as a game for me, even if others could get a few months out of it before figuring it all out.  So I'm glad I succeeded in the design goal of making the game not only interesting to others, but also interesting and challenging to myself.  That's harder than one might think.

That just means you are a very good AI programmer. Please never work for a robotics firm. ::)

Offline Harry

  • Jr. Member Mark III
  • **
  • Posts: 88
Re: Want to read about the game's creator losing pitifully to his own AI?
« Reply #14 on: January 17, 2010, 01:31:41 am »
Nice job relocating the AAR forum!  I hope it gets more traffic now, too.