im not making any suggestions. just showing the games we played with other members using AI autopilot to try to see which is best on the 1v1 strat template. i'll be conducting research on other templates as well.
I like beeing distributed in 3 different places, like Wunderwaffe. But an AI is very 'predictable'. As you See the moves (or better: the most wanted territories) are often the same.
This plus the delay of cowzow lead to waffes loss of scandinavia. And made its first positional advantage useless.
That's why income and delay made the difference in this game imo.
(No idea if it makes sense at all. Just to Start discussion)
Are you using the Flash client (the website) or Unity client (the app) to test how 'good' AIs are? In Unity, Cowzow does decent orders, but in Flash, the doesn't build the orders well...
Just common knowledge forum advice! Stop making a new post for each game, edit your first post. sorted data is much more user friendly, than going through countless pages and spam posts to get relevant info. If this a research thread, then please let us know what is your research aim and so on?
You have to turn off the fog, for the bots to work properly. Reason is that the bots (at least, as Wunderwaffe was ported to WarLight) had to be stateless, so they have no idea which bonuses the opponent controls. Cowzow hiting Nigeria strong turn 10 seems to imply that he knows more however the next turn he kinda obviously has no idea about Wunderwaffe controlling South America and I'm pretty sure that a "seeing" cowzow which won the AI competition would have broken South America instead of this nonsense with taking the wastelanded bonus. Also this turn you see Wunderwaffe attacking the big stack of cowzow which turned out perfectly however the reason wasn't Wunderwaffe beeing super smart but Wunderwaffe believeing cowzow to be at base income.
