For those of you interested in seeing in progress AI updates I've been working on over the holiday you can try the latest build here:
https://dl.dropbox.com/u/8051911/FE-119frog.zip
Just unzip the EXE into your FE directory.
The biggest changes should be on larger maps. However, be on the look out for longer turn times since I've had the AI do more analysis to try to make it smarter.
Tested on my better laptop, and the AI is much more annoying and the performance isn't as impacted.
So I would want to keep this improved AI, but it needs to be a slider.
One suggested gameplay change: allow non-fortresses to build walls and defensive improvements. I think it needs to be easier to build up static defense.
So what happens when the game is updated through Steam. Will it automatically update it to the then current build, or will we have to reinstall to remove the beta stuff?
I fixed this this afternoon. Though it's only on my machine right now. It'll be part of 1.2.
Nice! Looking forward to 1.2. Any ETA or changlog coming up soon?
Thanks boy!
I am really looking forward to 1.2. The game just keeps getting better.
Two things I hope are fixed in the next version are:
1. Stuck heroes as described in some of the above posts. Worst instance of this I saw was the enemy's sovereign stuck adjacent to one of its cities as I approached. It would have been a tough, interesting battle if the sovereign had defended the city.
2. Enemy units remaining in a city after capture. This is the only game-wrecking error I have encountered, and it's happened twice. I documented the latest instance in this post:
https://forums.elementalgame.com/438866
Thanks, Frogboy, for all your work on this great game!
That's a Kael thing. I just program.
1. Should be fixed by the linked EXE.
2. I've never seen this before. Obviously shouldn't be possible. Do you have a saved game with it?
I'm in a large game, turn 200, ridiculous AI and I believe the AI is significantly better now. It still like to put a stack on idle a few turns, or don't try to concentrate enough but there is much more 'brain power' behind its moves I believe.
So, good job Brad!!
Any possibility to enliven the game with pro-active diplomacy and dialog? I like when the AI chit-chat in Gal Civ, saying 'you believe I don't see you massing at my borders' or messages like that, which convey the feeling the AI understand the world and see what you plot?
Yes. Please see my post which has a link to save games:
EDIT: Here is my original post when it happened the first time. I don't have any saves for this one, but a couple people mentioned having something similar happen to them.
https://forums.elementalgame.com/438008
also militias don't appearing when fighting multiple battles per turn attcking city.
Is this fixed?
Got some weird save game bug with this. I failed bad at a mob, and reloaded. But then the mob was damaged from the fight before? My stuff was still at full health so very weird.
(One Obsidian golem that I killed was at 1 hp, and easy as hell to get next time too)
Here is the unofficial change log of what I put in. Note that my balance changes are proposed, they may not actually go in officially.
Brad (1:00 PM)
General+ Version string updated to v1.20+ Fixed glitch that caused the arrow cursor to show instead of the hour glass when the player dragged the map between turns (yea, obscure but it bugged me)+ Reduced ground cover on terrain with the exception of deserts
AI+ Fixed bug where AI knowledge trading didn't always reduce tradeable knowledge+ AI more intelligent about when and what it builds in its cities+ AI evaluates whether it should be training archers/catapults/mounted units more effectively+ Fixed but that caused AI to disproportionately choose the first level up perk (like Assassin)+ AI more aggressive about getting to goodie huts+ Monsters still raze city but it no longer salts the land+ Higher level monsters will now only spawn in barren land (no more dragons as neighbors)+ Fixed bug where AI units would sometimes get stuck (like watching while their city was taken over or not going out to get goodie huts or going after an easy kill unit) (this was a pain in the rear to fix)
Balance (proposed)+ Anicent temples now provide 2 knowledge per turn+ Twilight honey now provides a faction wide -2% to unrest+ Gold deposits more common but gildar provided reduced from 6 to 3+ Orchards now provide a 3% HP bonus faction wide+ Default # of supported players per map increased on small, mediu, and large+ Grain provies 25 food instead of 20 by default+ Pioneers now cost 75 gildar to produce+ Archer based units take longer to train (typcally 1 to 2 extra turns more)+ Modified rarity of different world resources and their availability to have more variety+ Resources are substantially more common+ Goodie huts are slightly more common+ Quests are slightly rarer+ Shards are slightly rarer+ Monster lairs are rarer on default settings (dense setting actually increased and sparse has fewer monsters)+ Tech multiplier slightly increased+ Epic pacing multiplier increased+ Default turn limit reduced to 800 (from 1000)+ Over encumbered reduces initiative by 8 (was 6)+ Heavy ecumberance reduces initiative by 5 (was 4)
You are such a tease, but I do agree with your decision to leave the "power" in the hands of the head of development, (can be hard for some people)
Is there a particular reason for this?Just curious anyways.
