Slitherine

Is there AI in the game? When I select a game with human/AI I end up playing both sides.

6. CONTENT
---------------
- the build contains no active AI yet, and therefore you can only play single-player hotseat, or multi-player

As build's notes states, this build contains AI code, but inactive (as it was not ready to be activated yet). I need to do some more work on the AI code, and a first AI build should be ready at most on next Monday, and it will make its way into open beta after a few days of testing and adjustments.

Are you able to tell us anything more about the AI at this stage, Dan? Obviously experienced players would say that the existing AI is quite poor in terms of its playing ability and really only useful for learning the game or doing the occasional experiment. Glaring defects included picking more leaders than is necessary in smaller armies; getting confused by terrain; leaving all "poor" units up by the camp instead of using them as second rank support; and advancing skirmishers too far so they are easily cut off and destroyed. Have you been able to address these issues and others with the new AI, or are such modifications further down the pipeline? Thanks.

We've always said this: the release version of FoG(U) will be shipped with most basic AI, tactical ancient warfare being the niche market it is, we can not afford to spend too much funds on it at this point. Neither too much time. And both time and funds are required in order to design an implement a really good AI. Most probably it will be worse than FoG(RB)'s one at launch.

After the initial launch, work on AI will be a constant, and each new addon will bring with it a better AI code (among other things). My main intention (same as with the AWG project) is to bring a really good AI to this genre. FoG(U) is just a tool needing to be finished in order to be used for development and delivery of this.

cothyso wrote:After the initial launch, work on AI will be a constant, and each new addon will bring with it a better AI code (among other things). My main intention (same as with the AWG project) is to bring a really good AI to this genre. FoG(U) is just a tool needing to be finished in order to be used for development and delivery of this.

Well, that is really good to hear. I think what a lot of us "veterans" would like is a really tough level to be available eventually on the AI so we could practice all sorts of things in preparation for the tournaments - so maybe there could be two levels of AI, "standard" and "advanced".

Actually, I would like to make an AI which should beat the most of us (at least on the highest difficulty level) And I will also want to have some other AI surprises (for example, an AI server)..

cothyso wrote:Actually, I would like to make an AI which should beat the most of us (at least on the highest difficulty level) And I will also want to have some other AI surprises (for example, an AI server)..

Yes, that would be great if the AI was that good. I also think a much better AI would have a beneficial effect in terms of the participation in our tournaments. At the moment, I think what happens with some players is that they stay "solo" for quite a while and just play against the AI - but then when they do decide to enter a competition they find it really hard because nearly everyone is much better than the AI. So then they quickly drop-out from the competitions and we don't see them again. But if, in future, the AI is much better (and offers different levels of challenges) then newer players should be able to bridge the gap between single and multi-player more easily and hopefully this will mean they will participate in the tournaments for many years to come.