Random Game Ideas: AI Personalities and Technology

I’ve been kicking around two ideas on the game lately.

The first one is that I’d like to create some better defined personalities for AI players. Maybe some AI personalities would be easy AI players, while others would be expert AIs. For example, maybe one AI player could be bold and swift. He tends to leave fewer defenders around when he goes for an invasion, choosing to rely on “the best defense is a good offense”. Maybe he feigns attacks, and then attacks elsewhere. A different AI personality would be more cautious and chooses to attack when he’s got his own territory well protected and he’s very likely to succeed in an invasion. The idea is that each the AIs personalities would have more defined styles of play.

At first, I thought maybe there would be some “best” AI personality. And why not use the best one all the time? For one thing, that’s not very exciting because it lacks variety. If there’s one excellent AI and two good AIs, it might be good to add all three of them to the game. Even the uncertainty over which type of AI you’re playing against could put the human player at a disadvantage because he doesn’t know exactly which strategy to use against them, and can’t anticipate what types of moves the AI will use. From that perspective, not only would a handful of AI personalities be more interesting, but the mixture might make it more difficult for human players to win the game, as well.

Players who are setting up scenarios could even set different nations to use particular personalities – to mimic the historical situation. (Example: I want the leader of this nation to be cautious. I want the leader of that nation to be daring, quick, and unpredictable.)

The second idea I’ve been kicking around lately is an idea about redesigning the technology system. (I’ve actually been kicking around this idea for a long time, I’ve just never implemented it.) Right now, players upgrade along a predictable path from unit A to unit B. But, what if there were a whole bunch of technologies with levels. For example, what if there was an “armor” technology and an “engine” technology and a “cannon” technology? Players could decide to create a new type of tank. They’d spend some time and money on a new tank design, and they’d be allowed to specify what types of attributes they wanted it to have. For example, they want a heavy tank with good firepower. Or maybe they want a light, fast tank that is cheap to build. They’d then spend some money, and based on their nation’s technology sophistication and some randomness, they’d have a new tank design a few turns later. The tank design might vary in quality and price. Maybe it performs badly in certain types of terrain. Maybe it’s a very good tank for the level of technology that player has. The player could then decide to build it. The key thing here is that tank designs would vary in terms of their attributes. Players would have to decide whether or not the design should be used or scrapped.

Because other players are doing the same thing, each nation has units of different quality and attributes. Players have take that into account when going to war against other nations. It would also mean that players have to adapt their strategies to the current game. They might’ve had good tanks in the last game they played, but poor tanks in this game. Or, they might’ve had good tanks 50 turns ago, but technology in the game has progressed, and they haven’t funded their tank technology enough to keep up with the latest designs.

I could see this system being used with all the different units in the game: infantry weapons, tanks, aircraft, etc. Maybe the player could get a chance to research some anti-armor weapons and they’d come up with something like an RPG. Maybe the player needs to field test some designs to see how well they work in actual combat. I could imagine some of the attributes being things like: attack/defense numbers, movement rate, movement rate on different terrains, movement range (for aircraft and missiles), production and resource cost, maintenance costs, etc.

It might also be interesting to vary the costs and research time of technologies each time the game is played. Players would have estimates of the time and cost of researching a new technology, but it’s not exact. Some technologies in the real-world seem like they’re not far off, but they never seem to arrive. (Example: Fusion power.)

It would be nice to add tactics into the system as well. For example, some technologies are good for a while, but your opponent learns some new tactics and reduce its valuableness. Or maybe some of your technologies are given a boost because you learn tactics that improve its usefulness. Tactics are something that would be improved and refined through combat experience. Maybe if the player has military academies, they learn new tactics more quickly. The longer a war drags on, the more time the enemy has to learn tactics that mitigate your technologies.

4 thoughts on “Random Game Ideas: AI Personalities and Technology

  1. Pingback: Empires of Steel Developers Diary » Blog Archive » Random Game Ideas: AI Personalities and Technology

  2. I like most of the suggestions.

    However, if randomness is built in to new designs, it would be good if on average, the quality of new designs for one player is on a par with the average for each other player, even if you have to be smart enough to know which ones to go for (difficulty of implementing this aside.)
    “Maybe the player needs to field test some designs to see how well they work in actual combat.” This would be very important, one thing I would like would be to be able to just “place” a tank say, on a test island, so that you could place your new design there to try out combat. Along these lines, possibly the player could capture enemy units, and if they do this, their design becomes your design after some period of time. Maybe you’d need a certain amount of over-kill in combat to allow a capture to take place, or a plane would have to be shot down over land to capture, or allow a badly damaged unit to be more easily captured by strong units. Also, maybe production of a once-enemy design would be significantly slower for a while, then ramp up to standard. The capture idea could apply to the game as it is now, where a bb is captured by some overwhelming attack formula, so that the player who has not researched bb’s of whatever level it is could start to, as described above. Finally, for now, it might be good to be able to add or remove, at will, various designs from the build list, so that the list is easy to work with.

  3. just to add a tiny amount to the above, I meant that a player who has not researched bb’s of whatever level it is could start to immediately or delayed build those bb’s if one is captured. I guess, to follow up, you’d have to get your captured ship to a suitable port without it being sunk 🙂 in order for this to happen. You’d have to shoot the plane down over YOUR territory. Maybe the plane (or any unit, for that matter) would have to start out undamaged before combat for it to be captured. Maybe there would be levels of tech so that you couldn’t capture a level 4 bb and build it if you didn’t have a dreadnought. You could capture up to level 2 initially, without any previous research, then capture a level 4 if you did this.

    On the field test of designs or just combat trial, could we have a standard combat island with various types of terrain, quite a large area, or sea, with the ability to add any list of any units, on 2 opposing sides, to order them about, as in the game? To modify this, maybe only be able to put on the opposing lists the current games’ units which have met in combat and where damage has occured to the oppositions’ unit. or where a certain number of combats have occured with that unit. Finally, be able to put your own units on the combat range to fight each other, to speed up practical understanding of the mechanics of the game

Leave a Reply

Your email address will not be published.

Please note: if your comment doesn't appear right away, it's probably because it was automatically put into the moderation queue.