I hope to see more bug fixes , my brother discarded the game due to the numerous bugs he encountered.It is good to see these suggestions, not sure I like the idea behind global bonus hp and unrest though, but thats up to you (Just another reason that more means more, instead of having lesser empire's having a fighting chance).
So many turns I probably would turn different, guess that shows the difference in taste, hope it gets good and it seems like something I want to play with for some time to see the differences, thanks for taking the time!
Sincerely~ Kongdej
Changelog looks interesting. Not sure about the balance changes but we'll have to try it and see. Are you going to update your download?
Awesome!!! I really can not wait for release on this stuff!!! Do monsters still raze the city on turn 1, or does it take them at least a few turns?
Here are my favorite AI fixes:
AI+ Fixed bug where AI knowledge trading didn't always reduce tradeable knowledge + Fixed but that caused AI to disproportionately choose the first level up perk (like Assassin) + Monsters still raze city but it no longer salts the land + AI more aggressive about getting to goodie huts + Higher level monsters will now only spawn in barren land (no more dragons as neighbors) + Fixed bug where AI units would sometimes get stuck (like watching while their city was taken over or not going out to get goodie huts or going after an easy kill unit) (this was a pain in the rear to fix)
AI+ Fixed bug where AI knowledge trading didn't always reduce tradeable knowledge
+ Fixed but that caused AI to disproportionately choose the first level up perk (like Assassin)
+ Monsters still raze city but it no longer salts the land
+ AI more aggressive about getting to goodie huts
+ Higher level monsters will now only spawn in barren land (no more dragons as neighbors)
+ Fixed bug where AI units would sometimes get stuck (like watching while their city was taken over or not going out to get goodie huts or going after an easy kill unit) (this was a pain in the rear to fix)
Here are my thoughts on some of the balance changes:
+ Default # of supported players per map increased on small, medium, and large (Great)+ Pioneers now cost 75 gildar to produce (Great)+ Archer based units take longer to train (typcally 1 to 2 extra turns more) (It is a start to making ranged less OP)+ Default turn limit reduced to 800 (from 1000) (Bad, why change this?)
Are ancient temples coming back into the sandbox mode? I never see them.
I like the balance changes listed. I think encumbrance penalties should be 50 and 25% though.
The balance changes look really interesting, especially the pioneer cost. I have to ask as I haven't played 1.12 yet: is the AI still able to break NAPs and declare war if I'm paying tribute? Because if it is, it should be fixed.
I don't like that pioneer gold cost one bit. It makes taking wealthy and selling all your items the most winning strategy.
If you want to limit early game pioneer spam, just make building a pioneer drain 20-30 population from the town it got built in.
I would support this as well. Anything to reduce pioneer spam. Hell, anything to reduce city spam in general. I hate having a ton of cities, but there is no benefit to having fewer cities and if you don't race to get your own cities, the AI will, and then they get all the resources and grow too powerful. Argh! Anyway, I would love to see some mechanic to make city spam less attractive.
I agree that the gold cost is the wrong way to go, but why would it take 20 - 30 pop to build an outpost? I know - it's a meta-game thing - but why?
