AI Theory

Benutzeravatar
UesugiKenshin
Admiral
Admiral
Beiträge: 3561
phpbb forum styles
Registriert: Dienstag 31. März 2020, 03:51

Re: AI Theory

Beitrag von UesugiKenshin »

The only logical way to make it fair is preset save games as the AI can't manipuate commodity prices. And when the AI and the humans have a balanced economy and military forces, then major battles will take place, not pointless skirmishes.

Death by a thousand paper cuts as probes leading to skirmishes and death by the mayhem of outrageous sabotage. These are not effective strategies for waging war.
Minor Race Database http://birth-of-the-empires.de/wiki/ind ... r_(Minors)
Minor Race Ships http://birth-of-the-empires.de/wiki/ind ... nen_Völker
Chrome and Opera autotranslates German.
Benutzeravatar
UesugiKenshin
Admiral
Admiral
Beiträge: 3561
Registriert: Dienstag 31. März 2020, 03:51

Re: AI Theory

Beitrag von UesugiKenshin »

The only way wars can be waged is deficit spending. You cannot simulate this in BotE. So you have to have a treasury with excellent cashflow and a LARGE surplus as upkeep is likely to exceed taxation and so wars break treasuries in history and induce peace negotiations.

The war cripples the empires' economies.

Every single thing I did for the sandbox save game at turn #700 has a military and economic and political justification.

A good scenario or campaign would do the same thing for a one empire versus another empire test.
Minor Race Database http://birth-of-the-empires.de/wiki/ind ... r_(Minors)
Minor Race Ships http://birth-of-the-empires.de/wiki/ind ... nen_Völker
Chrome and Opera autotranslates German.
Benutzeravatar
UesugiKenshin
Admiral
Admiral
Beiträge: 3561
Registriert: Dienstag 31. März 2020, 03:51

Re: AI Theory

Beitrag von UesugiKenshin »

The AI does not grasp efficient construction of infrastructure. See the advanced saved turn #700 topic. Earlier in that topic, I posted a formula for infrastructure that seems optimal. Maybe that shoud be the default setting for the AI. I think the AI would be in better shape. It tries instead to build too many lower level buildings as they seem to take less individual turn time, so the AI is not considering many moves ahead, just playing chess like 3 moves ahead instead of 10+.
Minor Race Database http://birth-of-the-empires.de/wiki/ind ... r_(Minors)
Minor Race Ships http://birth-of-the-empires.de/wiki/ind ... nen_Völker
Chrome and Opera autotranslates German.
Benutzeravatar
UesugiKenshin
Admiral
Admiral
Beiträge: 3561
Registriert: Dienstag 31. März 2020, 03:51

Re: AI Theory

Beitrag von UesugiKenshin »

See the advanced turn#700 discussion.
viewtopic.php?f=107&t=7294&start=60
The AI does not grasp commodity manipulation (which actually is why gold prices stay low yet traders using AI actually do make millions on even small price movements).

This means the human player can always beat the AI by making huge profits and it's not necessary to buy monopolies. Then outright buy whole starships every turn once learning advanced technology.
Minor Race Database http://birth-of-the-empires.de/wiki/ind ... r_(Minors)
Minor Race Ships http://birth-of-the-empires.de/wiki/ind ... nen_Völker
Chrome and Opera autotranslates German.
Benutzeravatar
UesugiKenshin
Admiral
Admiral
Beiträge: 3561
Registriert: Dienstag 31. März 2020, 03:51

Re: AI Theory

Beitrag von UesugiKenshin »

The best way to help the AI would be "AI only offensive tactics". It's not fair, but then they would get a slightly better chance of ground invasions.

