Posts: 4,671
Threads: 36
Joined: Feb 2013
(May 5th, 2024, 21:19)RefSteel Wrote: (May 5th, 2024, 20:16)Gavagai Wrote: (Also, I think people are putting way too much faith in AI-mods when they think they don't have any goofy behavior but I am in minority of one in that respect...)
But they're goofy and dumb in (generally) entirely different ways than the base game's. The attraction of watching Civ4 AIs running around doing their thing is partly just watching the AI "personalities" we've known and loved/hated for two decades running over each other (and/or failing to) in their particular and peculiar ways, often managing to surprise us along the way. And it's fun to laugh at dumb AI mistakes when the AI was developed as part of a multi-million-dollar game-making effort by a major corporation that was intentionally designing AIs that would be fun and interesting to play against, including for the large majority of gamers who don't go too deep into the game's mechanics and strategy. It's a lot less fun (at least for me) to say, "Ha-ha, this hobbyist's AI that's meant to play more competently and make the game more fun for expert players like me - and actually does so in many respects - still messes up a lot in a bunch of dumb ways!"
(There's also Sullla's well-documented avoidance of mods in general, for which he has good reasons, including some that relate to the above.)
Or to rephrase in fewer words: Many of us want to cheer for Mansa or Cathy or our other favorite AIs, and see them move up or down the rankings each season, and that doesn't mean as much if their behavior and the background it's playing out against suddenly or periodically changes.
I believe the idea that K-Mod AI obliterates / distorts AI personalities is a misconception promoted by those who either did not play against it at all or played a couple of games and quit after discovering AI now actually tries to fight back against invasions. In my extensive experience AI-personalities in K-Mod are more pronounced if anything because they are no longer that much buried behind randomness and incompetence.
Posts: 10,079
Threads: 82
Joined: May 2012
I would prefer a different AI, but mostly just because I'd love to see something other than pangaea. It gets tiresome, but any islands screw with the vanilla AI too much, so you need the former to enable the latter.
Still, I think Sullla's preferences are clear.
Erebus in the Balance - a FFH Modmod based around balancing and polishing FFH for streamlined competitive play.
May 6th, 2024, 16:40
(This post was last modified: May 6th, 2024, 16:49 by Eauxps I. Fourgott.)
Posts: 111
Threads: 2
Joined: Nov 2017
See, one thing I love about AI Survivor is that it's the only competition I follow where the participants DON'T ever change or learn anything. They're the same old idiots year after year and I think it's a lot of fun to see who will make it how far THIS time?
Too much stuff is changing and updating all the time. I'm glad to have something like this that remains mostly constant.
(Last season was also the most I've enjoyed this for years, I think, so that influences my opinion for sure.)
Posts: 587
Threads: 7
Joined: Apr 2016
Posts: 11
Threads: 0
Joined: Feb 2011
FYI - Folks are already getting pumped up over at CFC. An announcement has been posted on the front page. Survivor has become hugely popular among the IV denizens over the years.
Posts: 3,741
Threads: 26
Joined: Sep 2010
Unfortunately I'll be too busy this year to really take part in the competition so I won't be submitting picks.
In previous years I also did a random.org pick as well for the games. If anybody wants to take that up, I'd be honoured. I allowed the random picks to be almost unlimited with a few caveats, 1) if a civ was picked 1st, it couldn't also be 2nd or first to die (and so on), 2) winning date was picked on a roll of turns 200-450 (except a time win pick always got turn 500) and 3) wars were rolled on between 5 and 25 wars total (from empirical watching over the seasons I found these to be a reasonable outer bounds for wars in a game).
Travelling on a mote of dust, suspended in a sunbeam.
Posts: 5,607
Threads: 47
Joined: Mar 2007
Pretty sure random.org has beaten my picks the past couple seasons. Maybe I should submit random stuff all the time.
Posts: 3,924
Threads: 19
Joined: May 2011
Oh man I saw this come up but I totally forgot about the fantasy stuff and got late to it, oh well. At least I still have time for predictions! Always love this series.
Posts: 111
Threads: 2
Joined: Nov 2017
(May 23rd, 2024, 13:05)Brian Shanahan Wrote: Unfortunately I'll be too busy this year to really take part in the competition so I won't be submitting picks.
In previous years I also did a random.org pick as well for the games. If anybody wants to take that up, I'd be honoured. I allowed the random picks to be almost unlimited with a few caveats, 1) if a civ was picked 1st, it couldn't also be 2nd or first to die (and so on), 2) winning date was picked on a roll of turns 200-450 (except a time win pick always got turn 500) and 3) wars were rolled on between 5 and 25 wars total (from empirical watching over the seasons I found these to be a reasonable outer bounds for wars in a game).
Just saw this, checked, and saw that nobody else has picked up the torch, so I shall take the responsibility of keeping random.org alive in the picking contest!
I'm tightening up the parameters somewhat so that it doesn't spit out as many outlier results. For this week I went for more common ranges of # of wars and finish date: 7-17 for the former and 250-350 for the latter. I think those could maybe be adjusted more in future though; I'm thinking of an upper bound of 400 for finish date instead. Maybe extend wars back as well?
I'll also remove time victory from the pool for guessing future games since that is never going to happen. Also considering weighting the First to Die numbers to make "None" significantly less likely than the other options, but I'm not sure if that's going too far.
In any case, this week's random prediction was a Turn 262 Domination win for Hammurabi - seems pretty accurate, right? Other picks were Peter for 2nd, Boudy first to die, and 14 total wars.
Posts: 5,607
Threads: 47
Joined: Mar 2007
Thanks for picking up the torch, Eauxps!
Has there ever been an AI Survivor game where no AIs died? It is theoretically possible, I suppose, but so vanishingly unlikely that it should probably not be part of the regular random picks. And narrowing the end date range seems like a good idea as well; we have a pretty good idea at this point for what is likely.
For this game: 14 wars, and that is with Boudica eliminated first? Random.org is predicting a bloody game.... I guess it thinks the Hammer has finally had enough, and is going to dish out the pain to all those AIs who think they are warmongers.
|