Epic Thirteen - ETHIOPIA

[Sullla]

Ethiopia players, please post your report links in this thread. (Either linking to your own site, or to your RB thread report here in the forum.)

REMEMBER: All discussions are supposed to take place in the sticky threads, not attached to any reports. We are asking everyone to use the "Threaded" display option to help organize responses to individual games. Thanks!

[Qwack]

Very nice game. I think Ethiopia had a many possible routes to victory that could have been taken, so it should be interesting seeing what the other players did.

http://rb.llsc.us/Epic13/

[Kylearan]

Hi Qwack,

The idea of different players starting with different civ's on the same map is a great one. This makes comparisons very interesting.

I agree, but I have to say comparison would be even more interesting if all players started on the same spot, but with different civs. We did this with Civ3 at least once, and I really hope we will do this again with CIV.

Instead of building these monster defensive units in this situation, he decided to spam chariots which would be tremendously easy to attack when they are in a city(Which was exactly where he kept them)

Justinian did the same in my game, which made me roll my eyes too. Even though his archers would have been "only" CG1, that would have been a lot stronger than whipping 3 or 4 chariots...

Very good game, and great report! I found it very interesting to read about how managing a cultural victory has changed a bit in BtS. And your (attempted) managing of golden ages was very good, too!

-Kylearan

[VoiceOfUnreason]

Instead of building these monster defensive units in this situation, he decided to spam chariots which would be tremendously easy to attack when they are in a city(Which was exactly where he kept them)

Presumably, this is the improved AI at work, which builds counters to your armies, rather than doing something actually useful?

[Swiss Pauli]

This is where the odds analysis breaks the 'build counter-units' plan: chance of defeating the spear is too low, so it doesn't attack, then builds another chariot, etc.

[LKendter]

My report is at the typical location - http://www.kendter.com

[Blake]

The AI is limited in that it has no notion of whether it's fighting a purely defensive war (where it's units literally only defend), or a defensive war where it has to offensively eliminate attacking units. It simply errs on the side of defensive wars which involve booting the invaders out, rather than dying slowly.
If the AI is forced into a "die slowly" war, it probably means that it's playing at too high a difficulty level, it should find an easier human to play against.

To put it as simply as possible: I don't care what the AI does once it has already lost the game. The AI always plays as if it hasn't lost the game. When playing at a difficulty where the AI has lost from the start, you tend to get poor AI performance, WHAT ELSE are you going to get?

That was basically a design decision: The AI always plays as if it hasn't lost the game. I decided it was not worth the effort to have the AI determine when it has lost the game and then to implement plausible sets of "spiteful strategies". I rathered spent my time implementing as many "winning strategies" as possible.
You know part of the problem is? Once a human has lost THEY JUST LEAVE THE GAME. And part of what I did was try to make the AI play like humans, but I DON'T KNOW WHAT HUMANS DO once they've lost the game but are forced at gunpoint to play on (ANY VOLUNTEERS?). That also makes it hard to implement credible strategies . Personally I think the "Apathetic response" is as good as any.
(AND DO YOU HAVE ANY IDEA HOW MUCH HATE I WOULD GET IF THE AI SLAVED ALL IT'S CITIES DOWN TO SIZE 1 TURNING THAT POP INTO DEFENDERS? IT MAY BE THE MOST USEFUL THING TO DO WHEN DYING, BUT GOOD GRIEF, I'D HATE MYSELF, THAT'S A LAMEEE SPITEFUL STRATEGY)

And yes there are some fringe cases where it's worth stacking up defenders and hoping the enemy suicides their army or goes away, but the AI is innately terrible at identifying fringe cases, it's strategic/tactical comprehension is extremely low resolution compared with humans.

[VoiceOfUnreason]

Epic #932: Bad Connections

This will be our first Epic using the new 5.961 patch; we'll be exploring the possibilities of Blake's new Bitter AI patch.

Options: Spite quitting has been enabled.

Rules: You must successfully conquer all 6 of your opponents. Should any of the AIs abandon the game prior to your victory, your game is forfeit.

[darrelljs]

Well done Qwack...I had read your manufactured research game but thought of it as more of a novelty than a strong tactic. I'll have to look for opportunities to do so in the future.

Darrell

[Atlas]

Well done Qwack...I had read your manufactured research game but thought of it as more of a novelty than a strong tactic. I'll have to look for opportunities to do so in the future.

Absolutely! In fact in multiplayer Civ building research/wealth is the difference b/w the very best and 2nd tier players (who go to 0% research by expanding and then building nothing but units in their new cities). I find that there is always a time when building research or wealth is worth while in every game (usually in the early game, when you can build a new city and for the most part make it at least research neutral by immediately building wealth/research). The exception to that research neutral thing is that one city (on standard it is like city 6 or 7 that cost like 2x as much as the previous one).