Rght now the Omega can create unassailable defenses, if they choose to as they have infinite money. The only thing slowing down conquest is the maximim rate of speed of 4.2 sectors/turn and that is reduced by fleet operations as some starships are slower.
Minor Race Database http://birth-of-the-empires.de/wiki/ind ... r_(Minors)
Minor Race Ships http://birth-of-the-empires.de/wiki/ind ... nen_Völker
Chrome and Opera autotranslates German.
Benutzeravatar
UesugiKenshin
Admiral
Admiral
Beiträge: 3561
Registriert: Dienstag 31. März 2020, 03:51

Re: AI Theory

Beitrag von UesugiKenshin »

Since resource routes are critical, the thing that would make sense is some kind of piracy which reduces resource route efficiency. So not actual freebooters launching attacks, (though that would be fun), but either sabotage affects it or the AI only can reduce resource route percentages.

Then less human player construction can take place. Then ground invasions by the AI are technically easier in an indirect way.
Minor Race Database http://birth-of-the-empires.de/wiki/ind ... r_(Minors)
Minor Race Ships http://birth-of-the-empires.de/wiki/ind ... nen_Völker
Chrome and Opera autotranslates German.
Benutzeravatar
UesugiKenshin
Admiral
Admiral
Beiträge: 3561
Registriert: Dienstag 31. März 2020, 03:51

Re: AI Theory

Beitrag von UesugiKenshin »

The Khaoran seems doggedly determined to be aggressive and attack unescourted colony ships even when it's pointless to do so.

The Terrans and Heyoun are apparently programmed to explore an opponent's sphere of influence even when doing so is provocative.

A chief issue provoking unnecessary conflict is an inability to "open hailing frequencies" to elict a first contact. The AI has to do it meaning they get a first opportunity to attack.

Some minor aliens will attack immediately, but mostly they don't even when they could crush a helpless scout as a "spy" within their home system.
Minor Race Database http://birth-of-the-empires.de/wiki/ind ... r_(Minors)
Minor Race Ships http://birth-of-the-empires.de/wiki/ind ... nen_Völker
Chrome and Opera autotranslates German.
Benutzeravatar
UesugiKenshin
Admiral
Admiral
Beiträge: 3561
Registriert: Dienstag 31. März 2020, 03:51

Re: AI Theory

Beitrag von UesugiKenshin »

Why does the AI build ships to attack the player or other AI so early in the game? Because human beings want war to happen.

This starts in childhood as most kids avoid fighting, but some are picked on by bullies. So some learn to fight and the rest, while not wanting to get hurt themselves, do learn to love violence.

So my theory is the AI is warlike, not because it actually helps the AI empires to grow, but to entertain and challenge the human players.

On a 40x40 map with maximal stars and so a wide open galaxy for terraforming and having many colonized planets in sectors, instead the AI wages war too early as a distraction.

See, a small percentage will fight based on possessing the intellect, experience, confidence, a standing capable experienced ground invasion force, a sizable diverse group of veteran ships and fleets, knowledge of tactics and strategy and short, middle, and long term goals in a comprehensive plan.

That is a LOT of qualifiers. So most players will not be aggressive.

So the AI need not be aggressive...until it can afford the military upkeep, and it has a balanced yet growing economy, with adequate infrastructure, and has done better research, so the recruitment is beneficial.

Actually AI defenses are more important.
Zuletzt geändert von UesugiKenshin am Samstag 26. September 2020, 15:24, insgesamt 1-mal geändert.
Minor Race Database http://birth-of-the-empires.de/wiki/ind ... r_(Minors)
Minor Race Ships http://birth-of-the-empires.de/wiki/ind ... nen_Völker
Chrome and Opera autotranslates German.
Benutzeravatar
UesugiKenshin
Admiral
Admiral
Beiträge: 3561
Registriert: Dienstag 31. März 2020, 03:51

Re: AI Theory

Beitrag von UesugiKenshin »

Unless one has all those military qualifiers and economic REQUIREMENTS, and a good command of the principles of Sun Tzu, then whomever attacks first, ultimately loses the galactic war.