I personally like the Civ 5 solution of halting population growth in the city while training pioneers(settlers).
Since Derek removed/altered how production, research, and taxes work - taking the bonus away from the actual population and moving it to city levels- going 'tall' has basically become an invalid strategy. Spamming cities early is far superior to gaining city levels. In fact, I almost never go for food anymore until it's way too late in the game to be of significance.
I can only comment on bugs that affect the AI. I am not discounting that there are bugs in the game somewhere but I don't encounter them when I play. But I tend to focus on AI coding.
This is a very interesting thought. What do others think?
I don't think the Civ model would work for Pioneers because going up in population in FE isn't as critical as it is in Civ.
So far on my 1.19 frog AI improvement game (xxlarge lush, Heavenfall's mod with 7 AI, dense magic/resources, world and all AI set to ridiculous) -- noticeable improvement. AI is attacking more often, with more stacks, and going after mob lairs and unguarded loot. I've had to make 2 basic changes:
-I used to not recruit many troops from the start, instead I concentrated on infrastructure (same settings as above). Now I have to recruit a number of troops very early
-I used to sit back and wait for AI to come to me. Now when I'm declared on I have to go after them and 'roll up' a few of their cities, else I'd get overwhelmed. Rolling up a few cities gets me enough 'points' to get the AI to accept peace if I so desire, cuts back their strength, and lets me hit their sov/champs while they're still recovering from 'dying' in earlier battles (so they're not at full strength and so are easy to kill again).
These are good changes.
On the other hand, increasing AI difficulty levels increases 'strength' more than 'smarts'. AI still is pretty 'dumb'. Yes, it can now send a few stacks my way, but the threat is not from smart tactics/strategy, but from units with big numbers (attack, defense, health, etc.).
Current game I was able to force peace with the angel faction (relatives strengths at war's start was 100 for me and 200 for them, at peace time we were both ~200). At that time the frost giants were strength 1200 -- 6x my strengt! Their advantage reached 7x my strength as we began to fight. They had a stack of 1900 strength (epic) to my best stack's <500 strength. I anticipated a very tough battle, but was still able to wipe out their stack, taking no damage.
Tactics I used (that the AI needs to start using):
-stinking mud to slow down AI so my ranged units can wipe them out before they reach me
-the air spell that adds dodge vs ranged spells (I'm getting +60, making AI ranged units almost useless)
-ranged units (magic or archers) to wipe out advancing AI
-tremor and/or freeze, plus cloudwalk or the other spell to teleport my armies so I only need 1 'killer' stack vs many AI stacks
The above does require a decent pool of mana, but some advance planning and husbanding of mana isn't that hard.
My killer army just wiped out a 'stack' of 6 AI armies, each AI army had much greater 'strength' than my killer army. Combat at this level of difficulty should be exciting/scary, however, once a killer army is available combat is merely tedious.
Yes, I could stop using the above tactics, and will have to. It would be nice tho to have the AI 'smarten up' instead of me having to 'dumb down'. We're given a lot of nice spells to use, and not using them is counter to that. I was hoping that at ridiculous difficulty the AI would be sufficient challenge that using all the tools we players are given would make a good game. While the game was challenging early on, once 3-4 decent cities are up, and 1 decent killer army is available, it's challenge-over.
FE is a great game and has the potential to be even better. Good AI takes time and I believe it will continue to improve, thus my taking time to post here.
PS -- one possibly relatively easy improvement would be to have defeated sovs/champs appear not in the nearest city, as that makes it very easy to hit that city before the sov/champ heals back up to full strength, to make steamrolling the AI more difficult.
I personally like Heavenfall's suggestion. Spamming cities isn't an issue, by the speed through which this happens is in my opinion out of all proportion. Of course, giving the AI more rules to understand when it's a good time to do pioneers, given a pop drain--that's going to be tough on you. But it does add a neat strategic dynamic to things, I think.
There are many great features available to you once you register, including:
Sign in or Create Account
