Many games include some kind of central market for buildings, abilities, items, upgrades, etc. I use the term "central market" to refer to a central supply of valuable items that are available for all players to buy.
The simplest type of central market would be one where all items are laid out at the start of the game and can be bought for a fixed price. This is exemplified by Dominion's cards and Puerto Rico's building tiles.
Other games have only a subset of items available for sale in the central market at any one time. This little bit of randomness adds to replayability, but it can take away from the long-term planning aspect of the game if players don't know what will be available in future rounds.
Here are some examples of games with randomized central markets:
1) Spyrium has a 3x3 grid of cards laid out at the start of each round, and players can buy them through a process of placing and removing workers. Cards are drawn from different decks in different rounds so they are guaranteed to come at the appropriate part of the game.
2) Power Grid has a few of its power plants laid out and a series of auctions for the available plants. (I find that Power Grid doesn't do a good enough job clearing out plants that nobody wants.)
3) Smallworld has a row of races, and players buy the one they want by placing coins on all the ones that are lower in order. These coins provide an incentive for players to take weaker races in later turns. New races are drawn at the top of the market, so start out expensive and gradually get cheaper.
4) Ascension, a deck-building game like Dominion, has a small number of cards in its central market that are constantly refreshed. There is little or no possibility of planning ahead because players have no idea what will become available in future rounds. This is randomness taken to an extreme.
In the game I am working on, I need to design a central market for the buildings and ships available for players to buy. Some competing goals I would like to achieve are (1) balance, (2) replayability, and (3) strategy. Balance is easiest to achieve when the prices of items in the central market respond somehow to players' behaviour as in Power Grid's auction and Smallworld's coin-placing mechanic. Replayability can be achieved with some kind of randomness. But if it's too random, it could come at a cost in strategy by making it hard to plan ahead.
What kind of central market mechanics have you used in your designs? Are there any that you find particularly clever or interesting?
Totally agree. My theme is different enough, I think. It's about the founding of Hong Hong after the First Opium War, so 1842 to about 1866, on the cusp of the modern age. The companies operating at that time, like Jardine Matheson and Co, were more like modern Fortune 500 companies than swashbucklers. So I'm making a game about banking and credit with a nautical theme on the side.