Interesting AI behavior and economics with 3 AIs in game
Posted: Sun Dec 03, 2006 8:25 pm
Since my Internet connection is flaky, I haven't been able to play online, so I've been playing the AI alot.
I was playing a multiplayer map, the rivers one based on the British economy. I had three AI's, and I was playing at the 'Mogul' difficulty (2nd from hardest I think).
One of the AI's developed a fairly good sized rail network. I myself also sold lots of stock early to grab ahold of the bomb making industry before anyone else got to it (coal -> steel -> bombs -> profit!) so the first AI and I held a large portion of the map.
The interesting thing was the other two AI's. Despite the ability to expand, neither of them went beyond two cities. They simply hooked up their original city with another town nearby and began trading back and forth. The astonishing thing is that they were worth as much as the other AI who expanded more normally. BUT, when it came time to buy out these AIs, the stock was just as expensive, but the payout for scrapping their railroad was peanuts, since there wasn't any infrastructure. So of course I kept them as little cash generators, and then I took a good hard look at how they were set up to see how they generated so much money with only one short route.
In general, it seems that over time, passengers completely rule. Even though he was only transporting passengers on a very short hop, he raked in tons of money. He also wasn't spending any on infrastructure expansion. He didn't even properly start up a bi-directional industry between the same two towns that was waiting to be had. When I completed that and tagged a few extra freight cars onto the passenger train, the income went even higher. When I bought out the other AI, it was the same thing. Passengers with one freight car on the back.
Sure enough, when I went to look at the full-screen train stats window (something I should look at more often) passengers always beat out the bombs and medicine in the end. Just for yucks I strung a long dedicated passenger route around the map to two distant cities, and ran it with an F-series diesel. Huge piles of cash for each car, but long delivery time despite the speed. I don't know which is more efficient in the end, long or short. But the AI's seem to do incredibly well with short routes. I guess passengers regenerate quickly when you are both loading and unloading in a given city.
in the end, the two-city AI empire is not enough to win or prevent being bought out (I won handily) but the technique is interesting as a sub-strategy.
My own strategy has been to aggresively build up the high profit industries (bombs, medicine, packaged goods, etc.) and run everything else secondary. This means a lot of up-front expenditures, but once the industry loops are complete, the money just rolls in. Perhaps I should look at passengers more seriously in addition, and not worry about long distance passenger runs so much.
-M
I was playing a multiplayer map, the rivers one based on the British economy. I had three AI's, and I was playing at the 'Mogul' difficulty (2nd from hardest I think).
One of the AI's developed a fairly good sized rail network. I myself also sold lots of stock early to grab ahold of the bomb making industry before anyone else got to it (coal -> steel -> bombs -> profit!) so the first AI and I held a large portion of the map.
The interesting thing was the other two AI's. Despite the ability to expand, neither of them went beyond two cities. They simply hooked up their original city with another town nearby and began trading back and forth. The astonishing thing is that they were worth as much as the other AI who expanded more normally. BUT, when it came time to buy out these AIs, the stock was just as expensive, but the payout for scrapping their railroad was peanuts, since there wasn't any infrastructure. So of course I kept them as little cash generators, and then I took a good hard look at how they were set up to see how they generated so much money with only one short route.
In general, it seems that over time, passengers completely rule. Even though he was only transporting passengers on a very short hop, he raked in tons of money. He also wasn't spending any on infrastructure expansion. He didn't even properly start up a bi-directional industry between the same two towns that was waiting to be had. When I completed that and tagged a few extra freight cars onto the passenger train, the income went even higher. When I bought out the other AI, it was the same thing. Passengers with one freight car on the back.
Sure enough, when I went to look at the full-screen train stats window (something I should look at more often) passengers always beat out the bombs and medicine in the end. Just for yucks I strung a long dedicated passenger route around the map to two distant cities, and ran it with an F-series diesel. Huge piles of cash for each car, but long delivery time despite the speed. I don't know which is more efficient in the end, long or short. But the AI's seem to do incredibly well with short routes. I guess passengers regenerate quickly when you are both loading and unloading in a given city.
in the end, the two-city AI empire is not enough to win or prevent being bought out (I won handily) but the technique is interesting as a sub-strategy.
My own strategy has been to aggresively build up the high profit industries (bombs, medicine, packaged goods, etc.) and run everything else secondary. This means a lot of up-front expenditures, but once the industry loops are complete, the money just rolls in. Perhaps I should look at passengers more seriously in addition, and not worry about long distance passenger runs so much.
-M
