new AI is coming... GamesGod AI

[GamesGod]

Wow! This is such a fun AI! I was testing SilverTongue against it and GG was being pushed and out of nowhere five skirmishers find their way into the back of the eco. I can't believe it has that kind of micro in it!

Do u have replay this game? I Want see it

Thanks, yes, my AI is trying to find different places to attack he does not always attack from one side) I try to make my AI as strong and interesting as possible for players there will be an update soon that should noticeably improve the work of my AI, there will be a lot of fixes in it

[Rorrik]

I watched the replay and I was mistaken, the skirmishers had somehow chased a ram repair vill back into the base from that direction. When I was running it, I thought they had come out of nowhere, possibly from a sneak range.

[GamesGod]

I watched the replay and I was mistaken, the skirmishers had somehow chased a ram repair vill back into the base from that direction. When I was running it, I thought they had come out of nowhere, possibly from a sneak range.

hmmm.... u have old version my ai and your ai simple win u need download last version

[Rorrik]

I might, its the version of King of Arabia.

[marathon]

It is a fun AI to play against. I sometimes do 4v4 solo games with various combinations of AIs and GamesGod has earned a slot in my pool of preferred AIs.

It does not seem to produce lag. It fights and does not give up when things go bad.

It has shown good performance in FFA games, although it seems passive in these situations. However, being passive is often a plus in FFA.

Its fighting capabilities are not that good. It is a bit better than typical pre-UP AIs but still weaker than most modern AIs.

I noticed that in a game as Britons it had 17 lumberjacks and over 70 farmers in late game. Do not neglect wood. 17 lumberjacks is a number for Gothic Champions, not Briton archery. However, for late game situations it is good to block selling wood and only sell food (so allow more farmers than needed).

It tributes to allies, it keeps fighting, it relocates, it does all the good stuff that make it fun to play with. It has a variety of army compositions, so it is fun to play against. You don't really need to change anything, it is good already for long games.
But if you want to make it able to compete with top AIs, you would need to put some effort in it. A solid early game strategy and an improved fighting system are the 2 most important things.

[GamesGod]

Yes, thank you for the positive assessment of my AI's work, as for resource extraction, the economy is completely dynamic, and it extracts those resources that are currently lacking, but they are needed. If he got more food while playing for the British, it means that he had an acute shortage of food well, for example, for the production of peasants or the study of research:) my AI is not very attached to civilization, so if he plays for the British, he can attack with knights or someone else and not just archers this is done specifically so that a human player could not initially know the composition of the AI army. After all, it is quite obvious that if the AI produces only archers for the British, it is very easy for a person to win such an AI, it is enough to make knights, skirmishers and onagers... And my task is to create an AI that will win by non-standard behavior. So that a person could not "know" the enemy's army without intelligence So that a person has to organize reconnaissance my AI is primarily aimed at making it interesting for human players to play against it computer players, of course, have no interest at the moment, I am just engaged in improving the combat system, the system of assistance to allies, this will make it even stronger

[marathon]

Acute shortages should be addressed, of course. But violent retaskings cost a lot in the long run. My own dynamic system is proven to be less efficient than Horde's static system. I measured it.

Long ago I had conceived a way to quantify the depth of any given shortage or surplus, using the same logic that is used in the mathematical notion of interval. But the simplest idea is to undercorrect. For the rest of the shortage, simply change unit composition or use market. Anyway you will find your way.

Unpredictability is good. But the cost of tech switch is high in early game. Switching from archers into knights, then into champions then into pikes+cavarchers, this costs. Every strategy has to sacrifice options. Going for everything costs.
Crusade is not efficient enough for its potential, because it does a little bit of everything.
Eagles Revenge, Strongbow, even Boss, they all completely neglect popular options. Boss never builds a stable not even with Huns!

But I understand and I feel the appeal of such tech switches. Ruler had made an AI that I loved: it would switch between trash and gold army. Trash army could not hold the enemy, but as he was approaching his lines of supply lengthened, while own lines became shorter. Then, suddenly, block all trash and pump out cavaliers, champions, scorpions, arbalests. Enemy front collapses, expensive siege that follows gets shattered. Really cool concept!

Darius is one of BTG AIs, and it is included in BTG Best Strats. It is Persians, but in castle age it goes for crossbows, camels and scorpions. Then in imperial it follows stages, resulting into Elite Elephants if all things go north.

Tech switching in late game is not only cheaper, but also fun and interesting for long games. A huge, long game can become boring without some peppering. A suicide cavarcher squad in your trade, a sudden switch into champ flood, stuff like these can make the game very fun. And I have suffered a lot from your own onagers lol! I know they will present themselves eventually.


I did not yet play and read your AI enough. It is worth it. More feedback will follow then.

[GamesGod]

In fact, my AI has many different strategies, but some of them are disabled, and the other part is disabled on purpose as soon as my ai sees the accumulation of some type of army at the enemy my ai always dynamically changes the composition of its army based on what kind of army it sees from the enemy therefore, it is quite difficult to defeat my AI with one type of unit. But, this is if he was able to scout the enemy's army in time) of course, he, as well as a person, may simply not notice how 40+ paladins are running at him, because he is honest)) and simply will not have time to make spearmen or camels)) but if the reconnaissance was successful, then all these paladins will meet a large number of spearmen and camels in defense)) if of course there are resources for this in addition, when the game is 2x2 or 3x3, for example, there are situations when my AI creates an army against one enemy, with one type of army, and as a result another enemy attacks, a completely different type of army) of course, this immediately leads to defeat

[FireBall37]

I like ur ai, watching it in the Arena tournament was really fun. I love how diverse you have made it, it is really fun to watch. Will there be another update soon? I'm stuck on my ai

[GamesGod]

I like ur ai, watching it in the Arena tournament was really fun. I love how diverse you have made it, it is really fun to watch. Will there be another update soon? I'm stuck on my ai

Thank you, this is exactly the main goal of my AI, so that it would be interesting to play with it and watch its game but of course, at the same time, I try to do everything possible so that he plays hard of course, this is the latest version of my AI that is currently on the forum, it looks more like an entertaining one than a serious enemy/ally... But the update will fix it it's a pity, of course, that the update is so delayed, but I just have to make too many changes, because this is a universal AI, and not just playing with one civilization on some 1 map I don't know yet if I will indicate the list of corrections that I have made, most likely I won't, I wonder what changes the players themselves will notice