The primary antagonist then is a THREAT. It's called the top poppy theory. The taller poppy creates shade and so competes against the others, so in AI theory, attack the tall poppy. So this negates the efforts of the first attacker.

Others join in and so they deplete their military forces through attrition. Meanwhile whomever stays out of the fray, gains economic power and so can invest in research, and so make superior ships and troops.

Eventually they will win as long as they have a command of the tactics and strategies of Sun Tzu.
Zuletzt geändert von UesugiKenshin am Samstag 26. September 2020, 16:21, insgesamt 1-mal geändert.
Minor Race Database http://birth-of-the-empires.de/wiki/ind ... r_(Minors)
Minor Race Ships http://birth-of-the-empires.de/wiki/ind ... nen_Völker
Chrome and Opera autotranslates German.
Benutzeravatar
UesugiKenshin
Admiral
Admiral
Beiträge: 3561
Registriert: Dienstag 31. März 2020, 03:51

Re: AI Theory

Beitrag von UesugiKenshin »

It's actualy worse. Whoever attacks first as the tall poppy, and since most lack military forces (unless they are Terrans and absorbing minor alien pre-existing fleets), then the cheapest way to attack is espionage and sabotage.

That is why even Terrans and Heyoun, who get no morale bonus for doing so (unlike the Carta, Rotharians, and Omega) still sabotage as much as they can. The Terrans and Heyoun are acting against their own ethos because it whittles down their perceived enemies.

http://birth-of-the-empires.de/wiki/ind ... Moral.data
The columns are: Terran, Heyoun, Khaoran, Cartare, Rotharians, and Omega.
Successful espionage 0 0 0 2 2 4
Successful sabotage 0 0 -1 5 5 2
There is no morale increase for Terrans and Heyoun to do espionage or sabotage, yet they cnsistently utilize them. Of course the impetus for the Cartare, Rothariansm and Omega are to do sabotage, with the Omega liking espionage more.
Minor Race Database http://birth-of-the-empires.de/wiki/ind ... r_(Minors)
Minor Race Ships http://birth-of-the-empires.de/wiki/ind ... nen_Völker
Chrome and Opera autotranslates German.
Benutzeravatar
UesugiKenshin
Admiral
Admiral
Beiträge: 3561
Registriert: Dienstag 31. März 2020, 03:51

Re: AI Theory

Beitrag von UesugiKenshin »

A better AI would be defensive and prioritizes this based on actual threat by proximity and ability. But they would only react to the potential to gain pre-existing minor alien fleets through diplomacy as the major aliens lack the capacity to wage effectual war.

The AI would grow and then be fearsome...as it should do in a giant galaxy. You don't wage war effectually without fortification of the home sector.

Some players create raiding forces to negate colonization and escort a colony ship, and then steal that improved sector. But the AI focuses, like a laser beam, on M class systems as it can immediately take them upon finding them.
Minor Race Database http://birth-of-the-empires.de/wiki/ind ... r_(Minors)
Minor Race Ships http://birth-of-the-empires.de/wiki/ind ... nen_Völker
Chrome and Opera autotranslates German.
Benutzeravatar
UesugiKenshin
Admiral
Admiral
Beiträge: 3561
Registriert: Dienstag 31. März 2020, 03:51

Re: AI Theory

Beitrag von UesugiKenshin »

I presume the concern is about the AI being defensive and worried that clever aggressive players would emulate Viking raiders in the Dark Ages. Well, historically...that didn't work as you can't occupy and hold it. That failed with feudalism as the antidote.

