AgeofStoneWalls

[Tonto_DaVe]

Okay, I think I've done enough testing to feel some confidence in bug minimisation. So attached is AgeofStoneWalls.

It will play Aztecs (I know, disappointing) for the AoKH ladder, although it seems to play Spanish (given time) pretty well, along with Mayans, and I think it did decently with Koreans (and perhaps Turks, but I did not test them much). I haven't tested it with Brits, but I'd assume it'd still be fairly effective. It has been coded for every civ but Franks and Huns (i.e. it will not play those civs properly).

From my tests, Az Age is very effective versus the likes of De'gel, The Horde and most other AIs, but aggressive ones like ER, Eyries, and even Lade 1.0 Mayans it struggles, if not outright gets dismantled.

Edit: updated to (hopefully, but the buffer SN is pretty suss) limit dropsites in the center of the enemy base till it can defend it, and will get EEW as Az before bracer and chemistry 100%.

[marathon]

Very interesting strategy! A fast imperial with good booming too. I think that I've never seen something like that before.

A test against my Mayans proved its weakness against determined rushers. It's fine, every strategy has its known weaknesses.
A second test against Degel's Aztecs denostrated its strategy: it hit 32:xx Imp, its booming was impressive, with ~80 villagers and another 25 soldiers at 30', it reached 95 Elite Eagles at some point. Yet, it lost. It had the initiative for about half an hour and this is a feat on its own when against Degel.

- I believe that it could use some skirms. They save gold, they have some resistance under building fire, they are somewhat effective against Jaguar Warriors. Plus, all relating researches are taken for xbows anyway.
- It should take Forging a little sooner. Not because it is super effective, but because it takes a lot of time to get all blacksmith researches with only 1 blacksmith. Despite being in Imperial 4-5' earlier, it was only marginally faster with upgrades. And it never took Garland Wars.
- Perhaps it could use more rams.
- I had the impression that there were some imbalances in early-mid castle age, but the super fast imperial proved me wrong.

Now I'm going to test it against my Byzantines. I expect it to beat them easily. I'll study it. It's good.

[Tonto_DaVe]

Thanks for the quick test marathon, and the kind words, but the result is pretty surprising to me since it has been a while since Age lost to De'gel. When I first tested Chief it got blatantly dismantled versus De'gel, so I decided to go back to the drawing board and produced Age. Generally I've found every game over between 43 and 50 minutes.

I did a quick test again just a little while ago, and even with 2 -ing issues - losing 5-6 vils on the way to imperial because of a mining camp within the enemy base, and getting bracer before EEW - it still won in 45 minutes.

http://files.foraster.us/download.php?f ... 72a99abf9d is the rec.

If you recorded that game I would love to check it out to see what was going on. I don't think De'gel can respond to EEW in the castle age - no AI can imo. One guess I have is that maybe De'gel had a very nice hill on approach to her base that allowed her to defend against any sort of army.

Having said all that, I do have to agree with BSes/rams. I hadn't actually thought of a 2nd smith at all, despite very commonly doing it when I actually play on Voobly. Rams had been downgraded in use a couple days ago, from being a waste of castle-age res, to now probably being too little of a factor.

Edit: oh and ha, with the res imbalance. You're right, actually. Huge imbalance of res, but that's the fastest way to imperial with AIs - buying your way there. Hopefully when we can build more than one farm at a time that will change.

[marathon]

I didn't record that test. I ran it at slow speed and 3x speedhack (Cheatengine). An enemy castle played its role.
Degel had ~150 conversions when I stopped the game (it was already decided at that point).
What made me is that, although it had a very powerful eco with nearly 80 villagers at 30', and a not that small army too, it suddenly pressed the button. I didn't see it coming lol.

The strategy itself is very interesting and powerful against a wide spectrum of ais. Trying to change it to the better might result in ruining it.

For civs other than Aztecs the imperial isn't that fast. I tested Turks and they were beating my Byz until the (sn-minimum-dropsite-buffer == 0) bug hit Turks: they dropped a camp next to a Byz castle...

I clearly remember having said that I set sn-defence-distance to zero. But perhaps a small defence-distance might help villagers avoid endangered zones. I haven't tested it. Anyway, even your Turks would have beaten my Byz (they were 3000 pts ahead).

I actually played against your script. I couldn't beat it (I am a lousy player lol). I decided to test it on Regicide. I lost even faster hahaha. In a potential future Regicide or Fortress tournament this script will have a good chance to win as it is now.

[Tonto_DaVe]

lolz, fun stuff. cheers for those SN hints, I've realised I've got a lot to relearn when it comes to that stuff; a lot of improvements to the general behaviour that I can make.

btw, where can I find this 3x speedhack cheatengine thing? Read a lot about it, but I can't find it so far. Testing is getting to almost absolute torture levels in the dark and feudal ages.

[scripter64]

Hi Dave! Nice to meet you, and it's great to see your new AIs

You can download Cheat Engine here. I downloaded "CheatEngine60NoSetup.rar" instead of the installer. To use it, start the game, start Cheat Engine, select age2_x1 from the Process menu's running processes list, click "enable speedhack" on the right side, and then configure the speed setting using the controls that appear below.

[Campeador]

This script is really strong. It reaches to imperial quite fast, despite to train enough units for a good castle age war. In my first test, it defeated my Britons (despite to thie dynamic gatherer percentages and well balanced eco). I though that it was because they were unlucky in the Dark Age (with a 15' feudal time, while the most usual for them is 13:30-14:00). But in the second test it defeated my Mayans, despite to my 13:30 feudal (who built the need archeries inmediately once arrived to feudal age).

You script will be quite problematic for your rivals in the Ladder and a good addition for human players. I still prefer No Limits because I like superagressive scripts for Arabia 1vs1, but this is only personal taste. AgeofStoneWall is also very good.

[Tonto_DaVe]

I can definitely understand your point of view, Campeador. Super aggressive AIs are so fun - which is why I made No Limits - but I think I just got tired of scripting a Hun flushing AI as that's what I've been doing since 2005.

[MuRRay Jnr]

No limits is one of my all time favourite scripts. back when we did the scandinavia tournement, i tested mostly against no limits as it was probably one of the best scandinavia ai's at the time. still is.

[Tonto_DaVe]

murray man, you have no idea how cool it is to hear that 11.

Haven't done any work on anything AoC related for a couple weeks, but decided to upload the latest (8/5/11) version that was just laying around.