AI Deathmach 3

[Hunpecked]

My thanks to Master Belisarius for running the AI Deathmatches. Based on the Round 1 & 2 reports I started a solitaire game (my third SEIVg in history) including my Terrans and the TDM-ModPack Aquilaeians, Fazrah, Narn, Tessellate, and United Flora. After seeing the recent quarterfinal results I would probably substitute the Verdurans or Pyrochette for the Fazrah, but there's always next game.

The Terrans encountered the Fazrah first, after 30-odd turns. I was surprised and pleased to see the Fazrah using mines on the common border soon after war began. Unfortunately they never used mines or sweepers in sufficient numbers, and finally surrendered. Although definitely improved over the standard AIs, the Fazrah were still slow to expand and research compared to the human player. They also neglected to launch stored satellites on several planets. On the bright side, they did seem pretty good at building facilities to match planet values, e.g. building mines mostly on mineral-rich worlds, and so on.

The Aquilaeians and Tessellate seem to have similar shortcomings; the Terrans had brief trade alliances with both before they declared war, and on the comparison chart they were each at about half the Terran level in Research, Resources, and Planets. The Aquilaeians seem to have developed some Counter-Intelligence capability; my second Ship Bomb project failed. The Tessellate, despite their lagging research, have fielded a Cruiser against the mostly LC Terran fleet. Nevertheless, as of Turn 79 the Terrans definitely have the upper hand. Unless the so-far-unseen Narn and/or UF have built a formidable empire, the Terrans seem to have a lock on this game. Perhaps next time I should try the team play option (5 AI vs human).

I look forward to the rest of AI Deathmatch 3.

[Master Belisarius]

Hunpecked said:
My thanks to Master Belisarius for running the AI Deathmatches. Based on the Round 1 & 2 reports I started a solitaire game (my third SEIVg in history) including my Terrans and the TDM-ModPack Aquilaeians, Fazrah, Narn, Tessellate, and United Flora. After seeing the recent quarterfinal results I would probably substitute the Verdurans or Pyrochette for the Fazrah, but there's always next game.

The Terrans encountered the Fazrah first, after 30-odd turns. I was surprised and pleased to see the Fazrah using mines on the common border soon after war began. Unfortunately they never used mines or sweepers in sufficient numbers, and finally surrendered. Although definitely improved over the standard AIs, the Fazrah were still slow to expand and research compared to the human player. They also neglected to launch stored satellites on several planets. On the bright side, they did seem pretty good at building facilities to match planet values, e.g. building mines mostly on mineral-rich worlds, and so on.

The Aquilaeians and Tessellate seem to have similar shortcomings; the Terrans had brief trade alliances with both before they declared war, and on the comparison chart they were each at about half the Terran level in Research, Resources, and Planets. The Aquilaeians seem to have developed some Counter-Intelligence capability; my second Ship Bomb project failed. The Tessellate, despite their lagging research, have fielded a Cruiser against the mostly LC Terran fleet. Nevertheless, as of Turn 79 the Terrans definitely have the upper hand. Unless the so-far-unseen Narn and/or UF have built a formidable empire, the Terrans seem to have a lock on this game. Perhaps next time I should try the team play option (5 AI vs human).

I look forward to the rest of AI Deathmatch 3.

Thanks for your report! Really enjoyed it and always is good to know how the AIs perform against real palyers.
Was you using High Bonus? You should play with at least low bonus.. but medium or high are more challenging.
Please tell me your game settings... and yes, the AI vs human it's a challenging option (specially in small galaxies), but maybe you are ready to play in PBW and meet human players!

[Hunpecked]

Master Belisarius writes:

>> Please tell me your game settings...

Large quadrant, Galactic Edge (?), 1 Average Planet (2000M pop) but not all players required to start with the same planet size. Low research start, default costs. 2000 point game, took "Propulsion Experts" Advanced Trait on the theory that Advanced Storage was too advantageous. I min-maxed only to the point of diminishing returns, e.g. Strength 80%, not 50%. Aggression and Defense default. Intelligence, Mining & Construction 120%, Maintenance 110%. Cunning, Farming, Refining, Political Savvy & Repair were all 80% as I recall. The rest were default or below, because somehow I raised enough points to take "Hardy Industrialists". Culture was Scientists.

It occurred to me later that starting with Medium Homeworlds may have been a mistake. I suspect the AIs are much less adept than a human at dealing with a lack of resources, hence the slow growth of the three AIs I've met so far.

>> You should play with at least low bonus

Well, I dislike AI "cheats", but if that's what it takes...

>> and yes, the AI vs human it's a challenging option

Good news.

>> but maybe you are ready to play in PBW and meet human players!

BWAHAHAHA! Good one, MB!

Seriously, so far I've been using Tactical Combat all the way. I know this gives the human a big advantage, but at this point I just can't stand for the game to run my precious ships, even under my general direction.

Maybe after I've learned more about Strategic Combat...

[Mephisto]

Hunpecked said:
Master Belisarius writes:

>> You should play with at least low bonus

Well, I dislike AI "cheats", but if that's what it takes...

The Terrans, EA and Narn for one are optimised for "no bonus"-games. They use the resources more sparingly which puts them at a slight disadvantage in games where there are AI boni since they don't use them to the full extend possible.

[Blade W]

Yes, it's a common way to stop AI-s even at high bonus: mines, mines, mines.

[Hunpecked]

At one point the Fazrah actually broke through a field with 20-odd mines, but lost their sweepers and a couple combat ships in the process; the survivors were annihilated in the ensuing battle. Going the other way I had to sweep 30+. Overall I thought the Fazrah handled mines pretty well for an AI. I'm currently swallowing frontier colonies of the Aquilaeians and Tessellate, and my sweeper-escorted fleets have so far encountered no mines. It would seem prudent for an AI to deploy a few minelayers to its borders before declaring war. In fact it might be prudent even in peace.

Next game I'll try mine warfare in a gang-up-on-the-human conflict.

[Master Belisarius]

Hunpecked said:
>> You should play with at least low bonus

Well, I dislike AI "cheats", but if that's what it takes...

Ok, this is the main reason because you had not a decent challenge...
All the AI races I have designed, are optimized to play with AT LEAST low bonuses (specially the Tessellate because it's a "combo" race).

Also, remember to select High Difficulty... because if not, the AI will not use all the ministers!

[Master Belisarius]

Hunpecked said:
I'm currently swallowing frontier colonies of the Aquilaeians and Tessellate, and my sweeper-escorted fleets have so far encountered no mines. It would seem prudent for an AI to deploy a few minelayers to its borders before declaring war. In fact it might be prudent even in peace.

They usually lay mines fairly well, even in peace. The problem would be that they're lack of resources...
Can you send me the savegame?

[Master Belisarius]

Aquilaeian vs Pyrochette.
The Aquilaeian won by points at the turn 500. It was a long and weird game: the Pyrochette conquered the Aquilaeian's systems in the left side, but lost his own systems at the right side.

Verduran vs Narn Regime.
The Narn Regime won by points at the turn 500.
In some way, I have not expected a Narn's victory, because although his ships are a lot better than the Verduran's ships, the Verdurans built many many ships and fleets (and the AI is not good defending when the opponent has many fleets moving into their systems).

[Master Belisarius]

Blade W said:
Yes, it's a common way to stop AI-s even at high bonus: mines, mines, mines.

True. But think it mainly works in the early game.
Because most of the strongest TDMs AI, include "offensive sweepers". In the late game, a pair of those offensive sweepers can manage 100 mines without big problems.