Evolution of a Revolution : Industrial Revolution 2

[BeyondCloister]

The original version of Industrial Revolution has now been placed in history and the hands of the open source mob.

After a brief amount of thought I've decided to go ahead and work on the full version of the game hand in hand with the board game version.

I present to the world Evolution of a Revolution : Industrial Revolution 2 (EoaR:IR2).

Well actually there is nothing to see yet but I can tell you that I've already implemented the basic versions of the following features:

* multiple players
* simple AI

There is a lot more work to do before a version is made public but judging from the voter feedback comments I know some of you would be interested to read about this.

[BeyondCloister]

The whole project has taken a sudden twist and change of direction.

I've added a message window that is displayed on the main map scree. This displays the end of turn messsages and the thoughts of the AI. In order to just sit back and watch the AI play I've added a timer and by passed the end of turn screen.

This means that it is now possible to play the game as a real time game instead of turn based. This is just a side effect of the AI tests but I may well leave it in.

All my development for the near future will be on getting the AI to play the game and not bother with player interaction. In fact 'm more interested in using the game engine for this AI than making an actual playable game.

[BeyondCloister]

I've been doing some more playing around with the auto play AI version of the game.

I've tweaked some of the rules that the computer player uses.

At present it has 3 goals:

1. Build a factory.
2. Buy transport (this includes replacing transport destroyed in accidents.)
3. Send transport to a city with auto return activated.

By changing the number of horse and carts that the computer is aiming to own it is possible to see the minimum amounts required to break even or make a profit.

So far it appears that it requires 5 carts traveling to the the nearest city to break even. Adding a sixth cart means it takes about 190 turns to have more money than at the start of the game.

I'll report back on further developments and discoveries in the world of IR as and when they happen.

[keith]

With the different cities having different values, would the nearest city generate the greatest economic outcome, or would a city a bit further away that pays more be better? Also if you consider the bigger picture would oversupply into one city reduce the value received?

Regards
Keith

[BeyondCloister]

With the different cities having different values, would the nearest city generate the greatest economic outcome, or would a city a bit further away that pays more be better? Also if you consider the bigger picture would oversupply into one city reduce the value received?

Currently my tests have shown that with the prices I have at present there is no difference between selling to the lower priced closer cities compared to the higher priced more distant cities. I will be fiddling with the prices and supply and demand based on how see the computer players getting on. Last night I had 3 computer players running all using different target strategies and yet they were all ending up running out of money.

[PowerMacX]

Time for a GA?

[BeyondCloister]

Time for a GA?

What is a GA?

[PowerMacX]

Genetic Algorithms (you can find a lot of sample Java code, a basic GA can be just a couple hundred lines)
Basically, they are very good at solving multi-parameter problems, not necessarily in an optimal way but very good anyway, and are problem-independent, ie: you can reuse the same GA for a wide variety of problems, without changes!

[BeyondCloister]

I know about them, just never put the term GA to them.