Those kinds of players are rare
Minor Race Database http://birth-of-the-empires.de/wiki/ind ... r_(Minors)
Minor Race Ships http://birth-of-the-empires.de/wiki/ind ... nen_Völker
Chrome and Opera autotranslates German.
Benutzeravatar
UesugiKenshin
Admiral
Admiral
Beiträge: 3561
Registriert: Dienstag 31. März 2020, 03:51

Re: AI Theory

Beitrag von UesugiKenshin »

I believe that after modding sabotage, resource routes, and deritium, since information warfare and asymmetric warfare through sabotage are not available until many hundreds of turns later that BotE will seem to be more accurate. Why? Because intelligence networks take time to get in place. You can't have organized ones with facilitators and sleeper agents and spies until they can actually get to these sectors. It takes time for starship travel to develop on a huge galaxy. You can't be running sabotage missions if no one has even visited far to the northeast or far to the southwest, right?

So war will be conventional for the longest time. The way those agents would get in place would probaby be on slow freighter transports. In rare circumstances by cloaked vessels and by pirates paid to transport them.

The very best way to test my theories is by running the BotE beta multiplayer and playing as all empires, and when I do, I think I'll set up each of the six empires by building it as I do now. I focus on infrastructure, then defense, not starship battles and invasions that realistically would be futile. You can win engagements and invade, but it's not necessary and costs a lot of money...needlessly.

The issue is to play six empires for 600 turns is 3600 turns so likely 4-6 months of play. It's a huge time commitment, and I have 3 versions of a mod planned.

But if six players tried each doing one empire, it would get done in a month. And if saved as a single player game, the player would be strong as would each empire. Then I believe the AI would act far wiser as they are stable and with a large treasury and with high research and with better ships and troops and many colonial sectors.

It's the long game and it would be very entertaining and challenging to play.
Minor Race Database http://birth-of-the-empires.de/wiki/ind ... r_(Minors)
Minor Race Ships http://birth-of-the-empires.de/wiki/ind ... nen_Völker
Chrome and Opera autotranslates German.
Benutzeravatar
UesugiKenshin
Admiral
Admiral
Beiträge: 3561
Registriert: Dienstag 31. März 2020, 03:51

Re: AI Theory

Beitrag von UesugiKenshin »

The standard way to beta test is to have a mechanism to design the land and assign territory, place infrastructure, research levels, ships, troops, diplomatic relationships, and money. Then the idea I have could be built in a week and tested in a month...not six months.

Alternatively you could assign a large treasury and better starting infrastructure and a colony ships. So the AI would dispatch them and be managing 20 colonies not one. But we have no way to assign a large treasury either.

I could place multiple starbases in each of six empires with lots of colony ships on turn one and with a developed infrastructure, then scrap the starbases all for cash. So the AI then would have a large treasury and capability to do a giant wave of colonization.

But the AI is not wise about colonization. It likes to grab up all the M class sectors, even when it has no chance of holding them later. It's self-sabotage.

I guess I could do a half measure and play 100 turns for all six empires so that would only take a single month. The best test would be six hundred turns for all six empires for I know I can create six wealthy powerful stable empires...none of which would be worried about sabotage...yet.
Minor Race Database http://birth-of-the-empires.de/wiki/ind ... r_(Minors)
Minor Race Ships http://birth-of-the-empires.de/wiki/ind ... nen_Völker
Chrome and Opera autotranslates German.
Benutzeravatar
UesugiKenshin
Admiral
Admiral
Beiträge: 3561
Registriert: Dienstag 31. März 2020, 03:51

Re: AI Theory

Beitrag von UesugiKenshin »

In my opinion, beginners would play new empires.

Then better players would start advanced turn#250+ games. So they start in turn #251. They have a turnkey kingdom as do the following based on starting turn number.

Then middling players would start in turn#500.

Then great players would start in turn #600.

Then excellent players would start in turn#700.

Each has various galaxy sizes and challenges pre-set in these scenarios or campaigns. I think it would be far more fun. And in each case, the AI begins from a position of strength, not weakness.
Minor Race Database http://birth-of-the-empires.de/wiki/ind ... r_(Minors)
Minor Race Ships http://birth-of-the-empires.de/wiki/ind ... nen_Völker
Chrome and Opera autotranslates German.
Antworten

Zurück zu „BotE Android - Questions and Answers“